Latest / StartUp Health NOW Podcast / Use AI. Don't let it use you: Esther Dyson on human limits in an age of artificial scale

Transcript

Use AI. Don't let it use you: Esther Dyson on human limits in an age of artificial scale · StartUp Health NOW Podcast

• 0:00
  [instrumental intro music] Welcome back to Startup Health Now, the podcast where we celebrate the entrepreneurs and innovators who are transforming health. My name's Unity Stoakes, co-founder of Startup Health.
• 0:15
  Welcome, everyone. We're so grateful to be with all [laughs] of the amazing health transformers. There's lots of great members, investors, um, and most importantly, Esther, we're excited to welcome you.
• 0:28
  Just to note, Esther Dyson is traveling today, s- we will make do as best we can with technology. Uh, today is going to be a, a wide-ranging conversation. We wanna make this an interactive conversation.
• 0:42
  But maybe just to kick things off, where are you today? Okay. So I delivered my book last week, which doesn't mean, "Oh, you can read it now." It means it comes out early next year. And suddenly I'm now back on the road.
• 0:57
  Where I was, I'm in Lansing, which is the capital of Michigan, for a Boys & Girls Club meeting. And the book is called Term Limits: Human Time and AI Scale, and so it's about the value of human life.
• 1:13
  And of course, Boys & Girls Club is the epitome of helping underserved children be mentored, learn their own true value, become amazing human beings despite growing up in hard circumstances,
• 1:29
  and the epitome of what my book is around. And I was there for the selection of the Boys & Girls Club Youth of the Year for Michigan. I know the one from Muskegon, and so it was just
• 1:43
  the whole notion of charity is not about donating a lot of money and having your name on the building. Charity is about being engaged and caring, not donating. So it was, I had to be there.
• 1:57
  And here I am, and now I'm heading to Detroit Airport. [laughs] Thank you for sharing your time with all of us as well. Maybe let's just start with Term Limits and your core thesis.
• 2:07
  The title's evolved slightly over the- Yeah... past few years, so it sounds like it's Term Limits: Human Time and AI Scale. Yes. I think that kinda subtitle's really fascinating.
• 2:17
  Can you tell us what the core thesis is and how that might also relate- Yeah... to a lot of the innovators that are building and on this fireside chat? Yeah.
• 2:26
  So the core thesis in a sense is you think you know what term limits is, right? It's like politics. No. It's also term limits of human life, and the whole idea is rethink what you thought you knew.
• 2:38
  Rethink, "Oh, things shouldn't end. Good things should last forever." No. Good things have a time, and then there's probably a time to do something new.
• 2:48
  Things can change, but the very fact that we have term limits as humans gives us, if you like, permission to say, "I can't do everything. Let me find my priorities," rather than, "Oh, I'm going to live forever.
• 3:00
  I gotta do everything." And it also means you can't do everything, so you gotta share, and you have to pass things along. Don't delude yourself that you're the only one that matters.
• 3:10
  You wanna make a dent in the universe. You don't want to control everything. And it's, if you take it the right way, it's very calming. It's, it's okay to do what you can.
• 3:20
  Figure out what you can do best, and then find other people, help them do what they can do best.
• 3:26
  And then there's a whole lot more about what AI really is and what it isn't, and how we can use it and not let it manipulate us, and stuff like that. Tell us more about this concept of human time and then AI scale.
• 3:42
  There's this paradox going on right now where AI scale, it seems so fast. Uh, we're changing. The scale is epic and awesome, and then there's this constant of human time- Yeah... which obviously is much slower.
• 4:00
  And then you mix all this together, and it can be very confounding. Yeah.
• 4:04
  In particular, as you're trying to adapt, as you're trying to build, as you're trying to innovate within this paradox, how should we all be thinking about this transition today? The first thing is, "I'm so important.
• 4:17
  I need to live forever. I need to do everything. I need to control everything. I want more." And in a sense, there, there are two diseases of time. One is, "Time is too short. I need it now. I need it again.
• 4:31
  I need more," in real time, and that's addiction. And then the other is the yearning for more and for forever. You got some sense of, "I deserve not to die."
• 4:42
  And writing this book wasn't, "Oh, I know all this stuff I wanna tell people." It was more like, "I have all these questions I want to understand." And it turns out lots of these questions were addressed many
• 4:55
  millennia ago in various religions and so forth. But this notion that, yes, I will give you the knowledge of good and evil, but your life will be finite and so forth and so on.
• 5:05
  The very, yeah, it's value of human life, which you can't measure, and then it's the valuation of money, of everything else that you... And it goes back to a quote from Oscar Wilde.
• 5:18
  He knew the price of everything and the value of nothing. And, yeah, every chapter has a different epigram.
• 5:27
  And- So it sounds like it's relevant to, to everyone, not just innovators, not just investors, but everyone has term limits- Yeah... or should be thinking about them. Like it or not, you have term limits.
• 5:38
  So what do you advise as an advisor, as an investor, as a board member, to builders to keep that human-centered approach? The most important thing is
• 5:50
  don't outsource what's important, and what's important is the ethics of your business. What's important is, yeah, again, yes, your business is important. So is your family. You need to juggle that.
• 6:03
  You need to juggle long term versus short term.And but don't expect the AI. You still need people to have a company. And
• 6:14
  the most important thing in a sense is-- But there's also theory of business minds, which is they want more money.
• 6:22
  And you need to understand with the people you're working with, the people representing the companies you're working with, understand what motivates them. And don't be a sucker, don't be a cynic.
• 6:33
  Again, don't be too skeptical, but don't be too trusting. Learn to earn trust. Learn to give trust where it's deserved, and just don't be too bedazzled by AI. Use it, don't let it use you. Yeah.
• 6:50
  I'm gonna bring in a couple of transformers. Matthew Swanson, if you're there from Clutch Health, I think you have a related question or comment. Yes. Esther, really nice to see you.
• 7:02
  I think I was on one of your earlier, a couple of years back, spoke to the group as well, so nice to see you again.
• 7:07
  Clutch Health, and as I founded the company, Reciprocity Health, is really focused on finding the best ways to engage people in their healthcare and bring them into, into their health journey in a much more meaningful way.
• 7:20
  And so I was just really just wondering where maybe you've seen companies either go too far in building that layer of technology i-in the relationship building and maybe where you've seen it done well. Yeah.
• 7:34
  There are so many examples of just don't outsource the humans. Outsource the doing of the paperwork. So I have this notion of there's diabetes, which is sugar destroys your ability to process insulin.
• 7:52
  There's information diabetes, which destroys your mind's ability to process good information. So you consume it, but you don't absorb it or digest it. And then there's money diabetes, which
• 8:07
  destroys your moral fiber by turning it too focused on money, which has no value, only valuation, and get sucked into this vortex. And the people who work for you ultimately
• 8:21
  get corroded and corrupted if you do too much of that. So it's using AI too much and losing your moral fiber. But specifically, where I've seen it used well is actually two companies I'm an investor in.
• 8:34
  One is peer-to-peer counseling, and they use AI to support the counselors. Basically, they moderate chats among people, and the actual peer-to-peer support people themselves never see or talk to the AI.
• 8:51
  They're talking to the moderator. The moderator gets hints from the AI. This one person in the conversation seems to be in need of
• 8:59
  more serious help than peer-to-peer, needs a suicide counselor, or somebody's being disruptive, and they give hints to the moderator and makes the moderator's job easy.
• 9:09
  So what I like about that is it uses human scale, one moderator for maybe five people in a chat. And then there's one AI for several hundred moderators because the AI is...
• 9:22
  Yeah, you just duplicate it, and it's kinda like the right ratio. To actually talk to people, one person to four people is great. To talk to moderators, just one AI does it all. The second one is somewhat similar.
• 9:36
  It's focused on training neurodivergent people on workplace skills and interaction, and the same thing happens there. They have counselors and trainers who are then supported by the AI.
• 9:49
  All these really wonderful healthcare jobs, as you all know, require a huge amount of bureaucracy, credentials. How do you get paid through Medicaid? How do you establish your credentials? How do you...
• 10:02
  You got registered in this state but not in that state. That stuff, that's perfect for AI to do as long as the AI actually is properly trained to follow the particular state laws or- Mm-hmm... registration requirements.
• 10:17
  So AI to- Thank you... empower humans to be better and perhaps more efficient- Humans... not to replace them, right? Yes.
• 10:25
  And it's like I swim every day, and people occasionally, "You must really love watching swimming championships." And no, I swim for myself.
• 10:36
  I don't pay other people to do my swimming, and I'm not trying to watch other people. It's me. I don't want my swimming to be efficient. I want to experience it.
• 10:45
  And it's where I did a lot of the thinking for the book, in the pool. Love that. Thank you. Thank you. Thank you. I think the next question is related. Emmett Shaller from Held Health, if you could ask your question.
• 10:57
  It's so nice to meet you, Esther, and also thank you for having me, Unity. I'm excited to be here and learn. I'm Emmett Shaller, founder and CEO of Held.
• 11:07
  We are a consumer personal health operating system, and not unrelated to the topic of discussion recently with Julie Yu, as well as your post recently on Substack related to the addictive nature of products that have been built through many technology companies.
• 11:27
  Right now, we're building our optimization of key features tied to AI, and I can't help but try to optimize these metrics and for the key features that we're building toward things like time spent and engagement, which it's easy to go down the road of building products that are AI-enabled that could easily become addictive.
• 11:53
  And how do you think about, as related to what you're talking about here, how do you think about the benefits of engagement, possible addiction, and whether or not that's a bad thing if it's tied to consumer health and wellness?
• 12:09
  Can we go too far still if people are consuming more and more health? Yeah. The first red line again is consumingVersus experiencing. You can consume more stuff, but the, the challenge is how do you absorb it?
• 12:25
  How do you digest it? How do you make it part of you? And it goes back to this too much ain't enough. Too much is enough. Too much obsession with being healthy.
• 12:35
  Like, so I had cancer of the esophagus a while ago, and I try not to eat after 4:00 PM, which is difficult, especially after COVID, and there are now dinners.
• 12:46
  And I see my sleep score goes down on Aura when I eat a late dinner and so forth. At the same time, there's social events, there's people you love.
• 12:56
  I very rarely get sick, but my goal is not to be the healthiest person in the world. My goal is to have enough health that I can achieve so that I can enjoy my life and be productive and do all the other things.
• 13:11
  But everything in moderation is a much better principle than too much ain't enough. And there's orthorexia. It's when you get too... Your job is not to follow all the rules to the extreme.
• 13:28
  Your job is to follow enough rules to follow the other rules because life is finite. It's a set of trade-offs.
• 13:37
  People have said this before, if you really wanna be safe, never go out on the street, but then you're gonna die of being too sedentary. And so take a risk every once in a while.
• 13:47
  And yes, if you have a ninety-five percent chance of succeeding, the five percent where you don't succeed doesn't mean that was a bad decision. It was a good decision, and you got the wrong five percent.
• 13:59
  And at the same time, be very conscious. Most of us here on this call are-- we're in the healthcare business. We're doing something commercially successful. We have all been lucky. None of us died on the way.
• 14:13
  Most of us might have made a few mistakes, and we recovered and succeed. We're here. But a lot of the time, we had no idea how lucky we were.
• 14:21
  And that is what I've learned from spending time, ten years in some of the poorer parts of this country. People who can't afford to take risks because when they take a risk, they're gonna lose their home.
• 14:33
  It's not, "Oh, I lost my chance to win the contest." It's, "I no longer have a home." We all are extremely lucky. We have some built-in resilience that we don't even understand the value of.
• 14:45
  But take a risk every once in a while. That's good for you, too. Thank you. Thank you, Emma. Anytime you get an email from Esther, she has a great quote at the bottom, which is, I think, "Always make new mistakes."
• 14:57
  New mistakes. Next guest, next question, next health transformer, Piyush Gupta from Ambrosia. How are you? Nice to see you. I'm good. Nice to see you. And hi, Esther. Hi. I met you at Apollo House. Great.
• 15:11
  And it was great meeting you. You are the sweetest person I know, honestly. And you know- I must be too nice. No, you are actually. And I have two questions.
• 15:21
  And first question is, what is a widely accepted belief in digital health that you think is fundamentally wrong? That's a really interesting one. Again, it's whatever is good we need more of versus...
• 15:37
  And from a vendor point of view, of course, that makes sense. From the customer point of view, it doesn't. And the challenge is to have a business model where you succeed when your customer succeeds.
• 15:51
  And for a lot of stuff, you succeed when the customer no longer needs you. That's the ideal in healthcare. In health, different conditions have different ways to fix them. One is
• 16:05
  you need a certain drug to-- ins-insulin dealing with diabetes. And now it's really interesting. Are the GLP-1s going to be like insulin, you need to keep taking them?
• 16:17
  Or are they going to be more like a vaccination that will help you reset your body? And the best forms of health are the ones where you trigger the body to be healthy versus you make the body dependent on a drug.
• 16:32
  As a business, making it dependent on a drug is more lucrative. So it's moral. That's why we want human beings in healthcare, not bots that are trying to optimize profit. That's an answer to your question.
• 16:46
  No, this is perfect answer, actually. Yeah. And just a follow-up question is, we all know what is going on, and you raised a very good point.
• 16:53
  Companies benefit when people are basically become habitual of that product, continuously keep using it. Yeah. How to break that loop? Yeah.
• 17:02
  My personal feeling is if insurance, basically Medicare will stop giving incentive, basically reimbursement for the medical devices and start incentivize companies to find a cure, we will have a cure.
• 17:15
  I'm pretty cynical, but I'm not that cynical. I think if somebody actually had a cure, it would be quite a successful business. Yeah. And so I, I don't think... The bigger problem is it's not in the healthcare system.
• 17:31
  It's in the rest of the collective system, our government, in other words, that doesn't spend enough money keeping people healthy. Let's tax sugar and ultra-processed food and apply that money to schools and training.
• 17:48
  And again, we need more human people raising children. Everybody that was talking at this Boys and Girls Club thing, they were talking about mentors, about teachers, about...
• 18:00
  They weren't talking about AIs or devices or anything. It was the human beings that made them have the courage to step up and make something of themselves. So, you know, and I, I don't want to spend this whole
• 18:16
  session preaching. I wanna go and get people fighting to-Make a better system and to pay for what's valuable rather than what is high valuation. We don't even have the right language for it.
• 18:31
  We keep mixing up value and valuation. A couple of the themes that keep emerging. One is long-term spans or long-term thinking as opposed to short-term consumption, short-term hits, short-term addiction.
• 18:48
  Uh, one crazy notion I have is a potential positive of AI is its ability to think very long term. Yes.
• 18:57
  It's better to have a lifetime value of a customer than a one quarter value of a customer, and if AI can help steer businesses to have long-term value propositions for their customer base, maybe it will help infuse that into our health systems and health, uh, strategies as well in some macro or micro way.
• 19:20
  Yeah. It's great, except it's not only the customers that churn, it's the CEOs, and they don't think long term enough either, and we... But you're right.
• 19:32
  We can model the future much more effectively using AI, and- Yeah... my purpose in writing the book, as I have said, is it's not 10 timely tips for living a better life.
• 19:45
  It's more you need to think more deeply about a lot of this stuff, and now go and ask the government to. We got a couple more questions here. Rick, let's go to Rick Gersony from MedMovie. Hi, Rick.
• 19:57
  My name is Rick, and MedMovie, we focus on visual tools at the point of care. Right.
• 20:03
  My question is, as AI shapes the treatment recommendations and doctor time shrinks at the same time, what are your view of how to preserve the patient agency?
• 20:14
  And really it's like how do you view the guardrails that would make the point of care experience the best it could be, and then the follow-up be as best it can be, and then the patient have access to that information?
• 20:26
  Yeah. I think the doctors will have more time if you automate some of the boring stuff and let them spend more time with the patients. That's the huge benefit.
• 20:36
  I'd rather have more nurses because they're better trained in human contact. Nurses, in a sense, are almost more important. Doctors, more of that sort of intellectual work can be automated.
• 20:49
  I still want a doctor telling me I have cancer or something.
• 20:52
  But I think using AI to save humans time so they have more time to be human is, in theory, going in the right direction if we do it right, if we don't use the AI to create fake doctors giving fake testimonials and promoting fake supplements and so forth.
• 21:12
  So it's very interesting and related to your entire book about the shrinking time and how to increase that value, the quality of the time that you do have. Yeah, I guess what I'm trying to say is the time isn't shrinking
• 21:28
  because you can outsource the stupid stuff so that you could spend more time on the human stuff, whether it's more time with your patients or more time with your family. But AI should be relieving us of stupid friction
• 21:44
  so that we can experience the good friction, and the good friction is holding somebody's hand, delivering words of comfort through the ear. It may not be efficient, but it's what's needed at the moment. Thank you, Rick.
• 21:57
  Okay, longtime friend, longtime health transformer, Bentley Adams from Way. So I-- My question, Esther, is more of a macro level question.
• 22:07
  It might have been touched on a little, a couple different ways here, but I was looking for kind of a unique perspective from you on this.
• 22:12
  The way I look at you is you have this long-standing, really deep history with seeing technology and technological waves, seeing the hype cycles, seeing them drop, all these ups and downs.
• 22:23
  But is, what are the inherent limits of this, these tools? Do we look at these as tools just like we did wave 1.0 and wave 2.0 of, of software, for example?
• 22:31
  Is this just wave 3.0, and that there's gonna be a lot of wheat that's separated from the chaff, if you will, over the next 10 years? But in, in 10 to 20 years, what do you see the term limits or the upper bound of AI?
• 22:44
  To be honest, I really don't know because, point pardon me, certainly I think it's fundamentally different from these other ones because it really, it's going to force us to reassess our own humanity, and I hope in a good way.
• 22:59
  It's not being smart, it's being wise, and the fundamental question is: Are we going to be wise enough
• 23:06
  to use this stuff effectively, or are the smart people going to take it away and just be so good at creating predatory AI that we really get into trouble?
• 23:18
  What I do know is for the next five to 10 years, it's very much within our power to control these things or to somehow cede everything to them and let the tech bros take over
• 23:34
  and automate everything, and I think it is our way to defeat. And we're going to have a political time in the US as well where there seems to be a real shortage of products.
• 23:47
  It's not very encouraging, and it's looking like maybe we've almost hit bottom. I don't know. So there's really no simple answer to that one. Thank you, Bentley. We've got time probably for one more. Saba, how are you?
• 24:02
  Great to see you from Public Good Pharma. If you could introduce yourself. Hi from Honduras. I'm in a special economic zone in Honduras called Próspera. There's a biotech hub out here.
• 24:13
  My question was, what do you think is the strongest policy fix for misaligned incentives under the current sick care model which looks at these short-term profits?Single-payer healthcare paid for by the public with the understanding that we're investing public social infrastructure the way we need to invest in our bridges and our cell phone systems and our
• 24:38
  roads, everything else, right? You can't expect poor people to be able to fund the flourishing of their own children, but having those children flourish in the future is of public value.
• 24:53
  It's a public resource and a public asset. And that's the simplest thing.
• 24:56
  Then there's lots of other more complicated things that follow, including, again, taxes on sugar, better enforcement of various kinds of laws and regulations just on honesty and credibility and validation and so forth.
• 25:12
  But with all the savings that AI should actually provide, we should be spending more money again on raising children.
• 25:19
  And in the end, what we want is the human race, is the flourishing of the human race, and somehow we don't seem to be investing much in that.
• 25:27
  And I would say that we could extrapolate this out globally, not just single payer here in the US. Yes. The global opportunity, there's eight and a half billion people, nine billion people.
• 25:38
  So last topic, Esther, what patterns and signals are you learning most recently that you're picking up from these various different places you go? What's emerging as big insights for you?
• 25:52
  So much and, again, nothing hugely surprising. The people in poor places want dignity and respect, and they need a sense of agency. The children need more time with healthier parents, and of course,
• 26:09
  if you have unhealthy parents, somehow society or Boys & Girls Club should step in and help all the kids flourish. I think there's--
• 26:19
  The biggest problem is these divisions, and it's not just political divisions, race divisions, economy divisions. It's everybody who grows up thinks, "I'm normal, and this is how I grew up. This is what makes sense."
• 26:36
  I think everybody needs to spend-- I would like something like universal basic income as low asset bar that so that you can take that ten percent risk and go to college or train for a new job. But
• 26:53
  something, universal basic service that makes everybody spend two years before adulthood living and working in a place that's completely unlike where they grew up, and exactly how you create the regulations and so forth to make that happen.
• 27:08
  And there are predatory people that shouldn't be let loose in somebody else's neighborhood. But in general, this understanding that there's a different way to see the world. There's a different way to grow up.
• 27:21
  There's a different way to be normal.
• 27:23
  If you look at the social skills of people who are bilingual or who've lived in two separate places or-- They just, they have a better understanding that there are different ways to see the world.
• 27:38
  And in a sense, that's the most important. How do you get people to experience something different and understand that neither way is right? Nothing's all good or all bad.
• 27:50
  You should explore and understand the differences and see what's good and bad about them. Wonderful. We're so grateful for you sharing your time in between, uh, your travels.
• 27:57
  A couple of quick announcements as we wind down here. Esther's gonna be in real life at our Civic Health event in June. Yeah, in New York. In New York, June 16th.
• 28:09
  We're inviting all Health Transformers to join us there and, uh, it'll be an opportunity to continue this conversation in real life and all of us together in person, which is always wonderful.
• 28:19
  Most of all, Esther, wanna thank you for sharing your wisdom, sharing your time, and just being so supportive of the entire ecosystem.
• 28:28
  We, we greatly appreciate it and can't wait to read the book in 2027 when it's in print. [upbeat music] And finally, an invitation to all of the extraordinary Health Transformers.
• 28:43
  Whether you are a startup, someone building the future of health, or whether you are an investor, funder, or buyer of the solutions out there in the marketplace, this is an invitation to you to join the Startup Health member community.
• 28:58
  Learn more at startuphealth.com. Look forward to hopefully seeing you soon. [upbeat music]