Latest / The Vegan Report / 2025 Highlights: AI, MAHA, Vegan Nudges, Cultivated Meat | Science of Activism
Transcript
- 0:00Welcome to The Vegan Report, my name is Ryan
- 0:03and this is your special Wednesday Science of
- 0:06Activism episode. For today, a group conversation
- 0:10with two research scientists from Stanford University's
- 0:15Humane and Sustainable Food Lab, Seth Ariel Green
- 0:19and Jacob Peacock. We discussed some of 2025
- 0:24news and trends that affected animal advocacy,
- 0:29such as artificial intelligence, the election
- 0:32of Donald Trump, the rise of the Maha movement,
- 0:36the popularity of the vegan nudges concept among
- 0:40researchers, cultivated meat, and much more.
- 0:44Seth and Jacob also tried to convince me to start
- 0:48a cult. Don't miss that piece of the conversation.
- 0:53Please learn more about Seth and Jacob's work
- 0:56by visiting the episode's notes. A special mention
- 1:00of Seth's excellent substack regression to the
- 1:03meat. It is an excellent read. Follow the show
- 1:07for more Wednesday Science of Activism episodes.
- 1:10Share the episode with the vegans you know. Leave
- 1:13a good review. Any gesture of support counts.
- 1:18Thank you. Let's start this conversation by talking
- 1:21about some of the highlights of 2025 and I wanted
- 1:26to begin with artificial intelligence. That's
- 1:30one of the big buzzwords of 2025. There have
- 1:35been a few initiatives and things happening at
- 1:39the intersection of AI and animal rights. There
- 1:42was the Amplify for Animals initiative. I am
- 1:48going to release an episode with Thomas Richardson,
- 1:52who was one of the teachers for that initiative.
- 1:57I know of Ben Mazaya, who is training you know,
- 2:02chat GPD to become a vegan activist. So all great
- 2:06initiatives to support advocacy work and the
- 2:10work of animal rights organizations. And then
- 2:14a few comments, discussions on public forums
- 2:18and on podcasts about, you know, artificial general
- 2:23intelligence and how that could help resolve
- 2:29the problem of animal exploitation forever. So
- 2:32what are your thoughts on artificial intelligence?
- 2:37And what do you think we can expect for 2026
- 2:40with that technology? I saw this going my way.
- 2:48Um, you know, I'll just give a big copy out here.
- 2:50This is not my area of expertise or focus. Uh,
- 2:54but, but I can say that especially in terms of
- 2:57the effective altruist milieu, I have thought
- 3:00about the intersection of animal welfare and
- 3:02artificial intelligence. And one thing to think
- 3:06about here is maybe the distinction between the
- 3:09sorts of really valuable work that people are
- 3:11doing right now to get animal advocates to work
- 3:14more efficiently and embrace this potentially
- 3:15very productivity increasing tool. On the one
- 3:20hand, and then on the other hand, we have things
- 3:21like how do we make artificial general intelligence
- 3:25or superintelligence or whatever you might call
- 3:27it, A -S -I -A -G -I -T -A -I. How do we make
- 3:31that a transformative artificial intelligence?
- 3:34And these are all just debated terms of what
- 3:37exactly they are. How do you make that go well
- 3:41for animals? That's a really tough question.
- 3:45I think the first one is not easy, but it's relatively
- 3:49easy. You know, we got to train people to use
- 3:52an LLM to write the first draft of this marketing
- 3:55email so they don't spend an hour drafting it
- 3:57up. They spend 10 seconds getting the chat should
- 4:00be T to make a draft and then 20 minutes editing
- 4:02it. And they save some time. Time is money for
- 4:04animals or something like that. And they can
- 4:06help more animals doing that. Transformative
- 4:10AGI. That's. Really tough. I think it's really
- 4:14unclear how to make that go well for animals.
- 4:17From talking to people who do more work focused
- 4:20on AGI and AI safety, by and large, people don't
- 4:25think too much about what a post -AGI world could
- 4:29look like because it's just so mind -boggling.
- 4:31No one knows. No one's really sure of that at
- 4:34all. And people found it relatively intractable,
- 4:38is my sense. beyond your utopias and dystopias,
- 4:43like broad strokes. But I don't think people
- 4:44have many tools to distinguish between those
- 4:47different possible outcomes. And where that leads
- 4:50for animals, in my opinion, is more likely bad
- 4:53than good, unfortunately, which I fear may be
- 4:56a theme. I did listen to Seth's previous appearance
- 4:59on the podcast before this, and he ended with
- 5:01a bucket of cold water, was your phrasing. It's
- 5:04like, oh, wrong guy for the job here, too. One
- 5:12view here is that AI is going to take on the
- 5:16values that are sort of vaguely endemic to human
- 5:19society as it is now, which is not good for animals,
- 5:24or as it is when AGI is created. Obviously, that
- 5:28could be not that far away, and it's relatively
- 5:30unlikely if AGI is 50 years away or even 100
- 5:33years away. The chances of us making massive
- 5:37moral changes for animals in the next 50 years
- 5:39that would change how AI operates a lot seem
- 5:42relatively low But all of that's quite speculative
- 5:47quite speculative Yeah, so after you have anything
- 5:51anything you should add to add their Contradict
- 5:54yes, I've decided that because I'm on a podcast.
- 5:57I'm going to give the un -nuanced version of
- 5:59what I believe The Bloomberg journalist Zeke
- 6:06Fowkes had this really funny line about long
- 6:08-termism in general, which is like trying to
- 6:10forecast into the future and think about what
- 6:13is going to make that work for people, is that
- 6:16he said, this is regarding Sam Beckman Fried,
- 6:19that it allows each of us to be the star of our
- 6:22own sci -fi story where like our choices can
- 6:25change the world because you multiply the consequences
- 6:28by like the continuity of all existence. I think
- 6:32that's a very nice way of saying the phrase,
- 6:34but the phrase intellectually masturbatory really
- 6:37gets right to the point of what I think he's
- 6:40getting at. I think AI is cool. I use Claude
- 6:44like every day to do stuff that I don't feel
- 6:47like doing. Like, sometimes, I mean, increasingly,
- 6:51I think the answer to that is going to be like,
- 6:53my research is going to be like, well, why am
- 6:56I still here? I wonder that question about as
- 6:59researchers. As to what some kind of transformative
- 7:03AI system is going to do for animals, this is
- 7:07like a category error to me. It's not even the
- 7:09right, it's like the wrong terms are being applied
- 7:12that doesn't even make sense. Like describing
- 7:15one idea is taller than another, like no, ideas
- 7:17don't have height. What's going to happen when
- 7:20AI happens? There's no evidence that that's going
- 7:24to happen. There's just this projection stuff
- 7:27and then how will that affect animals? anything
- 7:29is possible. So I don't understand it. I don't
- 7:32understand why people find it so interesting,
- 7:34except for the previous sociological story I
- 7:36just told. And you mentioned a bunch of stuff
- 7:40at the beginning of this, like people who are
- 7:42working on this. I have to confess, I haven't
- 7:44heard of any of those people. I'm sorry, I just
- 7:46really don't think this is the right thing for
- 7:49us to be focusing on. Not that you shouldn't
- 7:52have asked us that question. Thank you for giving
- 7:54me the chance to cut this off my chest, but like,
- 7:57I just don't get it. Interesting. Have you heard
- 8:01of the organization, Science Futures? No. They
- 8:07tackle with the issue of AI becoming sentient
- 8:13and the concerns around that. Have you given
- 8:18any thoughts to that? Beloved internet weirdo
- 8:25Brian Thomasick wonders if we should be killing
- 8:27NPCs in video games. They have some utility.
- 8:31I actually think this is a relevant consideration,
- 8:33which is like, do we know that AI is not already
- 8:36sentient? Do we know that the next generations
- 8:38of video games characters will not be sentient?
- 8:45I don't see any way to contribute to this question
- 8:51meaningfully until we observe more facts. What
- 8:57about you, Jacob? Are you polite with your AI
- 9:01when you use it for work? Okay, so I've actually
- 9:06had explicit arguments with friends about this
- 9:08because some people say that I'm not polite with
- 9:10my AI when I work with it for work. I'm very
- 9:14direct in most avenues of my life, and I'm especially
- 9:19direct with AIs because I don't think it... causes
- 9:24much harm if i don't say please and thank you
- 9:26necessarily to an ai in fact it benefits me by
- 9:29keeping me under my token limit uh so no i'm
- 9:33not like uh scrupulously polite i'd say uh am
- 9:37i like screaming at claude cursing them out when
- 9:40they do things wrong uh no no but uh i definitely
- 9:44am direct and talk a lot and we'll say things
- 9:46like stop talking say this instead uh that i
- 9:50would say to a person yesterday i said to claude
- 9:52what the fuck, you can't do that, because it
- 9:55was making some stuff up for a grant application,
- 9:57which is like, OK, you can't. And then the response,
- 10:00you can see it thinking it's like, the user is
- 10:02absolutely right to be furious. What a sycophant.
- 10:05I mean, good Lord. So apparently it likes it
- 10:08when we're mean to it, so who cares? Yeah, so
- 10:13I guess to the point on AI sentience, Again,
- 10:19not, you know, speculating here and outside of
- 10:22things that I really have too much familiarity
- 10:24with. But is it possible? Probably is my sense
- 10:31here. At least it gets very fuzzy and hard to
- 10:34think about. If we think that pretty simple organisms
- 10:39are sentient or have a chance of sentience, you
- 10:43know, shrimp are pretty complicated, probably
- 10:44sentient. but we can keep going down the ladder
- 10:48to organisms that seem less and less likely to
- 10:50be sentient, different types of mollusks, cephalopods,
- 10:53probably sentient, sea hares. Not clear how sentient
- 10:59they are. What's a sea hare? It's kind of what
- 11:02it sounds like, but it's a type of roaming mollusk
- 11:06that, oh, sorry, a hare like a little rabbit.
- 11:09It's kind of a big blob. I'm holding this up
- 11:11for the camera that won't hear this. Yeah, so
- 11:15it's kind of a rabbit -shaped blob tends to be
- 11:17darker in coloration that crawls around the ocean
- 11:20floor eating things up So If we think that it's
- 11:28possible that really simple organisms have some
- 11:31degree of sentience and sentience probably isn't
- 11:33a binary it's probably scale and Then we can
- 11:37make really complicated machines artificial intelligences
- 11:40that perfectly simulate one of these animals,
- 11:46are they then making a digital sentience? Not
- 11:52clear. I don't know. But it seems like a distinct
- 11:55possibility that this could happen. Although
- 11:58it's not clear that it's coming soon. And it's
- 12:00not clear to me that it's likely. Okay. Another
- 12:04big highlight of 2025 was, of course, the election
- 12:09and the election of Donald Trump. I know you
- 12:15have written about the Maha movement, Seth. I
- 12:21feel like the vegan movement as a whole has not
- 12:24yet considered seriously how to do outreach to
- 12:30conservative audiences. And maybe it's a bias
- 12:34since most of us identify as liberals, according
- 12:39to the studies out there. and define our veganism
- 12:45or our compassion for animals as being something,
- 12:50you know, politically on the left. I've seen
- 12:53that too. So what do you think of the idea of,
- 12:58you know, reaching out to conservative audiences
- 13:02and making the case for plant -based living to
- 13:07those, to conservative people? Sure, sure. Broadly
- 13:13speaking, I'm all for it. I have no qualms with
- 13:16people working with anyone to encourage them
- 13:19to eat more plant -based or animal -based products.
- 13:23There's always going to be an effect of this
- 13:25question, but from the evidence we have, my sense
- 13:27is that it's not that strong a moderator. There's
- 13:31a couple of studies that use that as a moderator.
- 13:33My recollection is modest effects. And yeah,
- 13:40generally all for it. For our non -scientist
- 13:42listeners, a moderator effect is like, would
- 13:46it be stronger for this group than that group?
- 13:48So the question is, does an appeal to plant -based
- 13:51dining work better on liberals than conservatives?
- 13:55There is, as Jacob said, some very messy evidence
- 13:57on this. I don't really know what to think about
- 13:58it. I think it's true that in general, liberals
- 14:01are more likely to be vegetarian. I think vegans
- 14:04are weird enough that I'm actually not sure I
- 14:06have a good sense of what our politics are overall.
- 14:09In America, the fastest growing vegan segment
- 14:13is African Americans. Jennifer Chanin told me
- 14:17that recently. Very interesting. African Americans
- 14:21also tend to have more socially conservative
- 14:23views on questions like gay marriage. So there's
- 14:26like something that's like maybe there might
- 14:29be, if we really dug into this stuff, real vegans,
- 14:32as opposed to vegetarians, we might discover
- 14:35it's less of a liberal monolith than we expect.
- 14:41But in general, I said I was going to do this,
- 14:46that I was going to start reaching out to Maha
- 14:49Podcasts. And my basic reasoning was that I think
- 14:53a general distrust of orthodox thinking about
- 14:58medicine and health is a really good starting
- 15:01point for thinking about what's wrong with factory
- 15:04farming. I think we could also make, especially
- 15:07on impacts on communities, I think we could also
- 15:09make a really good case that there is a proper
- 15:15relationship between humankind and animals that
- 15:19is custodial and guardian -like, and that we've
- 15:24really messed that up. This isn't a novel idea.
- 15:30There's a conservative writer named Matthew Scully
- 15:32who wrote a book called Dominion, not affiliated
- 15:34with the documentary that makes this case that
- 15:36we have these obligations to animals that we're
- 15:39under by God that we're not fulfilling. And that
- 15:43doesn't seem to move the needle much. So what
- 15:46are we going to do? I don't really like... Sure,
- 15:49it's promising. I just don't have any concrete
- 15:51ideas about how to make this happen. And I think
- 15:54the... general obsession with protein, which
- 15:57I would call one of the big 2025 trends pushes
- 15:59in the other direction. I would add and thank
- 16:02you, Seth, for flagging on my on the jargon there.
- 16:04I will try to keep things layman friendly. I'd
- 16:12add there that the there's always this kind of
- 16:16trade off in designing meat reduction interventions
- 16:19where you can target those people who you think
- 16:22are most propents to change and most likely to
- 16:24change. In this case, you might say, oh yeah,
- 16:27we're going to try and work with liberals to
- 16:28get that, because they're already interested
- 16:29in the climate. They don't eat as much meat already,
- 16:35so maybe they're more interested. If you do that,
- 16:40you might get less impact just because there's
- 16:43less impact to be had. They already weren't eating
- 16:45that much meat. So even if you eliminate all
- 16:48their meat consumption, you won't have that much
- 16:50impact. Well, we should really go for the high
- 16:54value targets. There is a study a few years ago
- 16:57that suggested an 80 -20 Pareto distribution
- 17:01where 20 % of people eat 80 % of the red meat.
- 17:04And there are some people that just eat beef
- 17:05on beef on beef on beef. So he's trying to get
- 17:08these 20 % of people to change. And that's where
- 17:12you get the opposite problem, where it's like,
- 17:13sure, it's the most magnitude of effect to be
- 17:15had, the most impact to be had, focusing on those
- 17:18people. but they're probably the most intransigent
- 17:21they're eating beef for three meals a day so
- 17:23it's kind of be potentially quite hard to get
- 17:25them not to eat beef for three meals a day very
- 17:28hard to imagine that eating beef three meals
- 17:31a day yeah i've had yeah i've had i ate me at
- 17:34one point in my life but like a hamburger in
- 17:36the morning i guess steak and eggs or whatever
- 17:41you can like you could figure out a way to put
- 17:42it in but Well, to be clear, I was being facetious.
- 17:45I don't actually know that the study documents
- 17:47people eating, but the numbers are such that
- 17:49that would almost have to be true because I think
- 17:51it's like my recollection is it's like 20 % of
- 17:53people eat 50 % of all beef. The average American
- 17:56eats like almost like two thirds of a pound of
- 17:59meat per day. So if we have someone who's eating
- 18:01a lot more than that, that means they're probably
- 18:03eating more than a pound of anyway, this is a
- 18:06hard problem. I don't really have a good like
- 18:08anthropological theory of people who are that
- 18:11most serious carnivores. yeah except for the
- 18:13Jordan Peterson types who are like really they
- 18:16really are eating steak three times a day on
- 18:20the like outreach to conservatives more generally
- 18:22i'll say that i do think there are some reasonably
- 18:25promising ideas that people are pursuing here
- 18:29and i say reasonably promising in so far as we
- 18:32as a movement basically have almost nothing currently
- 18:35it's like very very very minimal no one's doing
- 18:37anything so i know someone's just at uh conference
- 18:43doing some tabling there just to get the word
- 18:47about animal welfare out and uh indicated they
- 18:49had a lot of success um you know and go ahead
- 18:53and white coast white coat waste as well i think
- 18:56that they're doing like getting rid of animal
- 18:58testing which is like maybe they're on ramp yep
- 19:01yeah using a kind of libertarian approach of
- 19:03like this is wasteful government spending and
- 19:05we should you know eliminate this waste um And
- 19:09to be clear, I'm not unaware of the pitfalls
- 19:14of this sort of work, especially Turning Point
- 19:16USA is a pretty controversial organization for
- 19:20good reason. And we should be wary of supporting
- 19:24them in some ways. But I think that is a consideration
- 19:29that one has to measure, I think. Yeah. Shout
- 19:33out to the Wilberforce Institute. They were the
- 19:36ones. doing activism and the turning point events.
- 19:42Recently, I think they had a big summit and they
- 19:44were present talking about veganism and animals.
- 19:49If I'm an investor and I have some money in alternative
- 19:54proteins in specifically lab grown meat, the
- 20:00results of the last election would make me feel
- 20:04a bit anxious. Do you think the maha crowd is
- 20:09going to be an obstacle for the introduction
- 20:12of lab -grown meat to the diet of everyday Americans?
- 20:18Yeah, my sense, and I haven't looked into this
- 20:21that much, but my sense is absolutely, I die
- 20:23hard to oppose it, right? Yep. I mean, America
- 20:29is not the only market though. Like in many ways,
- 20:33entrepreneurs tendency to focus on America because
- 20:35we're wealthy, which makes sense because if you
- 20:38were like selling a, um, like a startup or whatever,
- 20:43you know, a software, but everybody eats. I would
- 20:46say take this to the, take this to like Israel
- 20:48or Singapore and try to build from there. Yeah.
- 20:53And Singapore has been discussed especially for
- 20:55their limited domestic agricultural production.
- 20:59So this would be like a self -sustaining thing.
- 21:02Yeah, exactly. Yeah, I don't really understand.
- 21:06I mean, I get, I don't like, what is it? It does
- 21:08still take. Okay. So you're building your artificial
- 21:12meat company. You're like soil and disgusting
- 21:15thing or whatever. And it's like a, what? A huge
- 21:18vat? Is it like in a skyscraper? That's terrible
- 21:20marketing for cultivated meat. You're building
- 21:22a brewery full of rich, nutritious. Is it like
- 21:26a brewery? And that's the metaphor that's most
- 21:29often used. I'm gonna try and reverse that metaphor,
- 21:33but... Okay, it is fundamentally a... Actually,
- 21:37well, it's a fermenting process, so that makes
- 21:39a lot of sense. You just have to have, like,
- 21:40some kind of exposure to sugars and whatever
- 21:43is happening. Well, the cell culture process
- 21:46is different from the fermentation process. There
- 21:48are multiple approaches for making all proteins.
- 21:51Okay. Yeah, I know. That's a little... I'm thinking
- 21:55of... Okay. And neither here or there but yeah,
- 21:58my point is that Singapore probably still wouldn't
- 22:00have the space to be self -sufficient on the
- 22:02food front Even if it did do this, it would probably
- 22:04help but you're still gonna anyway still gonna
- 22:07take land. I Would guess that as well I certainly
- 22:12haven't yeah, I just saying you could vertically
- 22:14you can go up better Yeah, yeah Yeah, I certainly
- 22:20haven't seen and although I wouldn't be shocked
- 22:22if like GFI or someone has an analysis out here
- 22:24of like, you know what would be the land use
- 22:28footprint for cultivated meat, hypothetically.
- 22:32But I will say for my part, I'm just not super
- 22:35optimistic about cultivated meat in general.
- 22:38Obviously, I would prefer that Maha not block
- 22:42people from selling cultivated meat, even just
- 22:45on free market reasons. Not that I'm a die -hard
- 22:48libertarian by any measure, but yeah, I generally
- 22:50endorse the idea that people should be able to
- 22:52sell. things if they want to sell them and there's
- 22:55not some obvious reason they shouldn't be able
- 22:57to And obviously if it can help animals that's
- 23:01much much better. They should definitely be able
- 23:03to sell those things that they help animals And
- 23:06I think cold baby could do that, but not sure
- 23:08it's terribly likely that will do a whole lot
- 23:11Jacob do you want to tell us about three? Commonly
- 23:13cited reasons that you do not believe will be
- 23:15sufficient for takeoff of cultivated meat Well,
- 23:19that's so funny that you said that. How many
- 23:21was it again? What a coincidence. So some folks
- 23:25might have heard of this and actually wrote this
- 23:27paper focused on plant -based meats. But I do
- 23:30think they probably apply somewhat to cultivate
- 23:32meats as well. There's this pervasive idea in
- 23:36the alternative protein space that if only we
- 23:39could achieve plant -based alternatives. that
- 23:43competed on price, competed on taste, and competed
- 23:46on convenience with animal -based meats. If we
- 23:48could hit those three criteria, something really
- 23:51good would happen. We get a lot of people switching
- 23:54over. We get a really bulk shift. The details
- 23:59vary from source to source, but the implication
- 24:02and the way it's presented is if we hit this
- 24:05trifecta, then we kind of win. start eating plant
- 24:12-based meat alternatives, cultivate meat, plant
- 24:14-based meat, fermented, and stop eating animal
- 24:16-based meat. And I did a review of this thesis
- 24:21conjecture to try and understand the evidence
- 24:25supporting it. And unfortunately, the evidence
- 24:28is really quite weak, and a lot of the evidence
- 24:30in those sorts of scenarios suggests that the
- 24:35effects might be quite small. to give a really
- 24:38brief overview of that. Fundamentally, there
- 24:40is this idea that was put forward in various
- 24:43places that human food choice is mostly dictated
- 24:47by price, taste, and convenience. That's sort
- 24:49of the 80 -20 or 95 -5 of human food choice.
- 24:56But empirically, that's just not really true.
- 24:59Conceptually, it doesn't make a lot of sense,
- 25:00conceptually. But most importantly, Empirically,
- 25:05that's just not the case. There's just so many
- 25:07factors that go into human food choice. It's
- 25:09pretty complicated It's got a lot of interactions.
- 25:12It's not really this kind of straightforward
- 25:14economical rational human process It's just a
- 25:18much more complicated situation Okay, but allowing
- 25:22that maybe there wasn't like great foundational
- 25:23evidence for this sort of theory We can look
- 25:27at what would happen Do we have some tests hypothetically
- 25:30or in like some conscribed environment if we
- 25:33did provide these alternatives that were taste
- 25:35competitive and priced the same and about as
- 25:38convenient. And across the board, when we look
- 25:41at this in the more optimistic studies, which
- 25:44you might talk about the ways they're more optimistic
- 25:46in a bit, you're looking at like maybe 15, 20,
- 25:5125 % of people eating those alternatives. Maybe.
- 25:56But then in the More realistic studies that reflect
- 26:00what happened in a dining hall like what people
- 26:03actually chose at a university dining hall Even
- 26:06with a bunch of advertising a bunch of climate
- 26:08-related promotions You could be looking at just
- 26:11a few percentage points or or no effect at all
- 26:14in terms of do you have in mind when you're Saying
- 26:16that which studies Oh specifically mal in 2020
- 26:22Malin, it's in the 20s, is the big one. M -A
- 26:27-L -A -N, right? M -A -L -A -N, yeah, we can
- 26:30put it in the show notes. And that's the big
- 26:32one that uses ground impossible meat in a UCLA
- 26:37dining hall to test. And then we have some of
- 26:41these, they're called hypothetical discrete choice
- 26:43experiments. I'm not gonna rattle off the whole
- 26:47list. I can put the whole paper in the show notes.
- 26:51Or if you want to talk about any of this specifically,
- 26:53Seth, I'm happy to rattle off the list if that's
- 26:55helpful. No, no. I did have a couple of questions,
- 26:57though. It's interesting, first of all, that
- 26:58you're talking about. You initially distinguish
- 27:00between plant -based. By the way, for our listeners,
- 27:03this is like Jacob and I talk about. This is
- 27:05what we talk about. You distinguish between plant
- 27:08-based meats and cultivated meat. But my recollection
- 27:10is at least some of the survey evidence you pointed
- 27:13to in the study, like the survey of Swedish people,
- 27:16was like blurry on this question. it asked them
- 27:21like, what would you eat something that was like
- 27:23a like a fossil? I can't remember the specific
- 27:26wording, but I don't think like, do you think
- 27:27the evidence is really clear on this distinction?
- 27:32Some of the evidence is some is not. Yeah. Like,
- 27:35for example, there is one study in there, I believe
- 27:37Van Loo, where they actually included a cultivated
- 27:41meat in the choice set. So it's like, Oh, would
- 27:43you want this cultivated burger, this animal
- 27:45based burger or this beef burger? And in that
- 27:48case, obviously, it's not clear at all. And the
- 27:51way it was reported, I wasn't able to tease out
- 27:53just the effect of the plant -based burger. So
- 27:55I was optimistic. I was like, imagine there's
- 27:57this additional option. And I'll just give that
- 28:00for free, if you will, in terms of... But yeah,
- 28:05the Carlson study, I don't remember the exact
- 28:08phrasing, but it's possible it's kind of ambiguous.
- 28:10I forget to. My guess would be I can probably
- 28:13pull it up because I have this paper. at hand.
- 28:16My guess would be that most participants would
- 28:18not interpret it that way. Yeah, they're going
- 28:21to think like some fake meat, but not like some
- 28:23science experiment. My other question is, are
- 28:28you making a claim about absence of evidence
- 28:31here, or do you believe this is an evidence of
- 28:33absence, or sorry, EOA or AOE? Yeah, so just
- 28:39to give some background here, what Seth's asking
- 28:42is, am I claiming that we don't have evidence
- 28:45that if the plant -based meats were at price,
- 28:49taste, and convenience competitiveness, they
- 28:50wouldn't displace animal -based meats? Or am
- 28:53I claiming that we do have evidence that if these
- 28:56conditions were achieved, not much would happen?
- 29:00Yeah, it's a question of do we not know or do
- 29:02we know that this would not work? Yeah, and I
- 29:04think it's closer to the latter. I'll add the
- 29:06evidence I adduces in the shape of the ladder.
- 29:11Maybe that answers the question right there.
- 29:13I adduce a bunch of evidence, so yes. There is
- 29:18evidence to this effect and across the board,
- 29:22I think the evidence points. I generally just
- 29:25use majority, so 50 -50 as the threshold of like,
- 29:30I'd say certainly in some of the ways this kind
- 29:31of conjecture has been phrased it sounds like
- 29:34it would be a lot more than half so half felt
- 29:35pretty conservative to me and the evidence we
- 29:38have pretty strongly suggests that the effects
- 29:40would be under half yeah but you and i both know
- 29:43that when people make those kinds of claims they're
- 29:45being deliberately slippery right they're saying
- 29:48like a cascade but a cascade isn't a number so
- 29:51it's actually kind of an it's like a hard thing
- 29:53to like you're you're right to try to say like
- 29:55what is the threshold where we would count it
- 29:57but like for us I would say like if 20 % of people
- 30:01converted to this kind of the plant -based analogs,
- 30:04I'd be stoked about that. Absolutely. That would
- 30:06be like, yeah. Yeah, I would be very happy with
- 30:09that. And to your point of the cascade, by nature
- 30:12of having some evidence on this and direct evidence
- 30:15at that, or more direct evidence, the paper really
- 30:18focuses on current consumers, right? Because
- 30:20for want of crystal balls, I'm unable to say
- 30:25what future consumers would do. But that's not
- 30:27to say that there's not ways to collect evidence
- 30:30that might inform our expectations of future
- 30:32food chefs. But I think that's quite a bit more
- 30:37speculative than what we have now. I think I'm
- 30:41more optimistic on the PTC thesis than you are.
- 30:44But there is a sense in which what I understand
- 30:47to be the contribution of this paper is that
- 30:51a lot of these claims are really maximalist.
- 30:54And let's say, what kind of effect sizes should
- 30:57we expect? you know, as I've said in other contexts,
- 31:01like one to two percentage points at the scale
- 31:03of reductions in meat at the scale of Costco
- 31:05is a big deal, but it's not like radically changing
- 31:08factory farming. But if this were like, I would
- 31:12say like, if we achieved PTC price, taste, convenience,
- 31:18parity, and of course, by the way, the T is the
- 31:22tricky one because taste is a matter of like,
- 31:24I don't like it, even if it passes a blind taste
- 31:26test, people still might say, no. Oh, interesting.
- 31:30We should get into this. Go ahead. I disagree.
- 31:33Taste is not just this. It's not just your mouth.
- 31:36It's like, I have a taste for authenticity or
- 31:38whatever. That's what I believe. I would expect
- 31:4310 % to 20 % shifts within 10 years of that.
- 31:48Wow, 10 % to 20%. Yeah, that's right. If parity
- 31:50were, if we really had parity, but I just think
- 31:53that parity is like this. And I think people
- 31:55will, you know, like how many, do you know anybody
- 32:00who still drives a stick shift? Cause they like,
- 32:02like a feel of it. And it's like a couple of
- 32:04them. Yeah. It's like, that's like a thing. It's
- 32:06like, is it, is it easier? Is it better? Like,
- 32:07no, you're a lot more likely to still on the
- 32:09highway, but like people like it. Okay. So the
- 32:12thing that I said that I disagree about is I
- 32:14really think that the convenience is the more
- 32:17slippery one. I think we actually probably agree
- 32:21with issues with taste, but we should talk about
- 32:22it more because I really liked some of what you
- 32:25wrote in one of your taste regression to the
- 32:27meat posts. But the convenience one is really
- 32:32slippery. And it's funny, one of the reviewers
- 32:34of the paper complains that I like to complain
- 32:37rightfully complained. Were you one of the No,
- 32:45no, I didn't they would not ask me to review
- 32:46your I was gonna say it wouldn't make any sense
- 32:49It's like although it's like not that big a field
- 32:51it could happen someday to us. Oh, definitely
- 32:54But but anyway, I'm not complaining about that
- 32:57as I'm happy I made the edit I'm telling an anecdote
- 33:00here about how helpful the reviewer was. So hopefully
- 33:03they won't be offended So they're like, yeah,
- 33:06you really need more citations about convenience
- 33:08and what it means to be being in and it's like,
- 33:11well, part of the problem here is no one's saying
- 33:13what's it mean to achieve convenience parity.
- 33:16And lo and behold, someone also ran a systematic
- 33:19review on the definition of convenience in food
- 33:21choice studies, and came up with the same situation
- 33:24where they're like, yeah, people just say convenience,
- 33:26and it means all sorts of things. So I think
- 33:30like, when to get to like a croc Seth on like,
- 33:34Oh, do you think like, believe it more than me?
- 33:36Or do I believe it more than you? The PTC hypothesis
- 33:40that is you really have to get into these like
- 33:43really annoying brass tacks of C and T especially
- 33:46but also P where it's like convenience does that
- 33:50mean like one feasible interpretation if it's
- 33:53equally convenient anywhere you go in the world
- 33:56and you want to buy a piece of meat right next
- 33:58to it is like the same product but plant -based
- 34:03And if that's your definition of convenience,
- 34:04that's like a huge change in the world. That's
- 34:07like the entire supply chain has doubled. There's
- 34:09a extra supply chain for plant -based meat. There's
- 34:12a plant -based meat analog product for the hundreds
- 34:15and thousands of meat products that are in the
- 34:17world and animal products that are in the world.
- 34:19Okay. So I think what people probably actually
- 34:21mean when they say convenient is like, eh, not
- 34:23too much work. Not like convenience parity, but
- 34:25like my personal informal definition of not too
- 34:29much work. Yeah. I agree that's what they mean,
- 34:33but I don't agree that's like sufficient to actually
- 34:36stake a disagreement on because like Oh to me
- 34:38convenient is every restaurant I go to has it
- 34:41Otherwise and again, I'll have like search around
- 34:43for the right restaurant just today I left one
- 34:45grocery store because they didn't have me on
- 34:48me and I went to another like it's that's inconvenient
- 34:50So we're not at convenience parity. Yeah, so
- 34:54I Was gonna ask if either of you has read Warren
- 34:58Velasco's food the key concepts. No All right.
- 35:03This is a book whose main, one of whose theses
- 35:06is that the main drivers of food are not PTC,
- 35:11but rather identity, convenience and responsibility,
- 35:14with identity and convenience being the big two.
- 35:18If I might read, Velasco defines convenience
- 35:21as, convenience encompasses variables such as
- 35:23price, availability and ease of preparation,
- 35:27which are all related to the requirements of
- 35:28energy, time, labor, and skill. In other words,
- 35:31convenience involves concerns such as, can I
- 35:34get it and can I afford it? So for him, this
- 35:38is like, I actually think that that's probably
- 35:40the right way to think about convenience in the
- 35:42sense of price is part of like, eh, how easy
- 35:46is it to get this versus like, what do I have
- 35:48to give up to get it? Out of curiosity, Has anyone
- 35:56of you ever had lab grown meat? No, no, interesting.
- 36:04No, never. Um, okay. There was a moment when
- 36:09it was like available at mission Barnes grocery
- 36:11in Oakland. But I talked to someone who worked
- 36:13at or sorry, yeah, I think it was someone who
- 36:16worked at one of those companies. And they said,
- 36:18like, this is all just for show. Like they can't
- 36:22produce it at enough rate. Like it would be hundreds
- 36:24of dollars per pound if they were actually putting
- 36:26it in the grocery store. So this is just like
- 36:28a marketing thing. Oh yeah. I sort of took that
- 36:32as a given. I didn't know that. I thought it
- 36:34was going to be like dozens of dollars, like
- 36:36something I would like maybe buy. Oh, I see.
- 36:38I see. Yeah. But if I had, you know, if I was
- 36:41in charge of that whole enterprise of introducing
- 36:46lab grown meat, I would send samples to researchers
- 36:51at the front line of studying the effects of
- 36:55my product on markets and people. Yeah, I feel
- 37:00it's not smart to not have you partake in that
- 37:09experience of tasting their product. It might
- 37:15be because they actually know that people will
- 37:17think that it is not achieved taste parody Yes,
- 37:24interesting Okay, um this is a very fascinating
- 37:30and clarifying conversation because I've been
- 37:35torn all throughout, you know this past year
- 37:38about this issue. I actually released an episode
- 37:42uh, which was a debate hosted by, um, poultry,
- 37:48poultry concerns, something like that. Um, you
- 37:51might know the organization, they hosted the
- 37:53poultry concerns United. Yes, exactly. Um, did
- 37:57the debate, you know, are you for or against
- 37:59a lab cone meat? And I, you know, reposted that
- 38:04debate and there are, you know, actually a big
- 38:08chunk of the vegan uh, community and vegan activists,
- 38:12animal activists who do not believe in lab grown
- 38:15meat. Um, for ideological reasons, I don't think
- 38:21that, I think they are skeptical about, you know,
- 38:25the success of the product, uh, in the market,
- 38:28but, um, at the end, they don't believe in that
- 38:32as a solution, they believe in, you know, changing
- 38:35minds. And that's how in converting people to
- 38:40a vegan lifestyle and not in the market and the
- 38:46economy, making the shift that we all need to
- 38:50see happening. What do you make of their position?
- 38:57That's dumb. I don't have a That's obviously
- 39:02dumb like no wrong, but like luckily this is
- 39:06that this is like I even if it's a mainstream
- 39:09opinion among vegans it's like It only matters
- 39:14if someone who matters says it like when the
- 39:17you know new all person in the office of tenant
- 39:20protections in New York City says that like Private
- 39:23property is like a white supremacist tool. That's
- 39:24like, whoa This isn't this is like left Twitter
- 39:27and it's like part of policy If some vegans get
- 39:31into power and start saying this thing, which
- 39:33again, I think is dumb, then it's a problem.
- 39:35Until then, it's like, okay, whatever, they're
- 39:36talking on the internet. Yeah, I mean, I also
- 39:40don't agree. I think that cultivated meat's a
- 39:44perfectly good solution. I say perfectly good
- 39:47solutions and it is in itself a reasonable solution,
- 39:50even if it might have issues that prevent it
- 39:53from being a good solution. I don't see... And
- 39:59trying to steal man the argument, right? Some
- 40:02cultivated meat companies still use fetal bovine
- 40:06serum. The industry is moving away from this.
- 40:09Nonetheless, even if it did require a fetal bovine
- 40:12serum, it's a tremendous improvement. It's much,
- 40:15much less animals would need to be harmed to
- 40:19generate FBS for cultivated meat. But again,
- 40:22the industry is already moving away from that.
- 40:25Some people might object to just taking cells
- 40:28from animals, period. I think that objection
- 40:32is pretty far -fetched in terms of that sort
- 40:39of procedure is incredibly non -invasive and
- 40:41not very painful at all. Is it like a check -swap
- 40:45effectively? I don't know what approach they're
- 40:48using exactly. either it be a very small biopsy,
- 40:52which is slightly more invasive, or a cheek swab,
- 40:54or a small skin swab. In any event, it's much
- 40:59less invasive than the current system of chopping
- 41:02off 100 billion chickens' heads. So I don't know,
- 41:06guys, which one do we do? And the scale there
- 41:10is much smaller, right? That's the whole idea,
- 41:12is you replicate the cells so it's not like you
- 41:14need a chicken every time you make this product.
- 41:17Well, they do present those arguments against
- 41:20lab -grown meat, but I think it has to do more
- 41:23with what Seth alluded to, which is, you know,
- 41:26this anti -capitalist ideology. Which I'm against
- 41:32in all of its guises for the record, not just
- 41:34this one, but please, Jacob. It's funny that
- 41:38I'm being asked about anti -capitalism and...
- 41:41Marxism capitalism again in the same week In
- 41:45the new year because it happened like at the
- 41:47New Year's party as that too and I was like,
- 41:48oh man I really never felt like I developed a
- 41:51level of sophistication to have like a nuanced
- 41:53view in this argument So I'll just make make
- 41:56an easy maybe an easy cop -out to say like Yeah,
- 42:02I just don't think these terms end up being defined
- 42:04very effectively or it's you know, very few people
- 42:07actually endorse like a absolute free market
- 42:11that's utterly unregulated and vice versa. I
- 42:16don't know the extent. I'm sure many people endorse
- 42:18like fully centrally planned economies, but in
- 42:21practice, like very few of either exist. So yeah,
- 42:28I definitely think like planned economies probably
- 42:31less good than free markets. But that said, there's
- 42:35like weird niches where Walmart just runs a massive
- 42:39internal planned economy and seems to do a pretty
- 42:44good job of it. But they're also effectively
- 42:46a dictatorship, so that's not ideal from a different
- 42:51perspective. But generally, it seems far -fetched
- 42:56to me. I'm not too worried about the fact that
- 42:58it's a capitalist solution to the problem. a
- 43:07general comment. Ryan, I think that a lot of
- 43:10this stuff boils down to expressive benefits.
- 43:13Most of politics can be explained in terms of
- 43:16expressive benefits, in terms of what does it
- 43:18feel good to say and be seen saying. This is
- 43:26very speculative, but like when I think about
- 43:28vegans, I know they're coming at it from like
- 43:31two camps. There's the people who have a strong
- 43:35moral intuitionist sense about animals' dignity
- 43:38and worth. And then there are those of us, and
- 43:42I'm putting myself in this category, and perhaps
- 43:44Jacob, although, you know, my heart also hurts
- 43:49for animals. We're like the abstract camp. We're
- 43:52like the, you know, 80 billion is 80 billion
- 43:55times, a chicken's dead is 80 billion times worse
- 43:58than a single chicken being hurt. So like, we
- 44:01have these sort of big picture considerations
- 44:05that require many layers of abstraction to understand.
- 44:11And this is as opposed to populist thinking,
- 44:15which in its leftist vegan form comes out as
- 44:18that's bad, that's gross, that's capitalist.
- 44:23No abstraction necessary. That's like just those
- 44:26are the bad guys. Don't do that. I think this
- 44:29kind of politics is really harmful when Trump
- 44:31does it, and I also think it's very harmful when
- 44:34vegans do it. But, like, I don't think that...
- 44:40I don't take it... I don't worry about it too
- 44:42much because I think, again, a lot of it is just,
- 44:44like, the benefits of being seen saying this
- 44:46on Twitter or whatever. Yeah, I guess I'm a little...
- 44:50That feels a little cynical to me insofar as
- 44:52I imagine... Again, not knowing the actual people
- 44:56saying this. So I imagine a lot of them are saying
- 44:58this on Twitter, but it's not really to like
- 44:59some grand audience. Although maybe that makes
- 45:02it all the more aggressively beneficial. Yeah.
- 45:05You ever been to vegan circle jerk, the subreddit?
- 45:09No, don't spend time there. Yeah, anyway, I just
- 45:14think it's like a situation of much like cage
- 45:17free. A lot of vegans have opposed cage free
- 45:20for a long time. incremental progress seems good.
- 45:25Winning things seems good. And not saying do
- 45:29that at any expense, compromise any value. But
- 45:33I think one needs to identify the clear and concrete
- 45:38harms or at least clear and reasonably likely
- 45:41harms beyond this is capitalist. Therefore, we
- 45:45should be an aesthetic objection. It's about
- 45:48like it's like qualities of unbeauty. I'm not
- 45:52sure like I don't sure that's like the position
- 45:54I don't know maybe that's the like expressive
- 45:56belief position but I think like the steelman
- 45:59position is something like, you know, this is
- 46:01enforcing a Like capital -driven hierarchy. Yeah,
- 46:05it's like dot. It's not well Yeah, no, there's
- 46:08a version I mean and personally I would like
- 46:10the part of this that I am on board with is that
- 46:12I think that blended meats There if blended meats
- 46:17which are like, let's say 30 % mushroom blend
- 46:21and 70 % beef. If those have the capacity to
- 46:26replace rather than compliment meat demand, I
- 46:29would like to see that demonstrated rather than
- 46:31assumed. I kind of think, I kind of, my picture
- 46:35of blended meats is I don't, I see them like
- 46:38rising in parallel to meat consumption as like
- 46:41a specialty product. Like might be wrong, but
- 46:46I'm not like very excited. I'm skeptical until
- 46:49we see evidence of displacement. Yeah, I guess,
- 46:53um, to close the conversation on politics and,
- 46:57um, you know, conservatism that we started earlier,
- 47:02um, I know of an initiative that will launch.
- 47:06I won't say more than that. Um, I think that
- 47:10efforts to get conservative on board for, you
- 47:16know, on the about veganism and plant -based
- 47:20living and all of that are going to take a new
- 47:26scale. So yeah, I think those investors are aware
- 47:32of the obstacles ahead and I think that's my
- 47:37theory that they want to start making more outreach
- 47:42efforts to conservative audiences. So we'll see
- 47:45more of that. Um, and, you know, about the, the
- 47:50more anti -capitalist crowd, um, I guess my position
- 47:55is there's lots of energy, you know, if we look
- 47:59at it from, um, amoral, just, you know, politically
- 48:03cynical way, um, there's a loss of energy, uh,
- 48:08with that crowd and they're in good position
- 48:12for appealing to people who also share their
- 48:15belief system. And so I don't want them to sabotage
- 48:21efforts from other vegans or animal advocates
- 48:27who have great initiatives. But at the same time,
- 48:31I feel like they should have a place under the
- 48:37animal advocacy movement umbrella. What do you
- 48:42make of that? I very pro big tent I think to
- 48:48the extent possible and you know, I as someone
- 48:52who's been publicly critical plant -based meats
- 48:54Some might say that I'm not practicing what I'm
- 48:57preaching But you know, I will I do disagree.
- 49:00I think like reasons discourse and disagreement
- 49:04is valuable and you should do it and I'm not
- 49:08suggesting that these will say anti -capitalist
- 49:11vegans who oppose cultured meat are not engaging
- 49:14in recent discourse. I don't know what discourse
- 49:16they're engaging in. You just told me this happened,
- 49:18so I don't really know too much about it. So
- 49:20I think having a reasonable discourse about that
- 49:22is fine. But I do think we should avoid open
- 49:28sabotage of one another's efforts. For example,
- 49:33in my work, I've taken great pains to... Not
- 49:37that it was really hard to do, but instead of...
- 49:39tried to be very careful to focus on the empirical
- 49:43and scientific issues surrounding plant -based
- 49:44meat, as opposed to the movement and advocacy
- 49:47issues of like, oh, you know, who says what,
- 49:50who are the actors and whatnot. I've tried to
- 49:52maintain my interests as primarily what is the
- 49:57human psychology and human behavioral science
- 50:00that underpins the interaction of plant -based
- 50:02meats with animal -based meats. Very much agree
- 50:06with that. I think what Jacob and I fundamentally
- 50:08share in common is a commitment to empiricism.
- 50:11The kinds of issues that we're interested in,
- 50:12if I might speak for both of us for a second,
- 50:15can be resolved with experiments. Like we believe
- 50:20that plant -based meats, if they achieve XYZ
- 50:23conditions, will move the market's Q amount.
- 50:27That's like a testable proposition. And if I'm
- 50:31wrong, I will say, I'm wrong. I was wrong. It
- 50:34was better or worse than I expected. I do not
- 50:37have much interest in the, I know this is very
- 50:41dismissive, more like aesthetic view of social
- 50:47change where it's like it's like thinking about
- 50:50this broader system stuff in a way that's like
- 50:52people find like inequality inherently unlikable
- 50:56so they want to get rid of it. I don't. And likewise
- 51:00recently there was a debate where I think it
- 51:02was like farm kind Said something criticizing
- 51:05for January and like on social media. I saw some
- 51:08people can like Seeming to get up in arms about
- 51:13this now There are of course two confounding
- 51:16factors here The first of which is that there's
- 51:18performative outrage and the second of which
- 51:20is that the algorithm rewards performative outrage
- 51:22So I have no idea what people actually think
- 51:24about this, but whatever I saw some of this on
- 51:26the internet and I was thinking like farm kind
- 51:30is doing There's like an empirical question here
- 51:36Which kind of messaging works better? I don't
- 51:39know and Because I don't know like it's interesting,
- 51:42but I don't like I don't care You know what?
- 51:45I mean? I don't mind like personally invested
- 51:47in the question of the January versus whatever
- 51:49farm kind was saying instead I would just say
- 51:50like there's an answer to this question. I don't
- 51:53know what it is I'm curious about that, but till
- 51:56I know I'm like, you know Gnostic Definitely,
- 52:00there's lots of pathos around those issues. And
- 52:06yeah, we should be more intellectually minded
- 52:08about them. Well, should we like I share, I have
- 52:12my values that I don't have any basis of saying
- 52:14it should be yours. Like this is, you know, you
- 52:16get it. Or anyone's. Well, you know, that's the
- 52:20reason why I started the science of activism
- 52:23initiative on this podcast, because I I felt
- 52:26strongly that there was uh, too much, you know,
- 52:30this is my personal opinion. Um, and I will,
- 52:34um, just scream about it online and, you know,
- 52:39people will react emotionally and it's just the
- 52:42never ending thing. And, um, when you have so
- 52:48many researchers, actually studying those issues
- 52:50and coming up with answers, not just questions
- 52:53and, um, emotional arguments. Um, it, you know,
- 52:58it feels good. It makes you smarter and it's
- 53:01clarifying. So yeah, I, I, I do tend to bring
- 53:08the discourse more on the intellectual side and,
- 53:12and just calm down people, like, um, bring down
- 53:17the temperature. in terms of all of those debates
- 53:21going on with people who have their hearts on
- 53:25their sleeves. And yeah, I guess those are my
- 53:29thoughts. I don't know, man. We always said that
- 53:32the meta question is not resolved. We don't really
- 53:34know whether empirical research is a superior
- 53:36method of changing the world rather than online
- 53:40shouting. That's also an empirical question.
- 53:44Well, you I want this podcast to be this big
- 53:51tent and representative of different opinions
- 53:54and views. At the very least, researchers and
- 53:59research and empirical studies should be included.
- 54:04And there are so many influencers out there.
- 54:06How many make the effort of giving voice and
- 54:12attention to that side of the world of activism,
- 54:17not that many minority of the minority. And so,
- 54:22uh, I, yeah. Well, I guess I'll ask you both
- 54:25a question. Pretend you could, you're like set
- 54:28your life is D and D in this little scenario
- 54:31and you can set your character stats to whatever
- 54:34you want. And then you can maximize into something
- 54:36else. If you, and you still have your, like your
- 54:40current goal of making a vegan world, what would
- 54:43you be doing if you could do anything, like if
- 54:45you were skilled at whatever you want to be skilled
- 54:46at. Like, I think I might be doing more delicious
- 54:54vegan food. I'm just not a very good cook. Like
- 54:58I'm doing what I do because I can't do other
- 55:00things. Do you have something else in mind that
- 55:03you would do if you could do anything? Yeah,
- 55:09I think maybe, Jacob, you feel the same way,
- 55:12but I'm already doing what I can and what I know.
- 55:17Yeah, but my question is, what if you could do
- 55:19anything? Well, you know me well enough, Seth,
- 55:24to know I'm just going to nitpick this question.
- 55:26I mean, anything? Besides just like using your
- 55:30force power to persuade people. Yeah, just get
- 55:33the objective. I mean, among human abilities
- 55:37that exist. Okay. Use for me to career change
- 55:42all together and I'm a little bit in the middle
- 55:44of a career change So this is a little bit of
- 55:46a live question to some extent I'm sure you would
- 55:48do think about this because you you're smart
- 55:50Jacob You can you can be something else if you
- 55:52want to be so yeah Yeah, I'm a little bit open
- 55:57-minded about it. I've looked into going to donor
- 55:59relations donor development Just to help move
- 56:02more money into amlavixi and amlavixi research
- 56:05perhaps ideally I've also thought about working
- 56:10as a campaigner. So just upping my skills as
- 56:14a campaigner to actually run the campaigns that
- 56:17a lot of animal advocacy work relies on. What
- 56:20do you have in mind? Yeah, KTree is a classic
- 56:27campaigning issue, but, you know, people have
- 56:29campaigns against cosmetics companies for animal
- 56:32testing, campaign against SeaWorld famously for
- 56:37holding captive animals. I'm not sure those are
- 56:40the campaigns I personally would pick to work
- 56:41on, but all campaigning skill type things. How
- 56:47about you, Ryan? Well, I kind of do some things
- 56:56around that with my current work in fundraising
- 57:00with Dharma Voices for Animals of the organization,
- 57:06but I think I would uh, adopt like a position
- 57:11of a religious leader of some kind. Um, and then
- 57:15preach the good word about animals. It's a great
- 57:18answer. Uh, I, I do think there are so many parallels
- 57:22between, uh, major religions out there and what
- 57:26we're trying to do. Um, starting with, you know,
- 57:30diets, uh, constraining diets. We were talking
- 57:32about, um, convenience of a vegan diet. Well,
- 57:41did religious groups throughout history tackle
- 57:44that question? You know, Jewish sects or Muslim
- 57:50immigration waves coming to countries where they
- 57:54don't have the same dietary values and they safeguarded
- 58:02their values and propagated them successfully.
- 58:06Um, what happened there? Um, also what we're
- 58:09asking from people is very much akin to converting
- 58:13to another religion. Um, you know, it's a, it's
- 58:17a big asks to become vegan and completely transformed
- 58:21the way you interact with, uh, different species
- 58:25and change your, uh, dietary habits and also
- 58:29your, you know, cosmetics and all of that. Um,
- 58:34Religious conversion is the closest thing that
- 58:36comes to money in terms of comparisons. I'm totally
- 58:41down with this. Like, you're going to start a
- 58:43cult and then instead of the cyanide pill you
- 58:45get like some tofu. It's like, this is the big
- 58:48reveal. Exactly. I mean, where is the vegan cult?
- 58:53Where is the vegan cult? That's a great question.
- 58:54We need more. I think they were called the ZZX
- 58:57and they weren't so good. Yeah, and Zazian's
- 59:01and DXE has also been accused of being a cult.
- 59:04So I know there are some options. DXE is direct
- 59:09action everywhere. They do good work. Well, you
- 59:15know, it needs to have the potential to scale
- 59:19up. So the leader needs to die and become a martyr
- 59:23of some kind. Yeah, I mean, the most successful
- 59:25cult in history is indeed very successful. Like,
- 59:30how pretty Christians are there. Exactly. I mean,
- 59:33so if we could start our own cult or, you know,
- 59:40inject our values through a cult, through, you
- 59:44know, a mainstream religion and see it propagate
- 59:47through those already existing structures, I
- 59:53think that's That's the closest thing to checkmate.
- 59:58The less cynical version of this just adds the
- 1:00:01URE onto that and talks about culture. We talk
- 1:00:04about identity and something that I personally
- 1:00:07think we could do more of is that like, let's
- 1:00:09pretend that you were running this is hypothetical,
- 1:00:11of course, pretend you were running or in the
- 1:00:13middle of designing a default study at a prominent
- 1:00:16Catholic university in the Boston area. Then
- 1:00:21one thing you might do is you might have a religious
- 1:00:24studies professor come and talk about why Catholicism
- 1:00:28and plant -based diets are consistent with each
- 1:00:30other. We can like layer these things on top
- 1:00:33of each other. And that's an issue where credibility
- 1:00:36matters. Like no one's going to care what I think
- 1:00:39about Jews and veganism. It's just, but like
- 1:00:44someone else can talk credibly about this. And
- 1:00:46that's good. I think it's very promising. I think
- 1:00:49I might know the name of that professor. I just
- 1:00:53have a hunch too. Can we get some initial stuff?
- 1:00:56Yeah, AG, not SG, but like, you know, there's,
- 1:01:00you know, there's like the center. Okay. So at
- 1:01:02first I haven't talked to Aaron about this, but
- 1:01:04Aaron Gross, professor at UCSD is a religious
- 1:01:06studies professor who also does food stuff. And
- 1:01:09there's a way to make this case that we can bet
- 1:01:13like the values you already hold are consistent
- 1:01:18with eating more. plant -based diets. I mentioned
- 1:01:21earlier, Matthew Scully, who's doing the same
- 1:01:23thing for Christianity. I think this is really
- 1:01:26hard. And I think it's like a long patient conversation.
- 1:01:31But I think this is a problem. I think that in
- 1:01:33conjunction with like convenience, this is a
- 1:01:35good one. I think it's also a hard sell for the
- 1:01:40advocates out there, since you know, I think
- 1:01:4473 % of animal advocates uh, consider themselves
- 1:01:50as non -religious or atheist. And so, you know,
- 1:01:55do they believe in making a difference with religious
- 1:01:59communities? Maybe they're more skeptical about
- 1:02:03that. Well, I think how much of that, go ahead.
- 1:02:08I just came back to the same vein of like, um,
- 1:02:12anti -capitalism. I imagine some of these 70
- 1:02:14% of atheists of vegan atheists are, uh, anti
- 1:02:19So they might have some principled view against
- 1:02:22it as well. But those of us who are utilitarian,
- 1:02:26who are total empiricists have to say like, well,
- 1:02:30if it gets the job done, that's our ideology.
- 1:02:34Whatever gets the job done. The good news is
- 1:02:36you already have a microphone, Ryan. I mean,
- 1:02:41yeah, that's, that's the beginning. You're so
- 1:02:43mild mannered. You're very pleasant to talk to.
- 1:02:46Let's get more spirit. Yeah. I mean, it's, it's,
- 1:02:53we were talking about hypotheticals. I'm not
- 1:02:57planning on becoming a cult leader anytime soon.
- 1:03:00We were ready to join. Well, I'm already, you
- 1:03:05know, on the Buddhist wagon. It's not as energetic
- 1:03:12as evangelists or Islamists or something like
- 1:03:17that. So yeah, maybe that explains my mouth manners.
- 1:03:22But you know, you see the Dalai Lama had so much
- 1:03:27and still has so much momentum and popularity.
- 1:03:34And he's not, he's not, he doesn't even speak
- 1:03:38English yet. You know, his presence has triggered
- 1:03:45that much support around the world, which is
- 1:03:48incredible. Well, if we could gain him to our
- 1:03:51cause that, that would do lots of good. I mean,
- 1:03:57I imagine he's pretty close. Like if he's not
- 1:04:01vegan, he probably eats very few products from
- 1:04:03factory farms in Dharamshala. That's also what
- 1:04:10I think. I do believe that he's a prime target
- 1:04:14for that. I believe that Buddhism as a whole
- 1:04:18is a prime target for that. But you know, I'm
- 1:04:22blowing my own horn. Yeah, but on the other hand,
- 1:04:25I remember once upon a time I was in a political
- 1:04:27science program and graduated for graduate school
- 1:04:29did not pan out, but I tried. And there was a
- 1:04:32professor who brought this mystery forward that
- 1:04:34was like two Buddhist countries. I think it was,
- 1:04:37um, or it was like leadership in Myanmar is crushing
- 1:04:42the Rohingya. This is not consistent with Buddhist
- 1:04:45values. Why? And it was like, cause they're people.
- 1:04:50You know what I mean? Like they're, I don't know.
- 1:04:52I lived in Thailand for a year. They ate a lot
- 1:04:54of pork. I don't think they treat their cat,
- 1:04:56their pigs very well. It's like people like me.
- 1:05:01Yeah. I mean, if you look at early Buddhist history,
- 1:05:05people sent assassins to the Buddha. There were,
- 1:05:09you know, warring factions in the monastic order
- 1:05:14he founded, and that was during his lifetime.
- 1:05:18So yeah, people are people. That's just how things
- 1:05:22are. That's interesting. I didn't know that.
- 1:05:24Yes, even his cousin, you know, tried to assassinate
- 1:05:30him and take over as the Buddha. Anyway, there's
- 1:05:34a whole history that we can point to. Jacob,
- 1:05:38did you want to add something before we change
- 1:05:41topic? No, I'm happy to shift that my religious
- 1:05:45knowledge is quite limited. You're part of the
- 1:05:4873%. Yeah, guilty as charged. Okay, so one last
- 1:05:57highlight of 2025 that I really wanted to discuss.
- 1:06:03Something that was also dominating conversations
- 1:06:06in my podcast in 2025, the concept of vegan by
- 1:06:15default and nudges. Lots of You know, I even
- 1:06:20attended the last AVA summit. That was a big
- 1:06:24thing. You know, I kept reading nudges, wondering
- 1:06:28what was that about? So what do you make of that
- 1:06:33new trend and new tactic of advancing the cause?
- 1:06:39Do you believe in it? What does data say about
- 1:06:43it? Oh, Seth, you have to take this one. You
- 1:06:45just wrote the post. I did, right. I recently
- 1:06:47wrote something called, I wrote a newsletter
- 1:06:51called regression to the meat. It's a pun on
- 1:06:54the phrase regression to the mean, which is the
- 1:06:56tendency for subsequent samples pulled after
- 1:07:00extreme initial samples to be more likely to
- 1:07:03the true underlying average tendency. And the
- 1:07:06regression to the meat pun is my, is the, is
- 1:07:09actually, I was thinking about defaults and not
- 1:07:11just when I thought of it, which is we have a
- 1:07:14lot of default studies and nudge studies. So
- 1:07:17the archetypal default study, and this is a situation
- 1:07:20where I really think defaults work, goes like
- 1:07:22this. You're at a university, you sign up for
- 1:07:25a conference, and then they give you a free meal
- 1:07:29for part of the conference. And the free meal
- 1:07:31says, tofu with string beans. And then there's
- 1:07:36a little thing that's like, please email us if
- 1:07:39you would like a meat option instead. So the
- 1:07:41basic idea is that you have to go out of your
- 1:07:45way to get meat. I believe this probably reduces
- 1:07:50meat consumption at the conference by about what
- 1:07:54the study's expected to, which is sometimes up
- 1:07:56to 73%. Most people just don't feel like it in
- 1:07:59terms of going through the effort to write the
- 1:08:01email. What we don't know is what they then eat
- 1:08:06at the next meal. Is there any probability that
- 1:08:09being defaulted into plant -based food at time
- 1:08:12one makes you perhaps more likely to have meat
- 1:08:16at time too? I don't know the answer to that
- 1:08:18question. I think at conferences, plant -based
- 1:08:21defaults probably work. I think in New York City
- 1:08:23hospitals, plant -based defaults probably are
- 1:08:26good. Because that's a time when you've targeted,
- 1:08:30somebody's had a health scare, they're in the
- 1:08:32hospital overnight, and then basically how the
- 1:08:34plant -based default system works. And the New
- 1:08:35York City hospital is the person, let's say you're
- 1:08:38bedridden, the nurse comes to your desk, your
- 1:08:41bed, and they're like, We have this like nice
- 1:08:46mushroom ragu, which by the way, not plant based,
- 1:08:50that was a terrible example. Mushrooms are not
- 1:08:51plants. But whatever. And then you say, no, I'd
- 1:08:57like something else. And they say, we have broccoli
- 1:09:00and General Tso's tofu. And then you have to
- 1:09:03say no like three times before they actually
- 1:09:06serve you meat, before they offer a meat thing.
- 1:09:09This I think is a pretty good time to be giving
- 1:09:14people plant -based defaults, because, you know,
- 1:09:16they're already, like, in the hospital. They
- 1:09:19probably are thinking about, like, mmm, am I
- 1:09:22living in the healthiest way? Like, this is a
- 1:09:24good time to reach people with, like, a new way
- 1:09:26of trying things. And it's very possible some
- 1:09:28of them are like, hmm, we're good. I'll try.
- 1:09:31I'll make this at home. But what about, like,
- 1:09:35restaurants and universities? There, I would
- 1:09:39say I'm a skeptic. Because first, the big reason
- 1:09:43is that like a default just don't work as well
- 1:09:46when you're thinking about like, imagine you're
- 1:09:48going into Chipotle and they say like, our burrito
- 1:09:53comes with tofu. Please let us know if you want
- 1:09:56one of the following meat options. Like in some
- 1:09:58ways that's a default, but just saying, I'd like
- 1:10:02some chicken, please, is incredibly easy. Like
- 1:10:05most of these default settings, like when you
- 1:10:07actually think about scaling up, you should think
- 1:10:08about. times when it becomes really, really easy
- 1:10:12to get the meat. You don't have to say it like
- 1:10:14four times. Because if you were at a restaurant
- 1:10:17and they made you have four interactions every
- 1:10:20time you wanted to order meat, you're going to
- 1:10:22go to another restaurant. Any restaurant that
- 1:10:25knows this is not going to make it very hard
- 1:10:28to get meat. So as things like scale up, I think
- 1:10:30we should expect the defaults to just become
- 1:10:32less default D and more just like a selection
- 1:10:36of choices with one preselected. And second,
- 1:10:44well, this is really the same point. People like
- 1:10:48meat, and as they're going about their normal
- 1:10:52lives, they're going to make the effort to get
- 1:10:54it. Like this is just, everybody who knows someone
- 1:10:59who eats meat is going to say like, oh yeah,
- 1:11:01that person eats meat because they like it, not
- 1:11:03because they're just like going through life,
- 1:11:05just going like with their eyes closed, just
- 1:11:06eating whatever is served to them. I just don't
- 1:11:09think that's a good model for how people eat.
- 1:11:11So I don't think this whole like nudge mindset
- 1:11:15of trying to get people to choose plant -based
- 1:11:18and then like unconscious automatic way. I don't
- 1:11:21think that's a very good model for how people
- 1:11:23choose food. Yeah, that's what I want you to
- 1:11:26go first. No, I agree with so much of what you
- 1:11:29said and I think it's a really good answer and
- 1:11:31gets to like one of my most fundamental defaults
- 1:11:36views, which is like defaults are not a very
- 1:11:40coherent class of things. It's very idiosyncratic.
- 1:11:45The conference ones, yeah, they seem really trenchant.
- 1:11:49And like a key aspect of all nudges, well, not
- 1:11:52all nudges, but conscious nudges, like a default
- 1:11:58that your brain has to do some processing to
- 1:12:02interact with and for the effect to happen to
- 1:12:05some extent. is that salience is really important.
- 1:12:10So to the extent that in a conference setting
- 1:12:13where you're like given these two options and
- 1:12:15it's like right there in front of you and you're
- 1:12:17like, oh, either I need to just go with this
- 1:12:19default or I need to choose to send this email.
- 1:12:23It's a very salient situation. And exactly as
- 1:12:27you outlined Seth, like in Chipotle or even worse
- 1:12:30in the grocery store, it's like. What are we
- 1:12:35doing here? How do we make some? Yeah, I don't
- 1:12:37know exactly I have no idea what that like maybe
- 1:12:39you like put them like a head or something and
- 1:12:42you put like the meat behind it but like even
- 1:12:44then Well, I don't get it and I guess the good
- 1:12:47news is we can exploit the diversity of the term
- 1:12:50default here because people You know do people
- 1:12:52do serving size defaults so you can call that
- 1:12:54a default and that's more amenable to like a
- 1:12:56grocery store context Or you could make the serving
- 1:12:59size smaller by default I'm very amenable to
- 1:13:02this kind of default. I think this is a good
- 1:13:03one. Like just make the spoons smaller or whatever.
- 1:13:06Yeah, yeah. Now, but that's like, I will say
- 1:13:10in the grocery store, that's like kind of unpopular.
- 1:13:12People don't actually like it when you make the
- 1:13:13food that you sell smaller, by and large, like
- 1:13:16shrink -flation. Yeah, they call it, exactly,
- 1:13:18shrink -flation, right? Like people will notice
- 1:13:21when you take the meat away. Yeah. Well, if it's
- 1:13:23salient, you could just do it a little bit. But
- 1:13:26yeah. It's like, oh, it's just, it's not sick.
- 1:13:31It's 15 .85 ounces of prime riff. I haven't been
- 1:13:38doing for ice cream for years. Um, so yeah, I
- 1:13:43think it's really hard to talk about defaults
- 1:13:45and nudges, good ideas, bad ideas, because it's
- 1:13:47just incredibly nuanced. And that's the other
- 1:13:50big issue with this idea is it's I'm remiss to
- 1:13:55even call it an idea because it's a dozen ideas.
- 1:13:58And then we get to the measurement and evaluation
- 1:14:00problem, which turns out to be a big problem.
- 1:14:03It's really hard to run the high quality studies
- 1:14:06that are needed to validate that these things
- 1:14:08work. It's very expensive. There's not a lot
- 1:14:12of people willing to do it. It's not very fun
- 1:14:13to do by and large, I'll say by most accounts.
- 1:14:19I can speak to this personally, getting a university
- 1:14:22to randomly assign meals is hard. They've got
- 1:14:26their own thing. They've like got their, like
- 1:14:28for them, like just keeping the ship afloat is
- 1:14:32hard. Like it's a lot of work to run a university
- 1:14:34dining hall. And then like, you know, I, some
- 1:14:37researcher from some school come in and be like,
- 1:14:39well, would you consider allowing us to evaluate
- 1:14:41the following? And then it's like, some schools
- 1:14:44say yes, but it's a lot of work. And randomizing
- 1:14:49is one thing, that's hard. But just counting
- 1:14:52is a huge barrier. We worked on this at Stanford
- 1:14:55where we're doing some counting of animal product
- 1:15:00usage and it's incredibly difficult. It's just
- 1:15:03a very hard thing to do. Oftentimes, generally
- 1:15:08speaking, you end up with packages where all
- 1:15:14the record the university have is this, and it's
- 1:15:18seafood medley frozen. And there's not even going
- 1:15:20to be vowels in there. It's going to be like
- 1:15:22S -A or S -F -D -F -R -Z -N medley. Yeah, this
- 1:15:28lack of data granularity continues to befuddle
- 1:15:31me. As in you're not sure what's going on with
- 1:15:35that, like why it is that way? Yeah, like you
- 1:15:38think that someone at the dining hall has a sense
- 1:15:41of like, oh yeah, it's like seafood frozen medley.
- 1:15:43That's like, OK, we could we could you could
- 1:15:46figure out theoretically how much pounds of seafood
- 1:15:50that translated to when you talk about units
- 1:15:52of frozen medley someone could it's just that
- 1:15:56the information is really dispersed in a huge
- 1:15:58bureaucracy Oh sure and like the other factor
- 1:16:01with a product like that for example it's sort
- 1:16:04of the chef special approach potentially where
- 1:16:07a manufacturer has a bunch of seafood to get
- 1:16:08rid of and they're like great here's the seafood
- 1:16:10medley it can change contents over time so it's
- 1:16:13really quite pernicious like that so figuring
- 1:16:15out if these things work is really hard. And
- 1:16:19calling back to when Seth was on the podcast
- 1:16:21previously, you gave a really good example with
- 1:16:24the social norms literature, which is also in
- 1:16:26the nudge domain. And there's just a huge amount
- 1:16:30of heterogeneity within the social norms nudge
- 1:16:33literature, where you get, oh, yeah, it worked
- 1:16:36at this dining hall, didn't work at this dining
- 1:16:37hall, and there's no obvious explanation of why
- 1:16:40it worked in some place and why it didn't work
- 1:16:42in others. which makes it makes it a really hard
- 1:16:46path to go down because then you really need
- 1:16:48evaluation at each and every stage and maybe
- 1:16:52contrast that with say like a very effective
- 1:16:55drug where you run a clinical trial the drug
- 1:16:59seems to work great doesn't seem to be associated
- 1:17:01with any serious adverse side effects or anything
- 1:17:04you put it out into the world sometimes you do
- 1:17:06some post you know post -tracking of distribution,
- 1:17:12make sure there aren't any, now that you're testing
- 1:17:13in more and more people, there weren't any serious
- 1:17:15outcomes. And then you're kind of good to go.
- 1:17:18You just, you're a doctor and you say, hey, you've
- 1:17:20got this, here's the drug, it works really well,
- 1:17:22problem solved. And you only need to run one
- 1:17:24very expensive clinical trial at the start. If
- 1:17:28it's possible that for nudges, it's like, oh
- 1:17:30yeah, we figured out one for this one place and
- 1:17:33it works. We spent a lot of time and money designing
- 1:17:36it and testing it and showing that it works.
- 1:17:39And we have no idea if it'll work elsewhere on
- 1:17:41the next one the sociologist Peter Rossi makes
- 1:17:44exactly this point in his like essay about which
- 1:17:46comes up with this like iron law of social evaluate
- 1:17:49of evaluation is that like any program as it's
- 1:17:53scaled up the expected impact is zero if He Jacob
- 1:17:59specifically says like there's no social program.
- 1:18:01That's as effective as a small pack spike small
- 1:18:03pox vaccine. Yeah, and there never will be like
- 1:18:07context really I mean I just read a nice paper
- 1:18:11in by Hunt Alcott called Site Selection Bias
- 1:18:13in Empirical Economics. And its basic point is
- 1:18:16that the researchers choose, have choices about
- 1:18:23where they make things happen. And they like
- 1:18:27the very qualities that make like, okay, let's
- 1:18:30say that we had a good experiment at Stanford
- 1:18:32because we have a willing partner there. correlated
- 1:18:35with propensity for eating plant -based foods
- 1:18:38in a way that does not necessarily scale to other
- 1:18:41locations. Yeah, it's very pervasive and it's
- 1:18:47a real problem for running that sort of work.
- 1:18:50I think it's still a good idea, I think it's
- 1:18:51still worth pursuing. It's interesting, recently
- 1:18:54I was actually considering founding a charity
- 1:18:58to work on getting grocery stores corporations,
- 1:19:04food service businesses to make pledges to reduce
- 1:19:08their meat usage. And as I was thinking about
- 1:19:10what that could look like early on, my conclusion
- 1:19:13was basically the first couple of years of this
- 1:19:16charity might involve no pledging whatsoever.
- 1:19:18It's just counting. All we do is count. And the
- 1:19:23counting issues are really, really difficult.
- 1:19:25It's perhaps quite underrated, in my opinion.
- 1:19:29maybe among like the, I'll say the vegan public,
- 1:19:31if you will. I was just like, how do we figure
- 1:19:34out what helping looks like? It's quite hard.
- 1:19:41I might add to that. A area of personal interest
- 1:19:46is universities that have made net zero commitments.
- 1:19:50So they say like, no, no carbon, no net zero
- 1:19:53carbon emissions by XY year. don't include food
- 1:19:57in that category and like on paper many of them
- 1:20:02are also like signatories to some like food reporting
- 1:20:06guidelines where they like should be tracking
- 1:20:08that stuff but like just getting a simple answer
- 1:20:11of like how much beef went into the dining hall
- 1:20:17last year so far we've been getting like we don't
- 1:20:21know Yeah, like schools, maybe some schools know
- 1:20:25but they certainly won't share it with others
- 1:20:27and I think that a lot of them like Jacob said
- 1:20:31like they have Metrics that are like a spreadsheet
- 1:20:37where it's like BF next row 279 and that's it.
- 1:20:44Yeah, and fundamentally the needs of Animals
- 1:20:48are not centered in the way university dining
- 1:20:52halls or any dining halls or any food business
- 1:20:54works by and large which is not surprising right
- 1:20:57when you sell frozen seafood medley you mostly
- 1:21:00keep track of how many boxes of frozen seafood
- 1:21:02medley you sell and when you buy frozen seafood
- 1:21:04medley and beef and a thousand other things from
- 1:21:06a dozen different suppliers mostly keep track
- 1:21:09of how much you paid and how many shipments you
- 1:21:10got more so than you know what species of fish
- 1:21:14was it how much did each individual fish weigh
- 1:21:16in the box and that's the sort of information
- 1:21:19that you need to understand the impact of those
- 1:21:21sorts of policies on animals. Once we get to
- 1:21:24fishes and crustaceans, this whole thing falls
- 1:21:26apart. It's so much easier to talk about, like,
- 1:21:29you know, like, land animals have some real virtues.
- 1:21:34Yeah, in terms of trackability. I don't know
- 1:21:38how much time we have, but there's two areas
- 1:21:40that I'd really like to talk to, like questions
- 1:21:43I want to ask you, Seth, if we if we can hard
- 1:21:45pivot. But I also, Ryan, if we have other things
- 1:21:48to cover. Well, not too much time. Okay, guys,
- 1:21:54cut it later. I want to hear what Jacob, I love
- 1:21:56being asked questions. We can always just call
- 1:21:59each other too. That's an option. Yes, that would
- 1:22:02be great. I will add it to the conversation because
- 1:22:05I, sorry, I have to leave now. No problem at
- 1:22:08all. Okay, great. Let me end by saying this was
- 1:22:13such an interesting conversation and I learned
- 1:22:15so much from you. Again very clarifying so many
- 1:22:21insights and Yeah, thank you so much for your
- 1:22:24work. Thank you for all of the research you're
- 1:22:27doing I hope I'm doing it. I will do and I'm
- 1:22:32doing a good job at you know bridging the divide
- 1:22:36between the researchers and the data and the
- 1:22:40activists because the activists are not up to
- 1:22:43date with whatever is studied, they're not even
- 1:22:47engaging with the studies, not as much as I wish
- 1:22:52they did. And so I am happy and honored to play
- 1:22:56a role in facilitating that conversation. So
- 1:23:02Seth, Jacob, thank you so much. This has been
- 1:23:05a pleasure. Yes, and Ryan, that's a beautiful
- 1:23:07mission statement. Thank you. Thank you. Thank
- 1:23:11you so much, Ryan. Yeah, hope we didn't alienate
- 1:23:13them too much with our crankiness. Not at all.
- 1:23:18I think they're used to some challenging conversations.
- 1:23:24I mean, I've had a few challenging ones before,
- 1:23:27so it's okay. But I always want to be called
- 1:23:31challenging. Love it. Thank you. Take care. Thank
- 1:23:36you. Bye bye. See you around. Okay, so first,
- 1:23:40when we're talking PTC, I have this hunch that
- 1:23:45for taste tests that are designed to identify
- 1:23:51plant -based meats that would reduce the most
- 1:23:53meat consumption if they're introduced and made
- 1:23:56available, that we should be testing in unblinded
- 1:24:02taste tests, so informed taste tests. with the
- 1:24:07idea that we would like actively shift beliefs,
- 1:24:10because like in a blind taste as you could theoretically
- 1:24:12afterwards say, by the way, that was plant based.
- 1:24:15Yeah, you could. Okay. But no, I want I want
- 1:24:19the taste as as they'd experienced it with the
- 1:24:23taste with the mind and the bias and their beliefs.
- 1:24:26And I have a couple different arguments for why
- 1:24:28I think this is and that the macro level view
- 1:24:30is just like, Yep, this is the way the world
- 1:24:32is and like people don't eat burgers blinded.
- 1:24:35So why would we measure them blinded? For the
- 1:24:40purposes I'm outlining. I do think there's other
- 1:24:42things for blind taste tests that are really
- 1:24:43useful. What else? Why on that do you want to
- 1:24:46see this taste test blinded? Number one is just
- 1:24:52naturalisticness. This is like the way people
- 1:24:54eat burgers, and to the extent that that's the
- 1:24:57case, we should test their burger eating in that
- 1:25:00way. is that blinding in for plant based meats
- 1:25:08is probably like fails adversely. So you don't
- 1:25:13actually get the like a pure estimate that you
- 1:25:15want like the blinded estimate. Because in practice,
- 1:25:19people will successfully guess like which ones
- 1:25:22the plant based meat and animal based meat. Or
- 1:25:25now we're not talking about culture meat in this
- 1:25:26context, we're talking about like the next generation
- 1:25:29of beyond meat products, right? Sure, sure. Yeah,
- 1:25:32we could. we'd go with that. Yeah, I haven't
- 1:25:36thought about this as much for cold -fade meat.
- 1:25:38And it's kind of interesting, and I haven't worked
- 1:25:40out all the implications of this, but if you
- 1:25:42have a really bad plant -based meat, then you
- 1:25:45get more de -blinding, obviously, because people
- 1:25:47like this as a piece of rubber, not a piece of
- 1:25:48meat. But if you have a quite good one, then
- 1:25:51you get less de -blinding, and then you get closer
- 1:25:56to your blinded estimates. just think about it
- 1:26:03now for the first time perhaps ever. I actually
- 1:26:08think that there are many situations in which
- 1:26:10people do blind test food and that's like not
- 1:26:13at all crazy like it to me like because you might
- 1:26:15go to I mean I can't think of any but like I
- 1:26:19just feel like You're just a nerd. Your friends
- 1:26:22do this. Yeah, I did. Oh my God. I did have a
- 1:26:25friend from high school who went to college and
- 1:26:28he conducted a double blind randomized control
- 1:26:30trial blowing weed smoke into a bottle of vodka
- 1:26:34to see if that would produce additional effects
- 1:26:36beyond just getting drunk. And it did. But he's
- 1:26:41a big nerd. I'm not I would never do how did
- 1:26:43he find that then the weed would be so obvious
- 1:26:46I think it's not enough that not I think if it's
- 1:26:48pretty like it's weed smoke I think it's like
- 1:26:49maybe you the other person might have actually
- 1:26:51worn a blindfold But I think you taste the weed
- 1:26:54smoke and smell it In the bottle, I actually
- 1:26:58wonder about that. I should ask him. It was 16
- 1:27:00years ago so We There might be situations in
- 1:27:08which you go to a restaurant and they're like,
- 1:27:11hey, do you want to try this new thing? And then
- 1:27:13they're like, by the way, it was frog or whatever,
- 1:27:16they'd say. But okay, putting that aside, I like
- 1:27:22this. I don't have a strong opinion about this
- 1:27:25question. I think we could test this itself because
- 1:27:27ultimately what we're trying to do, you're trying
- 1:27:30to go with naturalism. There's two different
- 1:27:34estimates we could be talking about. the first
- 1:27:36is your opinions towards plant -based meat and
- 1:27:40the second is this other thing which is in which
- 1:27:43the taste test is the thing we're trying to test
- 1:27:45it's like the updating not like would i go and
- 1:27:48try this at a restaurant but like sure sure like
- 1:27:51this is oh wow it's plant -based cool in which
- 1:27:53case like i'm not sure whether being blinded
- 1:27:55or not matters i think you're going for the first
- 1:27:58which is like yeah going to a restaurant and
- 1:28:00like trying it like oh whatever let's see how
- 1:28:01we go let's see how it is Yeah, so it's an estimate
- 1:28:05for general consumer acceptance. Yeah. Yeah.
- 1:28:07I generally like arguments in favor of keeping
- 1:28:11the design simple and saying like, this is what
- 1:28:14people are actually going to do. We should keep
- 1:28:15it that way. So then, but then we have the downside
- 1:28:18with that. We have like people's pre we're like
- 1:28:20more likely to activate those prejudice and not
- 1:28:23be able to go around them in the first place.
- 1:28:26If we let people bring all that to the table.
- 1:28:30Yeah, potentially although and it really it is
- 1:28:33interesting because it gets into like some really
- 1:28:35nuanced design questions Like if we go with like
- 1:28:40a sort of lady -tasting tea design and it's like
- 1:28:42yeah, you're gonna taste like You know three
- 1:28:44products one of them. I'm not sure I'm sorry.
- 1:28:48It's But I'll just explain the design itself.
- 1:28:51It's not But basically, if you're going to give
- 1:28:56them, say, a sequence of products, some of which
- 1:28:58are plant -based, some of which are animal -based,
- 1:29:00and say, hey, can you classify which are plant
- 1:29:01-based and animal -based? Then if they get flagged
- 1:29:09off, if they have a hunch that one of them is
- 1:29:13plant -based, then you're going to get a lot
- 1:29:15of those biases in. Even if it's blinded, they're
- 1:29:18going to project their expectations, potentially.
- 1:29:21And, but the other interesting caveat here is
- 1:29:23like, do you inform them ahead of time? Do you
- 1:29:25just have them walk into your room? Be like,
- 1:29:26here is an item to eat. Here's a piece of food.
- 1:29:29Eat it. What do you think? Well, it's more like
- 1:29:33you walk into a Chipotle like setting and you'd
- 1:29:35say like, Hey, we're trying this new thing. We're
- 1:29:37like, you know, and by the way, it's available
- 1:29:39for purchase. Like we can imagine that it's like
- 1:29:42a free sample at the door. Sure. Sure. Yeah.
- 1:29:46It's reasonable. I don't know. It's interesting.
- 1:29:49I think we both agree that taste is more than
- 1:29:51just like a fat. It's like taste is in the mind
- 1:29:55as much as it's on the tongue. I'm not sure.
- 1:29:58And I think that's like a very underappreciated
- 1:30:00fact in general is that there's so much psychology
- 1:30:02of taste. And that's why these sorts of designs
- 1:30:05appeal to me in general, because since the psychology
- 1:30:08of taste obviously influences consumer acceptance,
- 1:30:11having a design that captures the psychology
- 1:30:14in addition to the purely But and I guess my
- 1:30:19only pushback is it is totally plausible that
- 1:30:21Chipotle might or whoever Taco Bell might might
- 1:30:25say like hey, we have a new product We're not
- 1:30:27gonna tell you what it is. Do you want to try
- 1:30:28it? And yeah Burger King has actually done that
- 1:30:32for the impossible burger Like they did a little
- 1:30:34campaign like this a while back. Oh, is it like
- 1:30:36try our burger or whatever? Well, that's yeah
- 1:30:38by what? Yes Yeah, that's what so you should
- 1:30:40see but I do think that's gonna be the vast minority
- 1:30:42of trials in all cases You know, so it's certainly
- 1:30:47like You know outside of it and and like that
- 1:30:51whole deal has a little bit of a Hawthorne effect
- 1:30:53too Or it's just which is good good time to our
- 1:30:55next next question Where it's like, oh, hey Sorry,
- 1:31:00you're about to divide it. Oh sure, you know
- 1:31:03the fact that you're an experiment and the the
- 1:31:06Observation of that contrivance then influencing
- 1:31:08your opinions so you can imagine if you Were
- 1:31:11just went to the store window and they handed
- 1:31:13you a burger and it they'd actually secretly
- 1:31:16coin tossed whether you'd get impossible or beef
- 1:31:17burger it'd be quite a bit less skeptical. You're
- 1:31:20just like, oh, this is, of course, the burger
- 1:31:22that I ordered. Whereas if someone, go ahead.
- 1:31:26A lot of the best experiments, like the one you
- 1:31:27just proposed, are impossible because we don't
- 1:31:30want to kill people due to allergens, which sucks.
- 1:31:32Oh yeah, totally. If only we could take a few
- 1:31:37losses. For science. For the cause. Whereas at
- 1:31:43the door, when some guy's like, hey, you want
- 1:31:45to try our new product? Here's a sample. That
- 1:31:48triggers people's awarenesses. It's a sort of
- 1:31:51demand effect in a sense. It's like, oh, well,
- 1:31:54I'm trying a new product. I'm going to really
- 1:31:56think about the flavor in my mouth and that sort
- 1:31:58of thing. A funny anecdote on it one time I left
- 1:32:05some slices of like vegan Tofurky meat like sliced
- 1:32:08meat in their refrigerator And I'd made pancakes
- 1:32:11that's that morning at my parents house And my
- 1:32:13dad just came home and it was a round brown thing
- 1:32:15in a container and figured it must be a vegan
- 1:32:17pancake So put maple syrup on it and ate it with
- 1:32:19a fork and knife. I was like, huh? those are
- 1:32:22some kind of chewy vegan pancakes, but whatever
- 1:32:24they kind of had a Weird mouth feel like it's
- 1:32:28just a totally different product It was pretty
- 1:32:32far away and I swear my vegan pancakes aren't
- 1:32:34that bad. Do your parents think you're like their
- 1:32:36dipshit vegan so they're like, oh, he's making
- 1:32:38whatever bullshit he's making. My parents are
- 1:32:41done with whatever it is on again. No, no, they
- 1:32:45really appreciate it. They're pretty, pretty
- 1:32:47good. They did a lot of vegan cooking for me.
- 1:32:51I've won them over on a lot of dishes. Okay.
- 1:32:57Cool. Well, anyway, it seems like, you know,
- 1:32:59there's no, you don't have any screaming objections
- 1:33:01to that. No, no. I think that like you're asking,
- 1:33:04I have no problem with you asking an unrelated,
- 1:33:06like a related, but separate question, which
- 1:33:09is what this boils down to earlier when Jacob
- 1:33:11and I, this is for our listeners, we're discussing
- 1:33:13estimates and estimate is the thing you want
- 1:33:15to estimate. When we're talking about this in
- 1:33:18an experimental context, we're talking about
- 1:33:20every experiment is a little like microcosm of
- 1:33:23the world that you actually want to test. And
- 1:33:27you can imagine very different things. Like when
- 1:33:30we were talking about defaults earlier, one kind
- 1:33:33of default tests extrapolates up to one set of
- 1:33:38environmental conditions and others extrapolate
- 1:33:40out to others. So what we're talking about here
- 1:33:42is which question do we really care about testing?
- 1:33:45And I'd say both are interesting. Both trying
- 1:33:48plant -based needs blind and trying them bringing
- 1:33:52the full set of your beliefs to bear? I think
- 1:33:54they're both fine and good questions. Yeah. Well,
- 1:33:59I guess my question is, which has more bearing
- 1:34:00on consumer acceptance? I don't know. Fair. And
- 1:34:04ultimately, that's an empirical question. I think
- 1:34:06my hypothesis is that the informed trials have
- 1:34:10more. Yeah. But we could set up a design around
- 1:34:12this. You could have those are your two options.
- 1:34:16You could either get blinded or not blinded.
- 1:34:19There's a word for this Solomon to group design
- 1:34:22or something along with that. It's not quite
- 1:34:24a Solomon. Anyway, sorry and then Test like how
- 1:34:30much were you surprised by what you tasted or
- 1:34:33something like that? You said a Solomon two groups
- 1:34:36is a Solomon a solid for group design is one
- 1:34:39in which you have half the group half treatment
- 1:34:41and half control both take a pretest so you can
- 1:34:43it's It was analogous but not the same. We are
- 1:34:47describing a two by two factorial design in which
- 1:34:50try the fake meat, try the not fake meat, blinded
- 1:34:54or not blinded. That makes sense. Next question
- 1:35:00was about your blog post on experimenter demand
- 1:35:04effects. which I will summarize for the audience
- 1:35:07when we get up to this question. Oh, even better.
- 1:35:09Why don't you, yeah, you can summarize and then
- 1:35:12I'll ask the question because it might answer
- 1:35:13my question, which was somewhat like, oh, what
- 1:35:15do you mean by this term? In the social sciences
- 1:35:19and the behavioral sciences where Jacob and I
- 1:35:21work, it's common to try to throw people off
- 1:35:24the scent when you are testing them on something
- 1:35:28because you don't want them to do give you the
- 1:35:33answer that they think you want to hear, which
- 1:35:35is a real thing that happens in laboratories.
- 1:35:37People are like, you know, eager to please, and
- 1:35:40you get the guy with the white coat or the woman
- 1:35:42with the white coat coming in and asking these
- 1:35:44questions, and you're like, oh, they're an authority
- 1:35:45figure. I'll tell her what I think is the right
- 1:35:48answer to the experiment. Sometimes you can exploit
- 1:35:52this for your own benefit. As a researcher, the
- 1:35:54Milgram experiments on shocking are really about
- 1:35:57the white coat effect. But when Jacob and I are
- 1:36:01doing things, we don't want people to know that
- 1:36:04we're trying to get them to eat less meat because
- 1:36:08they might say, we, oh, like, okay, yeah, sure,
- 1:36:13I'll eat less meat. And then when we're not watching
- 1:36:17them go eat more meat, we want to get like the
- 1:36:18true effect of how much like the experiment itself
- 1:36:23as opposed to the thing being tested as opposed
- 1:36:25to the experimental apparatus is changing their
- 1:36:28behavior. In general, having run some studies,
- 1:36:33I am less worried about experimental demand effects
- 1:36:37than I used to be because I basically think that
- 1:36:40most people believe eating meat is a natural
- 1:36:43and fine thing. And if some experimenter wants
- 1:36:46them to not eat meat, you can be like, that's
- 1:36:49a weirdo. I don't care. And basically people
- 1:36:51tell the truth or like that's one story I just
- 1:36:54thought we could tell many possible stories.
- 1:36:56I think people when you ask them, like, are you
- 1:36:59planning to eat meat tomorrow? Even if you, like,
- 1:37:02give a whole spiel about why, you know, meat
- 1:37:03is bad for the environment, I think a lot of
- 1:37:06them tend to say, like, yeah, okay, well, I do
- 1:37:07plan to eat meat tomorrow, so deal with it. Interesting.
- 1:37:12Okay, so I do think there's some survey evidence,
- 1:37:15or if you ask people, like, I think I almost
- 1:37:19wrote a mean Twitter joke about this once because
- 1:37:21some people do annual surveying about, like...
- 1:37:24New Year's resolution intentions. Like a huge
- 1:37:26chunk of people intend to reduce meat consumption
- 1:37:28every year. Like a third will report it or so.
- 1:37:30Absolutely. Intentions to behavior are not...
- 1:37:34That extrapolation is... You're right. I guess
- 1:37:38the example I gave was misleading because it's
- 1:37:40not really like... Like I was talking about intended...
- 1:37:43Like what do you intend to eat tomorrow? Yeah,
- 1:37:45that's why I was surprised. They might lie to
- 1:37:47themselves about that question. What I mean rather
- 1:37:49is... In which case you are also caught up in
- 1:37:52the... Not lie, they'll tell a story. Like, you
- 1:37:56know, as people, back when Netflix used to have
- 1:37:59queues, DVD queues, as opposed to just you can
- 1:38:03get all the content, whatever you want, people
- 1:38:06would like, you know, have Citizen Kane on their
- 1:38:09queue, and then every single time they would
- 1:38:11just order Happy Gilmore or whatever. This is
- 1:38:13like a known phenomenon that like the you who
- 1:38:15is planning is different than the you who is
- 1:38:18choosing. Definitely. In that case, I think that's
- 1:38:22real. I'm talking about the differences like,
- 1:38:25what did you eat in the past 24 hours? I think
- 1:38:28people basically like those results basically
- 1:38:32line up with the alternative, which is like watching
- 1:38:35somebody eat or observing it through like, you
- 1:38:38know, a video camera or something. Secret observing.
- 1:38:40Yeah. Okay. All right. So let me just recap.
- 1:38:47So if I'm understanding correctly, you're saying
- 1:38:49you find 24 -hour recall self -reports credible.
- 1:38:54More credible than I once did, yeah. More credible
- 1:38:56than you once did. If we had something really
- 1:39:01good and some intervention that we really believe
- 1:39:08worked based on 24 -hour recall, I would then
- 1:39:12say, OK, well, I'm sorry. I've got to see some
- 1:39:13real -world evidence. But I'm saying that when
- 1:39:16people tell us, rather than when they, when we
- 1:39:19see evidence of, for the, this is hard to articulate
- 1:39:23precisely. For the most part, when we run these
- 1:39:28interventions, people self -reported me, changes
- 1:39:33are very small. They don't tell us that like
- 1:39:37some big thing is, but it just happened. They
- 1:39:39seem to like report null results most of the
- 1:39:42time. That lends credence to the idea that they're
- 1:39:47not just telling the experimenter what they want
- 1:39:49to hear. But if you found evidence of big results,
- 1:39:58you might think, okay, well, maybe this one in
- 1:40:02particular was really good at eliciting experimenter
- 1:40:04demand effects. counter -argument there to the
- 1:40:11null results, like the meta -analytic null results,
- 1:40:13which is to say like meta -analyses that show
- 1:40:15not much of an effect on average. In both self
- 1:40:19-reported and objectively reported outcomes.
- 1:40:22Yeah, yeah, is that maybe that's because the
- 1:40:24researchers took good protections against social
- 1:40:26desirability bias and like the designs were not
- 1:40:28all that susceptible to it, like blind was successful
- 1:40:30or something like that. Yeah, the problem is
- 1:40:33that - And the interventions don't work. Yeah,
- 1:40:35and the interventions don't. I think that one's
- 1:40:36more likely because - That's a given. Yeah, yeah,
- 1:40:38because a lot of these interventions are effectively
- 1:40:40impossible to blind to. Like, Jacob and Sethu,
- 1:40:45I think, was the last name. Your, like, re -analysis
- 1:40:48of smartphone analytics paper. Yeah, yeah, yeah.
- 1:40:50I can set through 2021. You guys do these, like,
- 1:40:54or 2023, whatever it was. The interventions,
- 1:40:58you'd have to be so dumb. not to realize what
- 1:41:01the experimenter wanted when you're watching
- 1:41:03like, you know, like reading a pamphlet about
- 1:41:06why animal agriculture is bad, you know? Yeah,
- 1:41:12well, I do. I would say, right, that element
- 1:41:16is definitely very difficult to blind, but the
- 1:41:19outcome measurement could be more blinded. And
- 1:41:21in that particular study, it's not. It's I guess
- 1:41:23that does have it was in your meta analysis,
- 1:41:25right? So it has at least 24 hours of follow
- 1:41:28up. Yes, it does. But you can find the follow
- 1:41:30up more successfully. Like Maya and I did that
- 1:41:32in our documentary study where we followed up
- 1:41:35like almost a week later and with an unrelated
- 1:41:38survey. Yeah. Looking at the evidence that I've
- 1:41:46looked at, I don't see a lot of evidence that
- 1:41:51that was. strictly necessary because, again,
- 1:41:54in these situations where it is impossible to
- 1:41:56blind, we also see no results. And I'm also thinking
- 1:41:59about a paper by Astrid Dannenberg et al. from,
- 1:42:04I think, 2024 about observability in which they
- 1:42:10find that social pressure intended to reduce
- 1:42:14people's meat consumption combined with some
- 1:42:16metric where it's like whether you eat meat or
- 1:42:18not is about to be public in front of all your
- 1:42:19peers. That, in fact, increases meat consumption
- 1:42:23among some participants. And from this, I sort
- 1:42:28of am building a model of the world where you
- 1:42:31try to socialize. Even if people think there
- 1:42:33is social pressure to eat less meat, something
- 1:42:35about this issue is so deeply ingrained that
- 1:42:39they're just not having it. Sorry, can you elaborate
- 1:42:45on that? Do you have a sense of was that most
- 1:42:47of the participants? Good because i'm sure there's
- 1:42:50some participants had some reactants or whatever
- 1:42:52Yeah, it was mostly that my recollection of that
- 1:42:54is that it was male participants on average after
- 1:42:57being told that they were being observed in their
- 1:42:59meat choices were more likely to eat meat and
- 1:43:03that was independent of whether they also got
- 1:43:04the treatment telling them that uh many of like
- 1:43:0860 of their peers are vegetarian, so like that
- 1:43:11was just There's this like other effect in addition
- 1:43:15to social desirability bias. Okay Rather, social
- 1:43:19desirability bias does not obviously go in any
- 1:43:22direction here. Even if the experimenter wants
- 1:43:25the person to eat less meat and that's like dead
- 1:43:27obvious to the subject, they know that like in
- 1:43:30the world at large, eating meat is a cool thing
- 1:43:31that cool people do. Yeah, well, maybe maybe
- 1:43:36this is a good follow up question here, because
- 1:43:39this is something I'm confused about is like
- 1:43:42how to think about these different biases. So
- 1:43:48I can think of three distinct but interacting
- 1:43:52effects, and I'm curious if you think about them
- 1:43:54the same way. So number one, I think about the
- 1:43:57researcher's preferred hypothesis and the participant's
- 1:44:00beliefs about the researcher's preferred hypothesis.
- 1:44:03Number two, participant's belief that they're
- 1:44:06being observed in the experimental setting, like
- 1:44:09a Hawthorne effect. And then number three is
- 1:44:13beliefs about what is socially desirable and
- 1:44:15what society would think you should do in this
- 1:44:17sort of situation. Yeah, so okay, so we have
- 1:44:21three here. I'm going to repeat them. There's,
- 1:44:24you're a subject in the study. You have, I'm
- 1:44:27not really totally clear on how the first and
- 1:44:29the second are different, but whatever. The first
- 1:44:30is - I can elaborate, yeah. There's, you're at
- 1:44:34a study and you know what is the research you're
- 1:44:36hoping to find in order to get her paper published.
- 1:44:39Second, there's this sense that you're being
- 1:44:40observed, whether by the researcher or by like
- 1:44:43the scientific apparatus or whatever it is. Yeah.
- 1:44:45And the third is your broader sense of like,
- 1:44:48when you leave the laboratory, what is the normal
- 1:44:50thing to do? Yep. That sounds about right. So
- 1:44:53the second is different than the first, because
- 1:44:55it's like, just as pure, like, it's just the
- 1:44:58effect of someone watching you, not particularly
- 1:45:00the researcher. Yeah, and it's it's about the
- 1:45:05fact of being observed more so than what the
- 1:45:08researcher believes you'll do under observation.
- 1:45:10You might know what the researcher wants but
- 1:45:15nonetheless be totally sure of your anonymity
- 1:45:17because you have blinding in that direction,
- 1:45:19so the second is distinct. We look at these three
- 1:45:23biases, are you asking which is dominant or how
- 1:45:25to think about their relationship between them?
- 1:45:28Yeah, I'm especially interested in which ones
- 1:45:29you think are more or less applicable in analogous
- 1:45:32research. Apropos this so I wasn't totally sure
- 1:45:37Which of these you were referring to because
- 1:45:40okay, these are not all biases your beliefs about
- 1:45:42what society thinks you should do is a Behavioral
- 1:45:47determinant that's the treatment some treatments
- 1:45:50might try to alter that or this could provide
- 1:45:53a dampening treatment on Dampening effect on
- 1:45:58treatment, but that's it. That's verisimilitude.
- 1:46:00That's a real -world realism. Yeah So I wouldn't
- 1:46:03call that a bias. I would call that a context.
- 1:46:08Well, I guess to the extent that you're not expressing
- 1:46:10it genuinely, it could be a bias of like, oh,
- 1:46:14in combination with number two here, I know I'm
- 1:46:17being observed and I know this is part of a study
- 1:46:20and I know the right thing to do is eat less
- 1:46:22meat. So that's what I'm going to do. Outside
- 1:46:25the study, I wouldn't care. Yeah, again, I don't
- 1:46:33see a lot of evidence that people play ball when
- 1:46:38they think the researcher wants to see that they
- 1:46:41are eating less meat. This is totally an empirical
- 1:46:44question. We could design a study to assess the
- 1:46:47experiment or demand effects specifically. Yeah,
- 1:46:50I was surprised there wasn't more about this
- 1:46:52because it seems actually quite tractable. Not
- 1:46:54that I know of. I mean, it's just not that interesting
- 1:46:56because it's interesting to us, but it's not
- 1:46:59like... What would we do with that in terms of
- 1:47:01a campaign? Oh, design studies better. Yeah,
- 1:47:05that's true. Well, this would be to the point.
- 1:47:09Our study on Chipotle, where we include the decoy
- 1:47:12items, it's kind of designed to mitigate at least
- 1:47:18two of the three. Yeah, probably mitigates the
- 1:47:24first two better than the third, really. Yeah,
- 1:47:28so I guess one thing one version of this if we
- 1:47:30wanted to replicate it if we would run the study
- 1:47:33again in a new sample and The treatments would
- 1:47:36be the same as they are except there will be
- 1:47:38a fourth treatment group that didn't see any
- 1:47:40decoy questions Yeah, no decoys. Why should that
- 1:47:43be six treatment groups because we have to replicate
- 1:47:45the whole structure Yeah, yeah, well and you
- 1:47:50could also just analyze the people who got the
- 1:47:53target task first, yeah But we did do that, and
- 1:47:58we did not see any. Sorry. So for context for
- 1:48:00listeners, the experiment asked people to select
- 1:48:05from one of many possible taco fillings, some
- 1:48:08of which were plant -based, some of which were
- 1:48:09animal -based. And the treatment was how many
- 1:48:11plant -based options they saw, and particularly
- 1:48:14whether they saw a chicken analog thing called
- 1:48:17chickenitas that we came up with in a Chipotle
- 1:48:19-like setting. This was embedded in a series
- 1:48:23of tasks about selecting also t -shirts and pens.
- 1:48:27And we randomized the order in which people saw
- 1:48:30the questions about t -shirts, pens, and tacos
- 1:48:34with the intent to throw people off the scent
- 1:48:38about which one we actually cared about. And
- 1:48:44what we found was a that additional plant -based
- 1:48:49options reduce meat consumption by about one
- 1:48:51to two percentage points. And B, that it didn't
- 1:48:55really matter. My recollection is like no effect
- 1:48:57on net about which of these tasks you saw first.
- 1:49:01Yeah, yeah. Although I would say, you know, that
- 1:49:04comparison is much less powered than the main
- 1:49:06effects of the studies. Yeah, because effectively
- 1:49:07then we're dividing by nine because we have like
- 1:49:09people who saw the different treatment groups,
- 1:49:11the different amounts of meat, there are three
- 1:49:13options. Then we have divided by three again,
- 1:49:18based on which of the three tasks you saw first.
- 1:49:22I think that's right. I think it's only divided
- 1:49:24by three maybe. I don't think anybody's going
- 1:49:28to care about this besides us, so I'm just going
- 1:49:30to can that whole line. Sorry about for bringing
- 1:49:33it up. We'll let Ryan edit that out. But OK,
- 1:49:40cool. Well, that's at least helpful to get a
- 1:49:41better sense of where you're coming from. uh
- 1:49:44exactly and how you think about these different
- 1:49:45sources of bias um this all brings me like my
- 1:49:49last sub question of this section which is how
- 1:49:52do you apply all of this to the dominion study
- 1:49:56um like lisa kramer and peter landry landry work
- 1:50:03jacob is talking about a study in which people
- 1:50:05in which university of toronto undergrads are
- 1:50:07allowed to are encouraged to watch the documentary,
- 1:50:11Dominion, which is this incredibly upsetting,
- 1:50:13difficult documentary about how animals are treated.
- 1:50:16And then either right away or a week after they're
- 1:50:22given like a chance to get a meal, have a meal
- 1:50:26compensated for them by the researcher. So it's
- 1:50:28like this, it's called an incentive compatible
- 1:50:30measure. And - I was going to flag one other
- 1:50:39salient feature of the designer, salient to my
- 1:50:41mind at least, is that as is pretty common in
- 1:50:45econ for this sort of study, they actually included
- 1:50:48a honesty disclosure, where they implore the
- 1:50:51participants to respond honestly. And they say
- 1:50:53that some studies have found that effective documentaries,
- 1:50:56others have not. And as researchers are really
- 1:50:58agnostic, and we just want to get to the bottom
- 1:51:00of the matter, we just want to hear the truth.
- 1:51:04Which I read and I was like, I don't think they
- 1:51:06should have put this in there. I think that would
- 1:51:08make things worse rather than better. I don't
- 1:51:09know. And it's true that there's an element of
- 1:51:12lost realism, because then it's not just watching
- 1:51:15a documentary. I think whatever effects that
- 1:51:18had are probably negligible relative to the net
- 1:51:20effect, which is that people who watch this 10
- 1:51:22-minute or 16 -minute segment of Dominion a week
- 1:51:25later are like 13 .5 % less likely to choose.
- 1:51:29Maybe it's percent. Oh, yeah, 35%. less likely
- 1:51:32to choose pork in their sandwich. My belief tying
- 1:51:36this all together is that a week later, having
- 1:51:44watched this 10 -minute segment in a lab in a
- 1:51:47study a week ago, all sense of the researcher
- 1:51:50is watching me is gone. Or not gone, but very
- 1:51:56small, negligible. At that point, I think what
- 1:51:59really counts is how did this documentary affect
- 1:52:03how I view eating pork? But again, these questions
- 1:52:07are, you know, like, we don't know. This is not
- 1:52:12in the data. This is in our beliefs. Yeah, yeah.
- 1:52:19Yeah, I think my feeling there was that like
- 1:52:21this, since they have this voucher that they
- 1:52:23have to redeem, there's like this physical token
- 1:52:25of the experiment. So it could be more salient.
- 1:52:27I was just like... Oh, yeah, here's the special
- 1:52:29research voucher I was given after I watched
- 1:52:31that because like the documentary itself is very
- 1:52:33salient. Yes. And so an actual striking and the
- 1:52:38concern that Jacob is raising is that like if
- 1:52:41let's say we started it, we told like the Humane
- 1:52:43League or whoever that watching Dominion is the
- 1:52:46most effective thing that anyone can do for plant
- 1:52:49based advocacy. And then they run it and they
- 1:52:52see smaller effects than those were that were
- 1:52:55observed in the study. Subsequently they might
- 1:52:57ask why did why was that what's what's going
- 1:52:59on and one possible explanation is that like
- 1:53:02This is just this contrived experimental thing
- 1:53:05where what's actually what's being what's driving
- 1:53:08effects is not watching dominion, but rather
- 1:53:11Being an experiment Yeah, yeah and having a measurement
- 1:53:15that's followed up like directly and that goes
- 1:53:18back to the researchers and putting it all together
- 1:53:20with like the observability study showing that
- 1:53:23like being observed doesn't really seem to do
- 1:53:25much for a lot of people or rather has backlash
- 1:53:27effects for some people, or more meat rather,
- 1:53:30and like the general finding that in the meta
- 1:53:32-analysis that I co -authored with Benny Smith
- 1:53:35and Maya Mathur, and which I summarized on a
- 1:53:37previous episode, given that that paper found
- 1:53:41that self -reported and objectively recorded
- 1:53:44outcomes, including those where people did not
- 1:53:45know they were in an experiment because it was
- 1:53:47just like something that was happening at their
- 1:53:49dining hall, were about the same, I am not...
- 1:53:54I think whatever effects are here are negligible
- 1:53:58compared to the net effects of watching Dominion.
- 1:54:04Gotcha. Gotcha. Reasonable. But again, unknown.
- 1:54:08When we run experiments, we have to take guesses
- 1:54:11about what the data means. Sure. Unfortunately.
- 1:54:16Cool. Well, those are my other questions for
- 1:54:17you. Thanks for walking through them. Sure. Thank
- 1:54:20you, everyone, for listening. I kindly invite
- 1:54:23you to share this podcast with the vegans you
- 1:54:25know. Let's encourage more people to take action.
- 1:54:30Again, thank you so much for caring and I will
- 1:54:33see you next Tuesday for a new episode.