Latest / Love Conquers Fear / #4: The Intersection of Humanity and AI: A Journey of Technology, Trust, and Transformation with Ethan Holland

Transcript

#4: The Intersection of Humanity and AI: A Journey of Technology, Trust, and Transformation with Ethan Holland · Love Conquers Fear

• 0:00
  I think we're pretty confident saying like you grew up after
• 0:02
  the Internet or before the Internet, I think we're going to
• 0:04
  have before AI and after AI. You know, this could possibly be
• 0:07
  a moment of Zen, but the moment of Zen may be the way the
• 0:11
  gazelle looks when it realizes the cheetahs got it.
• 0:14
  That's right. So I think with open source,
• 0:16
  I've always thought it's great to be able to fine tune your own
• 0:19
  model. And so that to me is a blessing
• 0:21
  and a curse because it seems very analogous to this
• 0:24
  vulcanization of the Internet where you have a vacuum or a
• 0:27
  bubble. Will this technology help
• 0:29
  humans, friends live happier lives?
• 0:31
  I think for sure when it comes to health, if your goal is to
• 0:35
  get rid of scarcity, you know, I wonder how even open source
• 0:40
  would democratize the power. The biggest fork in the road
• 0:44
  that I've been keeping an eye on is whether or not things will
• 0:46
  become open source or closed if we continue to go open source.
• 0:51
  That's one path. Very distributed and
• 0:54
  democratized very different path and then the other one is Warren
• 0:58
  Grand seeing eye I. Have concluded in my own mind
• 1:01
  this doesn't mean it's going to be true, but that open source is
• 1:04
  going to win and the reason why. Hey everyone, welcome to Love
• 1:12
  Conquers Fear, the podcast where I'm putting it all out there on
• 1:16
  the line to do absolutely everything I can to help make
• 1:21
  sure that humanity makes it to the age of abundance for all.
• 1:27
  And this is the first time that the podcast has been on the road
• 1:32
  and I cannot think of a better person to kick this off.
• 1:37
  Then one of my very, very good friends, Ethan Holland.
• 1:40
  We've known each other for a long time and we're really going
• 1:45
  to dive into a lot today on the AI front.
• 1:50
  He's been chronicling the rise of AI and what it means for a
• 1:55
  societally and all the tools every week since the launch of
• 2:00
  Chachi PT and he's now up to episode, I think 96.
• 2:06
  And so we're going to get into all that.
• 2:09
  But first, I just want to welcome Ethan and Ethan tell us
• 2:12
  about where we're at. We're doing an actual fireside
• 2:15
  chat and it's August. That's.
• 2:17
  Right. So we are on the Delmarva
• 2:20
  Peninsula, which is where I live, which is Delaware,
• 2:23
  Maryland and Virginia just east of DC.
• 2:26
  It's a pretty rural place to be having a tech conversation, but
• 2:29
  I think it's. It's a beautiful place.
• 2:31
  Pretty conducive. So we're on the Bay next to
• 2:34
  Assateague Island off of near Ocean City, MD.
• 2:37
  Yeah, this is absolutely gorgeous.
• 2:39
  And and it's a good metaphor for humanity because you and I have
• 2:43
  been having, you know, some really deep conversations
• 2:46
  actually since since I arrived. And it's really such an honor to
• 2:49
  be here. And I just want to start out,
• 2:50
  which is thanking you so much for our friendship.
• 2:54
  I mean, gosh, we, I think it's been over 20 years now, right?
• 2:58
  Yeah, almost 30. We're we're in that.
• 3:00
  Oh my gosh, OK, and 25. Yeah, yeah.
• 3:03
  That's right. That's right.
• 3:03
  That's right. Yeah.
• 3:05
  All right, So let's start out with a bit about you.
• 3:09
  Tell tell us, you know, how did you grow up?
• 3:11
  Tell us about your parents, like your big aspirations growing up.
• 3:15
  You know how you got to this moment of caring so much about
• 3:20
  AI? Sure.
• 3:21
  And interrupt me if I'm taking too long because I want to make
• 3:23
  sure we keep the keep on track. But yeah, so I was raised here
• 3:27
  on this peninsula, very rural area.
• 3:30
  And you know, I think I was learned.
• 3:32
  I, I learned to be curious in college.
• 3:34
  So I was what I'd call a jellyfish in high school where I
• 3:37
  just floated around and absorbed and I wasn't really, I didn't
• 3:41
  have a lot of causality for myself.
• 3:42
  And then I went to Swarthmore College outside of Philly and I
• 3:46
  had some great professors and I learned to get curious.
• 3:49
  And I think once you get curious, it's hard to turn it
• 3:51
  off. And so I was just lucky to be at
• 3:56
  the same time as the Internet, probably a little bit like you,
• 3:59
  our generation was just was in a good spot.
• 4:02
  So we got to kind of grow at the same pace as the Internet.
• 4:06
  We're the last generation that will remember life before the
• 4:09
  Internet, right? So I think you know that how I
• 4:11
  got here from there was I don't know how much we want to talk
• 4:14
  about How I Met you, but that is probably a good.
• 4:18
  Second, let's first talk about you and your and your childhood
• 4:21
  and OK. Yeah, so I'm, I'm an only child.
• 4:25
  I was raised, I was very fortunate.
• 4:27
  So I think I've have carried that with me as something that
• 4:30
  I'm self-conscious about, you know, being my own person.
• 4:34
  So I had a very successful dad who was a strong personality in
• 4:38
  in a small town environment. He was a judge and my mom went
• 4:41
  to Harvard. So here I am in Milford, DE and
• 4:44
  from 4th grade till when I was a a senior in high school, my mom
• 4:48
  commuted to Boston. She left every Sunday and came
• 4:51
  back on Thursdays and went to grad school.
• 4:54
  And my dad was a judge and was friends with Ginsburg and Scalia
• 4:58
  and all these people. So I had to kind of figure out
• 5:01
  who I was, so I think. And he became a Supreme Court.
• 5:03
  He was on the Delaware Supreme Court and that certainly informs
• 5:07
  who I am now is sort of deciding, you know, what do I
• 5:10
  bring to the table and, and who am I?
• 5:12
  And along the way I discovered I just love emerging technology
• 5:16
  and I'm pretty good at it. Yeah, you're very good at it.
• 5:19
  So I think, yeah, so I have AI have a big shadow and a lot of
• 5:22
  shoes to fill and I think I carry that with me.
• 5:24
  So my childhood in that sense, being in a rural area, I was
• 5:27
  not, you know, I'm not in next door, I'm not in like if you've
• 5:30
  read outliers, I didn't have a Unix machine in my high school,
• 5:34
  like Bill Gates or something, you know, so I was in a in this
• 5:38
  rural area. So I didn't get exposed to that
• 5:40
  sort of thing until college. And then we did have a Unix
• 5:43
  machine in college. It had all these resources and
• 5:46
  that was where I really learned how much I loved it.
• 5:48
  Yeah, my in my Internet game in college, which I started in
• 5:52
  1990, it was on running on a Solaris workstation, you know,
• 5:57
  in Unix. And actually I got I got kicked
• 6:01
  out of the school because it would consume all the resources,
• 6:05
  almost like a parasite because my game became very popular.
• 6:09
  It's a implement in a way. It was in a way, but but it was
• 6:12
  it was also hard to get kicked out of the CS lab.
• 6:16
  I was not ACS student and then the game like parasitically went
• 6:20
  to other servers until finally a company picked it up as the
• 6:24
  sponsor. They're like one of the early
• 6:26
  web hosting companies and but it but it apparently by 92 it was
• 6:31
  the most popular game on the Internet.
• 6:33
  But it all started with with Unix and C.
• 6:35
  Yeah, we had Sun, we had Sun Microsystems using machines and
• 6:39
  we had those Next boxes. Do you remember Next?
• 6:41
  I do remember next we had those in the CS lab too.
• 6:44
  Again, I was not ACS student, I was an MIS student, but but we
• 6:47
  but I spent most of my time in the CS lab that's.
• 6:50
  Probably a good thing to point out as I was an as an English
• 6:52
  major. So I was, I was an English major
• 6:55
  and then I got so into computers and I would basically Friday
• 6:58
  night said if I was drinking a six pack, I'd be in the computer
• 7:01
  lab. I wasn't at a party, I was doing
• 7:03
  whatever on the Unix. Machines.
• 7:05
  Yeah, yeah. And and I'm not a computer
• 7:07
  science person either. Right, right.
• 7:09
  Yeah, 'cause you were, you were an English major, right?
• 7:11
  Yeah, yeah. What made you decide to be an
• 7:13
  English major? Like what was the my, my mom, by
• 7:15
  the way, we didn't talk about this, but my mom, Brenda, she
• 7:19
  was a dual major between English and mathematics, and her true
• 7:22
  Love Actually was English, like the classics.
• 7:25
  She she just absolutely loved it.
• 7:27
  I liked obviously just learning it through through literature
• 7:31
  and and immersing myself. I started out as a bio major and
• 7:34
  then I just kept signing up for English courses and I had really
• 7:37
  good professors. So I was actually a black
• 7:38
  studies concentrator. So a lot of Harlem Renaissance
• 7:41
  poetry and things like that. And just learning about the
• 7:44
  world through reading was fantastic.
• 7:46
  And I stuck with it almost kind of just made its own path.
• 7:49
  It wasn't something where I went into school saying this is what
• 7:52
  I want to do, it's just what I ended up taking.
• 7:54
  And in a way, you were taking a very humanitarian approach to it
• 7:58
  'cause you were digging into layers of history through the
• 8:02
  black American experience early on.
• 8:03
  And that's very Swarthmore cause Swarthmore being liberal arts,
• 8:06
  very small school. I don't even think you're tend
• 8:09
  to clear your major till junior year.
• 8:11
  OK. So you're meant to be figuring
• 8:13
  it out at school? Yeah, very cool.
• 8:18
  And then and then that that passion for technology just kind
• 8:22
  of block me through that because I know, you know, you got an
• 8:24
  early start in e-commerce and you were, you and I were both in
• 8:27
  the very early days of of that renaissance and now there's a
• 8:31
  whole new birth of AI meeting e-commerce.
• 8:35
  We'll get into that later on. But but you know, you and I
• 8:38
  originally met I think at a shop that or conference.
• 8:41
  Correct in Las Vegas so we met so to start with how I got into
• 8:46
  the to that world partially was through English reading so I
• 8:49
  read Neuromancer as part of my curriculum a lot of great
• 8:52
  science one of my. Favorite books actually.
• 8:54
  You know, Ursula Le Guin, Left Hand of Darkness, all these
• 8:57
  really great sci-fi fantasy type books I was reading.
• 9:00
  So that opened my mind to thinking about humanity and
• 9:04
  technology. And then I had a a guy on my
• 9:07
  hall was named Justin Hall, and this was 1993.
• 9:13
  He was the first Blogger ever. And so you can Wikipedia.
• 9:17
  He's often considered the father of blogging personal.
• 9:20
  Publishing What's his name again?
• 9:21
  Justin. Hall, Justin Hall, He had a
• 9:23
  website calledlinks.net. It was very similar to Yahoo
• 9:27
  because back then you could literally document the whole
• 9:29
  Internet and just. I had to have run across it
• 9:32
  because I was that early. So he he was a over sharer, but
• 9:36
  in particular he he took a semester off and he went to help
• 9:39
  found Wired magazine. And this is before Federated
• 9:43
  Media, before banner ads and John Patel.
• 9:45
  This was really at the beginning of publishing online.
• 9:48
  I was I was one of their first subscribers.
• 9:50
  I imagine you were reading it beyond too, Bruce Sterling and
• 9:53
  all. All yeah, Bruce Spimes.
• 9:55
  He's a Texas guy, right Bruce? Bruce Bruce would give the
• 9:59
  single best keynote at South by Southwest.
• 10:02
  And one of the things if you're Bruce Sterling fan is he does a
• 10:05
  state of the world with my friend John Lipkowski every year
• 10:10
  on the well. And so it's a really fun read.
• 10:15
  It's a little bit snarky because Bruce is definitely a strong
• 10:18
  satirist and he's not so positive on humanity like I am.
• 10:22
  But but he but he's genius that the most intimidating dinner I
• 10:26
  ever had was with Bruce Sterling because I put him on this
• 10:29
  pedestal of, you know, one of my favorite sci-fi authors.
• 10:33
  And he did at this point, he definitely is much, much smarter
• 10:36
  than me. I mean he his his polymath like
• 10:39
  knowledge was just kind of off the charts.
• 10:41
  If we're going back to the well, Justin's mentor was Howard
• 10:44
  Reingold, who is a big well guy, all San Francisco ethos and the
• 10:49
  Grateful Dead and all the hate Ashbury stuff.
• 10:51
  So how so? That is a great way to introduce
• 10:53
  Justin's influence on me. Justin came by my my dorm one
• 10:57
  day and, and said, you know, you have to check out this thing
• 11:00
  called the Internet. And he, his pitch was his father
• 11:06
  committed suicide. And I think he was in high
• 11:08
  school at the time when his father died.
• 11:11
  And so here we are sophomore in college.
• 11:13
  And he said, look, I have been alone with this very hard
• 11:17
  journey of losing my father and I've never been able to find
• 11:21
  anyone like me. And there's this new thing
• 11:23
  called the Internet, and I found 10 people in this community
• 11:27
  online. I don't know if it was Gopher or
• 11:30
  Telnet or what he was using back then.
• 11:31
  It may have been Usenet. Some sort of yeah, Usenet.
• 11:34
  Usenet was like the original message.
• 11:35
  Board but he found people like him and he was like this is
• 11:38
  going to change the world because people who feel lonely
• 11:40
  and and you know me coming from Delmarva very rural area, I
• 11:44
  think that's a normal feeling. People are a little isolated and
• 11:46
  you don't you can't find your tribe.
• 11:48
  He was like look, everyone's going to be able to find their
• 11:50
  tribe all right now the irony is I don't think he predicted the
• 11:53
  balkanization it would cause all these tribes but he was looking
• 11:57
  at it as the you know, utopian version of the Internet right
• 12:02
  and so he made his his calling to teach people HTML right and
• 12:07
  to teach them how to blog how to use fetch or FTP and or or just
• 12:12
  you know do it right in the town.
• 12:13
  You know you could do the code on server.
• 12:16
  My game was only accessible via telnet.
• 12:18
  So he really broke the ice for me and showed me this bigger
• 12:22
  picture of why the Internet was good and exciting.
• 12:25
  And then from there, he taught me the tools through HTML and
• 12:27
  things, how to buy a domain, all that I can, all those things we
• 12:31
  don't know about anymore. Right?
• 12:32
  Right. And so he was a huge influence
• 12:34
  on me. And so I started blogging too.
• 12:37
  And what year was that? This is 93.
• 12:39
  I think I was blogging in 94. OK, yeah, early.
• 12:43
  And at that point, there's not a blog.
• 12:44
  There's no WordPress. You just wrote the HTML and
• 12:47
  uploaded it. I wonder if he ever played my
• 12:49
  game because I started my game. It was called Renegade Outpost
• 12:52
  and I started in 1990. He may.
• 12:54
  He probably did. And now the difference between
• 12:56
  him and I, we had a very interesting path.
• 12:58
  He overshared to the point that I, I, I want to speak for him,
• 13:02
  but maybe got addicted to it. But he shared everything.
• 13:06
  And so that open kimono approach was so was so early.
• 13:10
  The Internet wasn't as toxic as it is now.
• 13:13
  So as he grew and became more and more open and the Internet
• 13:15
  got more and more people, suddenly other people were over
• 13:18
  sharing. He wasn't the only Blogger,
• 13:20
  right? Right.
• 13:21
  And so he actually withdrew and his his solution was to stop
• 13:26
  because like anyone who over shares or gets addicted, you
• 13:29
  know, in that sort of sense, it was ruining his authentic
• 13:32
  moments with his friends and family.
• 13:34
  So it withdrew. And for me, you know, I was a
• 13:37
  kid and I was drinking and, and I, you know, I just figured to
• 13:40
  myself, maybe I don't need to say everything.
• 13:43
  Right, right, right now. And but he did give me that
• 13:47
  introduction that led me to just when I graduated, going into
• 13:50
  e-commerce. So it was not hard to, you know,
• 13:55
  it's pretty easy when and when you've already kind of grown up
• 13:57
  with the Internet and a legacy business has not.
• 14:00
  So my first job was at a catalog company, which is when I met
• 14:03
  you, I just graduated and they wanted to put their catalog
• 14:06
  online. So taking print and turning it
• 14:09
  into e-commerce was my first job.
• 14:12
  And that was Burpee. That was burpee, Burpee seeds.
• 14:15
  Burpee seeds, which had been around since 1800s.
• 14:17
  Yep, they started mail order chickens.
• 14:20
  OK, I didn't know that. I just thought they started.
• 14:23
  With the and I think the seed was seed to feed the chickens
• 14:26
  and that evolved to flowers and vegetables.
• 14:29
  I never realized that Burpee, a plant company started with or
• 14:32
  seed company started with chickens.
• 14:35
  This is this is a there's a good metaphor.
• 14:37
  Which came first, the the the chicken or the egg or in this
• 14:40
  case the seed? The chicken or the diapers and
• 14:43
  they they innovative with some really neat things.
• 14:44
  So there was a lot of storytelling opportunities
• 14:46
  where, for example, yellow corn was not yellow.
• 14:50
  They bred it to look yellow because people couldn't afford
• 14:53
  butter. OK, you know, and and think
• 14:55
  things like that and Victory Gardens during World War Two.
• 14:58
  It was a big part of their DNA right, right.
• 15:00
  But the The funny thing in hindsight is they did not
• 15:04
  believe anyone would ever want to buy anything online because
• 15:07
  you need to hold the catalog, which I think our generation
• 15:10
  would remember the j.crew catalog or.
• 15:12
  Yeah, sure, pays through it. That you want get those to your
• 15:15
  parents or whatever. So they they the big, big win
• 15:18
  for them was when they realized there's no printing costs.
• 15:21
  You can put more than one picture for each product, and
• 15:24
  you can make a longer description, right right then
• 15:26
  they started realizing this is kind of fun.
• 15:28
  Right, right. So one of the stories that we
• 15:32
  were talking about over dinner last night with regards to
• 15:35
  Burpee, that's one of my favorite stories about the whole
• 15:39
  history of bizarre voice, which you know, you, you are always a
• 15:43
  big fan of the company. And we, we pioneered customer
• 15:46
  reviews in e-commerce sites. When we started, there were only
• 15:49
  three retailers in the entire United States who had customer
• 15:52
  reviews. That was just 2005.
• 15:54
  That was that was two years before the iPhone launched,
• 15:58
  Facebook was closed to the public.
• 16:00
  It was just a handful of schools.
• 16:01
  There's no Snapchat, no Twitter, no TikTok, etcetera.
• 16:04
  So there's a social media wave about to hit.
• 16:07
  But early on there were a lot of retailers and brands that were
• 16:13
  really sensitive about it. And Burpee actually was one of
• 16:16
  those. So here you are, like with this
• 16:18
  background of being an early tech pioneer and adopter, you
• 16:23
  know, beginning of Wired magazine and everything else.
• 16:26
  And I remember your boss was particularly hard on me at the
• 16:33
  beginning. He became a big fan later on.
• 16:36
  But maybe we should just tell that story.
• 16:37
  I'd love for you to tell it from your lens.
• 16:39
  Sure, because I'll tell you my, I'll tell you my version from my
• 16:42
  lens. So my lens was so I, I, you
• 16:46
  know, we used core metrics, which was your first company, or
• 16:49
  at least the first I've worked with.
• 16:50
  Yeah, for all the analytics. And I was brand new.
• 16:53
  So a good segue to just how brand new I was was I didn't
• 16:57
  realize that salespeople had an ulterior motive.
• 17:00
  So the same week I met you in Las Vegas at shop.org, I flew
• 17:04
  out for the first. It was my first conference,
• 17:06
  first time in Vegas. I think I was in the Venetian
• 17:08
  for six days by myself. So it was like my.
• 17:12
  What is going on? Firehose in Vegas and all these
• 17:16
  people wanted to Take Me Out to dinner and I thought it was, I
• 17:20
  thought it was amazing. So like Yahoo at the time, it
• 17:23
  was like the big one. And we went out to dinner and
• 17:25
  then they had a suite at like town at night.
• 17:27
  There, but like the night has. Great, great things.
• 17:29
  And I just thought, wow, these people are my people, right?
• 17:34
  And then of course, at the end of the night, they think they're
• 17:36
  thinking like, when are we going to do a deal or something,
• 17:38
  right, Right. I'm like, what's next?
• 17:39
  Where are we headed? Right.
• 17:40
  Where are we going? Right.
• 17:42
  So I was a little naive and I didn't realize that it was a
• 17:44
  transactional relationship. And obviously I think that's
• 17:48
  what made you stand out because you were not transactional.
• 17:52
  But when I met you and we were talking about reviews, it just
• 17:57
  made sense because I was all about information was to be
• 17:59
  free, be transparent. You want to show you if you're
• 18:03
  great, showcase it because people believe they're friends.
• 18:07
  So kind of it's, I think in hindsight, all these things are
• 18:10
  very intuitive. Right and by the way, just so
• 18:12
  you know, because you may not know this, like we named bizarre
• 18:16
  boys Brant Barton and I, my Co founder after chapter 4 of the
• 18:21
  clue train manifesto, which that chapter, that book was written
• 18:26
  in 1999 and that chapter was called markets our
• 18:29
  conversations. So we're trying to symbolize
• 18:32
  that word of mouth was always the most important because it's
• 18:36
  authentic, right? It wasn't marketing.
• 18:38
  It was me telling you about a book I just read.
• 18:40
  We're talking about a lot of books last night and that's
• 18:43
  going to have a lot more impact on you than an ad.
• 18:45
  And we're trying to tell the retail market and the brands is
• 18:50
  like your customers are doing this all the time.
• 18:52
  They're talking whether you like it or not, and you should like
• 18:54
  it because it's driving yourselves.
• 18:56
  I think when you're young and when I was young, you don't
• 18:59
  realize the stakes, right? So in my mind, I just thought it
• 19:02
  very simple, you know, burpees, an old company, they're beloved.
• 19:04
  They've got the best seeds, the right product.
• 19:06
  Like we have a great customer base.
• 19:08
  This will just be a mirror. Yeah.
• 19:10
  And shine the light on just how much love our customers have.
• 19:14
  And I think the the in hindsight, the parts that's
• 19:16
  naive is when you've built a big house, you don't want to mess it
• 19:21
  up, right? So that conservativism of not
• 19:23
  wanting to and what if someone says something bad like is
• 19:26
  little they people thought I was being naive, right?
• 19:28
  Because all it takes is 1 bad apple to kind of poison the
• 19:31
  well, so to speak, right? And.
• 19:32
  Well, that's the fear. It's actually usually not the
• 19:35
  case, but yes, that is that is the old story.
• 19:38
  That was the fear at the time. And so I came in pretty eager
• 19:43
  Beaver idealist, you know, coming in saying we need to have
• 19:45
  reviews, not expecting any. Problem Bush rack here.
• 19:49
  Like, didn't you all grow up like I did?
• 19:51
  I love them as a slam dunk. And so I I brought you into I
• 19:55
  got got us you on the phone because I met you in Vegas.
• 19:57
  I'm like, I come back and I'm like those guys.
• 19:59
  You need to meet my friend Brett and hear about this new company.
• 20:02
  And I think at the time you had like Petco and maybe Golfsmith
• 20:05
  and. Compusa.
• 20:06
  Compusa but it's like 5 I think. Maybe we had five customers.
• 20:10
  Yeah, I can't remember how many. We had but and maybe that would
• 20:12
  be the same way to you. Can you can explain how the call
• 20:16
  went when you got? Home yeah so I mean on the call
• 20:19
  your boss got really animated and started to yell at me and I
• 20:25
  put the phone on mute on my side.
• 20:28
  This is pre like zoom or any of this stuff.
• 20:31
  And I told the the other people in the room, I said, man, this
• 20:37
  airplane is going down hard and we are going to crash and burn
• 20:40
  in flames, you know? And then I got the idea.
• 20:45
  I was like, Oh my gosh, OK, I know what to say.
• 20:48
  And they're like, what are you going to say?
• 20:49
  And I'm like, just just watch this will be interesting to see
• 20:51
  what happens. So I took it off mute and I
• 20:55
  said, I said, Don, I was like, do you know anything about
• 21:01
  Kryptonite bike locks? And he's like, I've no idea what
• 21:04
  you're talking about. Like what?
• 21:05
  What is this Kryptonite thing? And I said, well, if you go and
• 21:09
  search it on Google, the top in the top three organic search
• 21:15
  results, what's called SEO, you'll find that there's a
• 21:20
  YouTube video showing how to pick a Kryptonite bike lock with
• 21:25
  a big pen. And he's like, OK, that, that's
• 21:30
  interesting. And I said, look, you know,
• 21:32
  respectfully, Burpee has a lot at stake.
• 21:36
  It's an old brand. And do you want to be the next
• 21:39
  Kryptonite bike lock or do you want to know when you're having
• 21:41
  a problem like that? And then I put it on mute.
• 21:46
  And there was a long pause. I mean, I don't know what it
• 21:50
  felt like on your side, but it was like, felt like a very long
• 21:53
  minute on my side. And then he said, damn it, we'll
• 21:57
  do it right. And that's how Burphy became a
• 22:00
  customer. And then and then what happened
• 22:01
  after that? I was in the room so I think he
• 22:04
  was at his desk and I was here and I knew about the kryptonite
• 22:06
  bike lock thing and I was like, OK, here we go.
• 22:08
  It was similar thing and I think he looked at me and I looked at
• 22:11
  him and he he liked me. He ended up being in my wedding.
• 22:13
  So we he was a good. Dude no, he ended up being a
• 22:16
  huge champion. And he, he, you know, I remember
• 22:19
  who was the guy from Bain who did the Net Promoter scores at
• 22:22
  Fred, Fred Reichheld Fred. Reichheld and he was called.
• 22:25
  The ultimate question was the name of this.
• 22:26
  Book And so you brought this up when we were talking yesterday
• 22:29
  about detractors and evangelists and there's like these 4
• 22:33
  quadrants I. Think promoters.
• 22:34
  He had of the net promoters and that is a case study of what
• 22:37
  happened with with me and Don because we launched a product.
• 22:42
  It was one of our first physical products that you're called the
• 22:45
  Burpee seed starting kit. And that was meant to just kind
• 22:49
  of help people to seed starting is a very niche thing and it's
• 22:53
  difficult. Most people would just buy a
• 22:55
  plant at a nursery. You go buy a tomato plant.
• 22:58
  So the people that grow things from seeds are another level.
• 23:01
  But to try to make that more accessible, we had the seed
• 23:04
  starting kit that would hopefully set and forget it on
• 23:07
  your window sill, right? And that was our first product
• 23:10
  with ratings and reviews and it got creamed and it had all sorts
• 23:16
  of problems. And what was really fun to watch
• 23:20
  was how much Burpee listened. And then we released the seed
• 23:24
  starter two. It was our best selling product
• 23:27
  and five stars. And we basically were very
• 23:30
  honest and said, look, thank you for telling us about our
• 23:32
  products. Look what we did, thanks to you,
• 23:34
  it is now the best seed starting kit you could possibly get,
• 23:38
  right? Right.
• 23:39
  And I think that was probably, from your point of view a common
• 23:42
  story of some clients. Not, not at that point in time
• 23:45
  it it became more common, but those early case studies, you
• 23:49
  know, we're sitting in front of a fire, like become fire
• 23:52
  starters for a movement. Because really, when when Brant
• 23:56
  and I started Bizarre Voice, what was in our hearts and the
• 23:59
  reason that we were so passionate about it from the
• 24:02
  beginning was it actually was a power of the people movement.
• 24:05
  We wanted people to go back to the humanity of realizing that
• 24:11
  word of mouth was the most powerful form of marketing.
• 24:15
  One of the most beautiful things about human beings is that we're
• 24:17
  wired actually to help each other and to suggest things that
• 24:21
  we think each other will enjoy. And you know, it's not like
• 24:24
  we're looking at our watch and thinking why I just wasted 10
• 24:27
  minutes talking about my favorite movie.
• 24:29
  Like we're trying to create connection that way because we
• 24:32
  live in a, we live in a physical world and we're having this
• 24:36
  physical experience and we want each other to have a great
• 24:39
  experience. It is interesting because
• 24:41
  because it was before Facebook, that era, the word community was
• 24:47
  used a lot next to ratings and reviews.
• 24:50
  It was like your customer community because there was no
• 24:54
  or algorithms, people still went to websites as destinations and
• 24:59
  then these communities would be there and people would get to
• 25:01
  know each other and you had a lot of trust because you'd hear
• 25:04
  the people would share their photos too.
• 25:06
  We did a lot of that Furby because gardening, you want to
• 25:09
  show off your big tomatoes and stuff.
• 25:10
  So our community within because of bizarre voice.
• 25:14
  It was really fun and it was it was a huge win.
• 25:17
  And Don, I'm happy you had that experience with Don.
• 25:20
  Yeah, he became a big evangelist.
• 25:22
  Yeah, yeah. And, and, and actually in my
• 25:23
  career, I can tell you that the biggest detractors initially
• 25:29
  that you turned into, you know, you turn their skepticism in a
• 25:35
  way, through actual facts and experience into a promoter, they
• 25:41
  become the biggest net promoters because the skeptics can sell
• 25:44
  everybody else, you know, because they've already gone in
• 25:48
  thinking this is going to fail. And I was surprised that it
• 25:51
  didn't. But I really owe a lot to, you
• 25:54
  know, you making that bet on me to bring me in there in the 1st
• 25:57
  place could turn up very differently.
• 25:59
  And, and Don for taking the risk and, and you know, it's, it's,
• 26:03
  it's it's kind of cool that the product that launched was in a
• 26:08
  way, it's burpees kryptonite lock.
• 26:11
  True. That's a good point.
• 26:13
  And then and then it turned. Yeah.
• 26:15
  Well, I think the the other really big revelation was that
• 26:19
  you need a few bad reviews to to lift up the good ones.
• 26:23
  Yes. And so they're actually net
• 26:25
  positive. If you have a couple bad ones
• 26:26
  because people are going to get a dud, that's OK.
• 26:29
  Right, because it's authenticity.
• 26:30
  Yeah, it's, it's if it, if it was all positive.
• 26:32
  We don't live in a world where everything's positive.
• 26:35
  And not every criticism is applicable to every customer.
• 26:37
  That doesn't even matter sometimes.
• 26:39
  That's exactly right. Like 2.
• 26:40
  Sunglasses are meant to be cheap.
• 26:41
  Yeah, yeah, yeah. Very cool.
• 26:44
  So, all right, so Burpee, let's close out that chapter.
• 26:47
  It was a great, it's a great chapter.
• 26:49
  And I know you had many other chapters and you were a real
• 26:52
  pioneer in e-commerce and we could talk about that all day
• 26:54
  long. I want to go to the thematic of
• 26:58
  this this podcast, but right before I go to that, I just want
• 27:02
  to say one thing. Like on social media and in real
• 27:07
  life, I can tell you one of the best dads I know.
• 27:10
  Like it's unbelievable how great you are with your daughters and
• 27:15
  how amazing they are at dance and everything else.
• 27:17
  So I just want to make sure that's said because we didn't
• 27:19
  talk about family, but like part of what I love about you is
• 27:22
  you're an amazing human being, not just an amazing business
• 27:27
  leader and AI forecaster, etcetera.
• 27:30
  Likewise, and, and you've always had faith in me, which is
• 27:33
  something I think, you know, for anyone who knows you who's
• 27:36
  watching this podcast, you know, Brett is a tremendously
• 27:39
  successful person and he's liked me since day one.
• 27:43
  And it's never been about my company or what I can offer.
• 27:46
  It's always been a genuine, authentic relationship.
• 27:49
  And that's very special to me. And and the fact that you've
• 27:53
  always rooted for me, it means the world.
• 27:54
  So likewise that goes, that's, that's, that's what a true
• 27:57
  friendship's all about. And so with the book I'm
• 28:00
  writing, you know which, which again is called Love Conquers
• 28:03
  Fear, Humanity, AI, and the age of Abundance for All.
• 28:08
  And this podcast is a, you know, predecessor to it.
• 28:11
  I really want to know if first of all, you agree with me that
• 28:16
  within you know our lifetimes that we can manifest humanity,
• 28:22
  can manifest an age of abundance for all.
• 28:25
  Well I one thing I'll say to begin is that I am laser focused
• 28:30
  on this topic. So when Cat GPT came out, it was
• 28:36
  the first time since Justin Hall came to my room and said the
• 28:40
  Internet is going to change everything.
• 28:43
  And I was lucky enough to live through it and watch it.
• 28:45
  I'll actually blow off mobile, even though that's a pretty big
• 28:48
  part of the arc as well. You know, just the Internet
• 28:52
  itself, which led to all these other things, including mobile.
• 28:55
  Mobile social media. All that stuff, I think this is
• 28:59
  it. And so when I, when I first saw
• 29:01
  GPTI didn't see GPT, I saw what I consider multimodality.
• 29:08
  It was just, it was just a spark.
• 29:10
  So I think whether it's embodied robots or all the things we now
• 29:13
  talk about, I immediately woke up and thought, OK, this is
• 29:17
  Round 2. So I'm 51 years old now.
• 29:20
  And when I was in my 20s, I got to watch the Internet and I knew
• 29:24
  I was pretty good at it. I'm not usually too much of A
• 29:27
  toot my own horn type of person, but I definitely saw where the
• 29:30
  puck was going with the Internet and I thought, you know what?
• 29:35
  You definitely saw where the. I had just been given a
• 29:37
  wonderful gift of Emerson, who I had not read before.
• 29:41
  But Emerson self-reliance said that the eye is placed in front
• 29:46
  of a particular ray of light, so it should testify about that
• 29:49
  ray. And I thought that's kind of
• 29:51
  cool. Like if I have a gift, I need to
• 29:54
  use it and this is what I love. So not only is it sort of it's
• 29:58
  just it's just something that woke me up immediately to say
• 30:01
  this is your chance to really dive in and not be embarrassed
• 30:04
  about being a little foolish and putting your stake down.
• 30:07
  And so let's go right, right. And so in, in, in September,
• 30:11
  whenever open it, whatever. In November it was whenever it
• 30:13
  came out. ChatGPT.
• 30:14
  It was. It was November 30th, 2022.
• 30:17
  November 30th, right? So I sat on it and and was very
• 30:21
  aware of it. And then in January I posted to
• 30:24
  Facebook. This is a line.
• 30:26
  It's like a ring in the tree where we will look back on this
• 30:28
  and just like we did the Internet or before the Internet.
• 30:32
  I think we're pretty confident saying like you grew up after
• 30:35
  the Internet or before the Internet.
• 30:37
  I think we're going to have before AI and after.
• 30:39
  AI for sure. So I wrote.
• 30:40
  I wrote. Which I would say is much bigger
• 30:42
  than pre Internet. Much bigger.
• 30:44
  And so we'll get to that next. So I I wrote a post.
• 30:48
  I said, you know, this could possibly be a moment of Zen, but
• 30:52
  the moment of Zen may be the way the gazelle looks when it
• 30:55
  realizes the cheetahs got it. That's right.
• 30:58
  That was supposed. I was trying to remember what it
• 31:00
  was. It was that and it really sat
• 31:02
  with me. Because you see that look in its
• 31:04
  eyes, if you've ever watched a cheetah get taken down or take
• 31:07
  down a gazelle, they're finally just like, all right, right.
• 31:10
  Here we are, it finally succumbs, right?
• 31:12
  Right. That was what I wrote and I felt
• 31:14
  I made it public. I didn't do that in the past
• 31:17
  very much, right? And so I decided I would, I
• 31:19
  would really live. So I turned on my Facebook post
• 31:21
  to public, my Instagram. I never was like that.
• 31:24
  And and I decided I'm going to put it in writing and I'm going
• 31:26
  to commit. So that's step one.
• 31:29
  And then I became rabid about it, which we can talk about
• 31:32
  later. But my answer about, I think
• 31:34
  your, your question is, do I agree that humanity can make it?
• 31:38
  Is that that's questionable. To the age of abundance for.
• 31:41
  All to the age of. Abundance, by the way, for all.
• 31:44
  I mean, you know, one of the, one of the, the people I'm
• 31:47
  covering in my book is Sky Callaboo who is from Chad, which
• 31:52
  I knew nothing about Chad, but it's 187 out of 189 on the human
• 31:57
  development Index. Their GDP per person is $660
• 32:01
  versus ours around 85,000 in the US And if we can create an Oasis
• 32:07
  in Chad through the rise of AI, robots, etcetera, and all the
• 32:11
  clean water supply energy they need and we can create an Oasis
• 32:16
  anywhere. So I kind of start with that in
• 32:18
  the book because if we can do it in Chad, that means humanity is
• 32:23
  going to be really pretty incredible.
• 32:26
  I think the crux of this first question is, is it, is this in
• 32:30
  our lifetime? Is that a question?
• 32:31
  Yeah, Yeah. I think the biggest fork in the
• 32:35
  road that I've been keeping an eye on is whether or not things
• 32:38
  will become open source or closed, right?
• 32:41
  And so either one of those are so different, you know, and just
• 32:44
  in the past few months, we've seen the chatter be that America
• 32:48
  is becoming more closed. So, you know, is Meta going to
• 32:51
  release the next one is open source?
• 32:53
  Is Grok really open source anymore?
• 32:55
  And the Chinese models, I think for the top five open source
• 32:59
  models are now Chinese. And so if we continue to go open
• 33:03
  source, that's one path, very distributed and democratized.
• 33:08
  It was very different path. And then the other one is I, I
• 33:12
  think probably at the moment more scary to me, this idea of 1
• 33:16
  grand closed frontier model that, you know, it's like in
• 33:20
  Camasots. What was the name in Wrinkle in
• 33:22
  time? The, the, the the the bad
• 33:24
  computer, the computer, right. And I forget the details, but
• 33:27
  you know, I don't want to be camasots, right.
• 33:29
  Warren Grand seeing eye. So I think, I think the idea of
• 33:34
  it being in our lifetime is I don't know, I probably see I'm
• 33:37
  I'm not optimistic in my I'm 51, maybe for our kids, possibly.
• 33:43
  But the way I see it playing out is more in these incremental
• 33:46
  steps. So I remember the Apple Newton,
• 33:51
  which was a dud. Yep, I remember.
• 33:53
  But yeah, the Kindle is like in the iPad are successes.
• 33:56
  And so the things that came together.
• 33:58
  Before even the Newton, by the way, the rules they go.
• 34:01
  Which I didn't have. Which, yeah, which there was a
• 34:04
  great non fiction book written by the founder of that and about
• 34:09
  how it crashed and burned. But it was a really beautiful,
• 34:12
  like entrepreneurial read because it took you through the
• 34:15
  whole thing where you had the cardboard box that you threw on
• 34:19
  the table. I think Kleiner Perkins backed
• 34:21
  it. And like, you know, imagine
• 34:23
  this. And so there were there were
• 34:24
  there were other predecessors outside of Newton.
• 34:27
  Well, for me, the, the Newton was a big lesson in the sum and
• 34:31
  the parts. I think there's an analogy to
• 34:32
  that with AI. So with the Newton, we didn't
• 34:34
  have digital rights management, We didn't really have Wi-Fi.
• 34:37
  I don't think when it first came out you had to like plug your.
• 34:39
  I don't think so. Yeah.
• 34:40
  So you're having an infrastructure, DRM,
• 34:43
  distribution, content syndication, all that sort of
• 34:46
  stuff. Once those parts came together
• 34:49
  into the sum, now we have a product and with AII think, you
• 34:53
  know, we have medical advances, which will be fantastic.
• 34:56
  So just on their own, the ability to read an X-ray better
• 34:59
  will be wonderful, or the ability to predict the flood or
• 35:04
  you know, the weather. But I think currently these
• 35:08
  these best in class products are from different companies, right?
• 35:14
  And anytime I see that company competitive competition, I don't
• 35:17
  necessarily see anything utopian because that's sort of resource
• 35:21
  scarcity as opposed to this idea of a non resource dependent
• 35:26
  humanity, right? So if we if this, the idea of
• 35:29
  capitalism is so rooted into the model right now, right, I don't
• 35:33
  see an exit yet? Can I, can I push on that a
• 35:36
  little bit? Yeah, because for me, I was
• 35:40
  trying to figure that out. And I have concluded in in my
• 35:46
  own mind, this doesn't mean it's going to be true, but that open
• 35:49
  source is going to win. And the reason why is it was
• 35:53
  February 21st of this year that President Trump announced the
• 35:58
  Stargate project, which is 500 billion of funding.
• 36:01
  He announced that alongside my OSHA son, Sam Altman and Larry
• 36:05
  Ellison. And to put that into context,
• 36:09
  inflation adjusted, that is more money inflation adjusted than we
• 36:13
  spent on the space race versus the Russians or on the Manhattan
• 36:17
  Project versus the Nazis. And that amount of money was
• 36:24
  overshadowed the day before. So you can't even script this.
• 36:28
  February 20th is when DeepSeek announced the R1 model and it
• 36:33
  took until the next week for the market to realize that R1 was
• 36:41
  vastly vastly more powerful as an open first model then GPT 4
• 36:46
  was. Which if you go back in time, it
• 36:48
  was March of 2023 that GPT 4 came out and the moratorium call
• 36:53
  happened that 37,000 researchers in AI sign, including Elon Musk
• 36:58
  saying shut it down until we get a handle on AI safety.
• 37:03
  Like give us six months to figure that out.
• 37:05
  And of course, we're so far past the GPT 4 moment.
• 37:09
  And, and there's, there's multiple ways to look at that
• 37:13
  with the Chinese, like is this a provocation or is this an olive
• 37:17
  branch? I mean, I, you know, I don't
• 37:19
  know. I mean, I don't know what was
• 37:20
  going on in their mind, but it is open source and what it, what
• 37:25
  it told me in that moment is that open source will win
• 37:31
  because the weights can be either stolen, borrowed,
• 37:35
  whatever you say. So the weights from Open AI were
• 37:41
  taken by R1 by DeepSeek to create a model that, you know,
• 37:47
  you can't really believe what what, what the government says
• 37:50
  on this. But the Chinese government said
• 37:52
  it only cost, well, deep. See if the company said it only
• 37:55
  cost $16,000,000 to develop this thing.
• 37:59
  So that means that the frontier models definitely have a role.
• 38:02
  I mean, like Open AI is already at 12 billion of AR.
• 38:06
  OK, Two years, I think nine months now after the launch of
• 38:11
  GPT of ChatGPT. And there's never been a company
• 38:16
  that's grown faster than that in historysalesforce.com took 18
• 38:20
  years to get to 10 billion of AR ServiceNow took 21 years to get
• 38:24
  to 10 million of AR, right? ChatGPT got there in 2 1/2 years
• 38:28
  and thropic has grown I think 900 million of AR to 5 billion
• 38:33
  in the last 10 months. I mean, which like stunning the,
• 38:36
  the, the pace is stunning. So the frontier models are going
• 38:40
  to have a big role and they'll be out in front.
• 38:43
  But one of the big learnings for me at least in this AI arms
• 38:48
  race, which we've never seen anything like this in human
• 38:50
  history in terms of the amount of money and time, is that the
• 38:55
  weights can be borrowed. And therefore, like, you know,
• 38:58
  Yuzu, our youngest daughter, she's running 5 open source
• 39:01
  models on her home computer. And they are vastly more
• 39:05
  powerful than GPT 4 was. And her home computer is a
• 39:08
  pretty powerful computer, but it's not as powerful as one of
• 39:13
  the NVIDIA digits boxes, which can each run a 200 billion
• 39:17
  frameter model. And you can chain those
• 39:20
  together. So you could actually have five
• 39:21
  of those, you know, running a thousand 1000, you know, gig
• 39:27
  parameter or 1000 billion parameter model.
• 39:30
  So at least to me, that tells me that open source is going to win
• 39:37
  and that this really is going to be like electricity for the
• 39:39
  world and not just controlled in the hands of the few.
• 39:42
  So I think with open source, I would look at what we've learned
• 39:45
  from the Internet with the Balkanization, yes.
• 39:48
  And I do like the. Idea It has been vulcanized.
• 39:50
  I like the and that's a term I stole from somebody, but I like.
• 39:52
  It, but it's a great term. It's a good term.
• 39:54
  For you. But the, the, the I've always
• 39:57
  thought it's great to be able to fine tune your own model.
• 39:59
  Yeah. So if you have an open source
• 40:00
  model, I could make it G, you know, rated G friendly for kids.
• 40:04
  I could make it talk like a sailor or pirate.
• 40:06
  I could have, I could have whatever the raunchy model.
• 40:09
  But I, I can control it's politics, whatever I want.
• 40:12
  Right, Right. And so that to me is a blessing
• 40:14
  and a curse, because it seems very analogous to this.
• 40:18
  Falconization of the Internet where you have a vacuum or a
• 40:21
  bubble. So I think I like the open
• 40:24
  source piece. I always would assume there's a
• 40:26
  stronger model that the people don't have, whether it's
• 40:29
  frontier, just a better CPU or more, more compute, you know, So
• 40:34
  I think that would be something I'd love to hear your thoughts
• 40:36
  about. There's there's two pieces
• 40:39
  before I switch back to you. One is just will, will this
• 40:42
  technology help humans live happier lives?
• 40:44
  I think for sure when it comes to health, right, you know, or
• 40:49
  my ability to, you know, if I want to look up how to start a
• 40:52
  limited liability corporation, I can ask GPT 5 and it's going to
• 40:55
  tell me as well as a consultant at this point, right?
• 40:58
  So I think the the ability for anyone, just like the kid in
• 41:02
  rural Alabama can now watch an MIT physics lecture on YouTube,
• 41:05
  right? We all have a PhD in our pocket.
• 41:08
  That's a big deal for helping people have agency, right?
• 41:12
  I think the the the converse piece would be there's always a
• 41:17
  stronger model that the scare if your goal is to get rid of
• 41:20
  scarcity, you know, I wonder how even open source would
• 41:24
  democratize the the power, so to speak, because, you know, my I'm
• 41:28
  limited to whatever device I have, if any.
• 41:31
  Right, right. You know, no, it's, it's a very
• 41:32
  good question. And part of my thesis and the
• 41:35
  reason the book and the podcast, you know, are named Love
• 41:39
  Conquers Fear, is that the biggest challenge we have is
• 41:42
  ourselves. And what I mean by that is that
• 41:47
  because of the way we've evolved, we are locked into a
• 41:51
  bit of a scarcity fear based mindset, which sometimes tends
• 41:57
  to be a 0 sum. You know, there, there's this.
• 41:59
  The reason Stargate is getting that amount of funding, make no
• 42:03
  mistake about it, is that it's the same thing like the quantum
• 42:06
  computing race where we're like, if the Chinese get AGI before
• 42:11
  us, we lose. And the Chinese are like, if we
• 42:14
  get AGI, if we, the US, get AGI before them, they lose.
• 42:20
  And you know, we're a world of 8.2 billion humans.
• 42:24
  And at one point collectively, we have to wake up.
• 42:28
  And there are moments in time where we did wake up.
• 42:32
  You know, we woke up when it came to nuclear weapons, got
• 42:34
  really close with the Cuban Missile Crisis, a little bit too
• 42:37
  close for comfort. Could have been all over them.
• 42:39
  You know, it's one person standing in the way of that it
• 42:44
  it got close to a smallpox and there there was a point in human
• 42:49
  history that most people don't talk about or learn about 50 to
• 42:54
  70,000 years ago where we were down to anywhere from 800 to a
• 42:58
  couple of thousand humans. There's no way to know for sure,
• 43:01
  but the genealogical records show that we almost were
• 43:05
  extinguished during the Ice Age. And the theory is that a super
• 43:11
  volcano came on top of the Ice Age and blanketed the earth.
• 43:15
  And it was, it was such tough conditions that, that that could
• 43:20
  have been the end of the human experiment, right?
• 43:22
  The that could have been it. And instead we pulled together
• 43:26
  and collaborated and survived. And so that the part of what
• 43:30
  gives me faith in humanity is that our core nature is good,
• 43:36
  but there's lots of things constantly challenging that core
• 43:38
  nature. I mean, I'm very clear eyed on
• 43:40
  social medias and algorithmic manipulation, like trying to
• 43:44
  goal, seek to our bases fears enragement equals engagement
• 43:48
  equals selling more ads. You know, Meta pulled in $45
• 43:51
  billion of profit last quarter. You know, Mark Zuckerberg has
• 43:55
  announced he is spending hundreds of billions of dollars
• 43:59
  on a Manhattan size data center dedicated, but at least he's
• 44:03
  doing it in open source. At least Llama is open source
• 44:07
  and it's a challenger to to the models.
• 44:11
  And so so, you know, it's a yin and Yang thing, right.
• 44:14
  And then of course, the the media itself, which I'm a big
• 44:17
  fan of the Free Press like this, not a critique of the Free
• 44:19
  Press, but it's just a critique of the business model is also
• 44:23
  goal seeking to our most primitive fears.
• 44:28
  And that's why they're not going to push to you and me the latest
• 44:33
  climate change cure, possible cure, or latest possible cancer
• 44:39
  cure. They are going to push to us the
• 44:41
  latest crazy thing that some politicians said that we should
• 44:45
  all be pissed off about. And so if you get programmed all
• 44:49
  the time, you just assume it's a dystopia.
• 44:52
  And the reality is, in a world of 8.2 billion humans, there's
• 44:55
  always something bad happening every day.
• 44:59
  But there's also constantly many, many, many more acts of
• 45:04
  love and beauty and connection. You know, like what we're doing
• 45:08
  here happening every single day. Well, you said two things that
• 45:11
  reminded me. So the Bay of Pigs in the Cuban
• 45:14
  Missile Crisis was a big, big informative moment for me when I
• 45:18
  watched. I I wish I knew who was on the
• 45:20
  panel, but it was the the main players.
• 45:23
  We're all invited back at some anniversary.
• 45:25
  So I don't know if it's 20 years later or what.
• 45:27
  It was Apbs panel, OK, from all the people who are involved,
• 45:32
  right? And they all sat at a table and
• 45:35
  confessed that they all were scared of the other person and
• 45:39
  just didn't want to make a mistake because they thought
• 45:42
  they were going to lose. Right.
• 45:43
  So it's the ultimate, you know, prisoner's dilemma or whatever
• 45:47
  you want to call it, game prisoner's dilemma.
• 45:49
  Yeah, it is. Yeah.
• 45:50
  Because when they got to confess later, it was like, yeah, I had
• 45:52
  no desire to hurt you. I didn't even know you.
• 45:54
  I didn't want to. I didn't.
• 45:55
  I had no I'll will to you. I thought you hated me and it
• 45:59
  was as a kid. I don't know how old I was when
• 46:01
  I saw this, but it was part of my journey as an individual that
• 46:05
  when I saw that I thought, wow, that's worth remembering.
• 46:09
  Because why can't we just hit pause and have everyone come
• 46:12
  back 20 years later? Like when everything, because
• 46:15
  you know, we go back to Japan now, we go to Vietnam now you
• 46:18
  always go back as a tourist to Germany and everyone's friends.
• 46:22
  Why can't we just hit the pause button and skip to the friends?
• 46:26
  Part And what would it take for us to skip to the friends part?
• 46:29
  I mean, one thing I do answer because you asked me this
• 46:32
  question. So I don't want to, I don't want
• 46:34
  to be perceived as dodging it because I'm not dodging at all.
• 46:37
  Like you asked me about the energy problem.
• 46:40
  And part of my thesis in terms of us getting to the age of
• 46:43
  abundance for all is that similar to deep mind with with
• 46:48
  alpha fold and now alpha fold 2, solving the protein folding
• 46:52
  problem. And they've come out with other
• 46:56
  versions of alpha this and alpha that, which are really
• 46:59
  incredible. We are going to have a dramatic
• 47:02
  acceleration in material science.
• 47:04
  And I think that the energy that we need is constantly staring at
• 47:09
  us in the sky every every waking hour.
• 47:13
  And that, of course, is the sun. And so, you know, we just need
• 47:16
  the material science breakthrough where solar is
• 47:20
  indisputably the champion. And actually, if you look at the
• 47:24
  charts, it's already happening. And so, you know, you being a
• 47:28
  long time technologist, you saw the beginning of crappy modems
• 47:31
  and crappy printers and crappy computers and crappy bandwidth,
• 47:34
  all that stuff. You can kind of just see that
• 47:38
  we're going to have the world of unlimited, nearly unlimited
• 47:42
  energy from the sun. One thing that I, I, you also
• 47:45
  made me think of is, you know what I believe is one thing.
• 47:49
  There's also what the market believes and a lot of things
• 47:52
  like that. So going back to my naive young
• 47:54
  self at Burpee when I just walked in with an idealistic
• 47:57
  principle, 1 theme I've noticed is that businesses protect
• 48:00
  themselves and they only invest and people do as well that the
• 48:04
  market just that's true. You want to do what's in
• 48:07
  countries, what's going to be fruit, bring fruit.
• 48:10
  And so I wrote in my most recent newsletter, one of the most
• 48:13
  surreal things I'm witnessing is the conversions of the hyper,
• 48:20
  the hyperbolic emotional things people are saying with the
• 48:24
  investment. At the same time.
• 48:25
  So, you know, Meta got a lot of press for offering 100,000,200
• 48:30
  million pay packages to try to poach Open AI and other other
• 48:34
  employees. But what struck me is possibly
• 48:37
  even more powerful is I, I believe the rumor is 10 Open AI
• 48:43
  employees turned down $100 million.
• 48:46
  Right, right, right. So when when someone's willing
• 48:48
  to fund $500 billion in a data center, that's different than me
• 48:53
  at a bar with two beers being like, hey, this is going to
• 48:55
  change the world if someone's giving real money.
• 48:59
  Yes, real money. And, and you know, it may seem
• 49:01
  completely insane to someone watching this that that Meta
• 49:07
  would offer an AI engineer $100 million.
• 49:11
  But one of the things they're missing in that is that if
• 49:14
  you're spending 10s of billions on data centers and you can
• 49:18
  increase the efficiency, some of these engineers can increase the
• 49:20
  efficiency 3X. That is serious, serious
• 49:24
  leverage. I don't.
• 49:25
  It all comes down to the end of office space where you save the
• 49:28
  pennies, right? It's just a bigger, Oh my gosh,
• 49:31
  it's just at scale. It's all a space at scale.
• 49:33
  You know, that movie was made in Austin.
• 49:36
  It was, it was one of the original.
• 49:38
  So, you know, going back to, to, to your AI newsletter because I
• 49:43
  want to make sure people really understand, you know, you're by
• 49:48
  far and away like one of my favorite follows on LinkedIn.
• 49:53
  It's you for me, Ethan Malik, who is at the Wharton School and
• 50:00
  was actually the keynote at my 25th anniversary last year.
• 50:05
  Our youngest daughter, like literally went up to Ethan Malik
• 50:09
  afterwards. And it's like, can I just skip
• 50:10
  this high school stuff and just go straight here cuz.
• 50:13
  And you know, she's right actually to think that way
• 50:16
  because of how fast this exponential curve is rising.
• 50:22
  But anyways, that didn't happen. She's, she's in, she's in high
• 50:25
  school right now, but she's spending a lot of time on AI
• 50:28
  and, and robotics. So, you know, and the other is
• 50:32
  actually Morgan Linton, who I'm going to have him on the podcast
• 50:36
  to. He's the Co founder of this
• 50:38
  company Bold Metrics. So Deborah and I have been
• 50:40
  investors in since the beginning, but he takes much
• 50:42
  more AI engineering approach. So like just take us back to the
• 50:47
  beginning of this newsletter because your staying power here
• 50:51
  is like incredible. I think you're it's 35,000 links
• 50:55
  and counting. I mean, I don't know another
• 50:57
  human being on Earth that has chronicled it to that extent and
• 51:05
  just take us back to the beginning of that.
• 51:06
  So I think, you know, you have to play to your strengths.
• 51:09
  All right. I used to be very self-conscious
• 51:11
  that I was a little much, but I'm now leaning into it.
• 51:14
  Good, good. You know, all it takes is doing
• 51:16
  something consistently to do pretty great things.
• 51:20
  And so I, I, I've always been an extreme person.
• 51:23
  And so whether, you know, unfortunately, with toxic ideas
• 51:26
  like drinking, that was a challenge because doesn't matter
• 51:28
  what's in the fridge, I'm drinking 6 of it.
• 51:30
  So, you know, that that idea was a bad piece of my obsessiveness,
• 51:34
  but one when I was, I don't know if you remember a website called
• 51:37
  Delicious. Oh, yeah, I do the bookmarking
• 51:40
  site. Yeah.
• 51:40
  Yeah, I was. I've forgotten about.
• 51:42
  That I was, I've always been a I'm almost like a strange
• 51:45
  hoarder of, of information. Yeah.
• 51:47
  So back in the day, I don't know if this was the 90s still or
• 51:50
  perhaps early 2000s, but Delicious was the place you
• 51:53
  would go to bookmark stuff. Yeah.
• 51:55
  And this is back during XML and RSS feeds and when before these
• 52:00
  walled gardens like Facebook, you could aggregate and compile
• 52:04
  information and then share it. Yeah.
• 52:06
  And and you know, Dave Weiner was a big influence with OPML
• 52:09
  for me and. And RSSXML.
• 52:12
  Yeah, in the I. Think he was one of the authors
• 52:13
  of the Kluge and. Benefit.
• 52:14
  Yeah, I'm not. Yeah.
• 52:16
  So, so, so, you know, I did not realize it at the time, but just
• 52:20
  by the nature of being myself, one day I got a, a message from
• 52:24
  the head of delicious and, and he was like, you know, you, you
• 52:27
  are insane. You've got more links than
• 52:30
  anybody on this platform. I don't know how you physically
• 52:33
  do it. You clearly love the Internet in
• 52:35
  your free time and you have thousands and thousands of
• 52:38
  links. And so he brought me into Yahoo
• 52:40
  because Yahoo acquired them right right around the same time
• 52:43
  as Flickr. So I got to see Katarina fake.
• 52:45
  And I mean, this is like for me, I might as well be meeting like
• 52:48
  a football star, right? So going to Yahoo's headquarters
• 52:51
  were really exciting and I was the top user of Delicious.
• 52:54
  And so I realized, OK, I'm a little obsessive.
• 52:57
  So this is what I tapped into when it came to AI.
• 53:00
  So when when AI, when I saw GPT, I just decided I am going to
• 53:05
  without even realizing the symmetry of my delicious
• 53:08
  behavior, I started organizing links for myself.
• 53:12
  And so there's this has never been to publish a blog.
• 53:15
  I don't care if anybody ever reads it.
• 53:17
  It's just I want to know more than anyone else in the room.
• 53:20
  I mean, it might be the smartest in the room, but if you bring up
• 53:23
  retrieval, augmented generation or cohere or embodiment,
• 53:27
  whatever it is, I don't know what we're talking about.
• 53:30
  And so starting right after GPT came out in December, I started
• 53:35
  just manually reading everything I could and then organizing it
• 53:40
  privately. And then in September, which is
• 53:44
  several months later, I gave a speech to the National
• 53:47
  Association of Broadcasters. And that was when I decided I'm
• 53:52
  going to be the weirdo in the room because everyone thinks of
• 53:56
  AI even now is like a language bot.
• 53:59
  But I was like, no, there's this thing called multi modality
• 54:03
  where it's going to be able to also think and talk in sound
• 54:06
  waves motively or possibly, you know, through through
• 54:10
  understanding of pictures and context of what's in a room and
• 54:13
  be able to understand things. And then long term these things,
• 54:17
  just like the Apple Newton was too early.
• 54:19
  We're at the Newton stage, but there's going to be a Kindle and
• 54:22
  an iPad where all these parts come together and you're going
• 54:26
  to get robots that you can talk to because the natural language
• 54:30
  is an interface. It's not the, the language is
• 54:33
  not the product. The language is the first step.
• 54:35
  So it's a chit chat bot that can then, if you give an agency, can
• 54:39
  do really cool things. So I, I decided at that
• 54:42
  presentation with the National Association of Broadcasters that
• 54:46
  that would be my opportunity to publish something with a road
• 54:50
  map. And I was like, well, I got to
• 54:52
  put it somewhere. And then I realized kind of
• 54:54
  like, you know, there's the joke pictures or it didn't happen.
• 54:57
  I'm like, you know, I don't want to just be singing in my shower
• 55:00
  at home about AI you, you have to be braver than that, right?
• 55:03
  Right. I bought a domain.
• 55:06
  I got a wordress account and it was taking me more hours than
• 55:11
  there are in the week to publish each week, right?
• 55:13
  So I think it was like 80 hours. Right.
• 55:16
  Korea problem, Right? Right.
• 55:17
  So I ended up being backlogged by, I think I was four months
• 55:20
  behind, right? And I really wanted to quit.
• 55:22
  I were embarrassed because here I am, like, here's the latest
• 55:25
  about AI, Brett, right? It's four months.
• 55:27
  All four. That's all right.
• 55:28
  But I was doing it because I'm obsessive.
• 55:30
  I was doing it in seven day blocks.
• 55:33
  And I was like, look, if I just chip away at this, I'll have an
• 55:36
  archive of every week in like a history of this moment.
• 55:42
  And so from there I was like, well, this is a good opportunity
• 55:45
  to force myself. If everyone says AI is a
• 55:49
  productivity booster, maybe I should use it.
• 55:52
  But I went from using Word docs and Google Docs and copy and
• 55:56
  pasting. I actually got carpal tunnel.
• 55:58
  I used, I used AI to write Python scripts and I used them
• 56:02
  to make I made a Chrome extension, which was probably
• 56:05
  the coolest moment for me when I used to read, I think it was 600
• 56:09
  clicks of my finger to open all these tweets and then copy and
• 56:13
  paste their their title and their description in the URL
• 56:16
  into my blog. And I made AI made a Chrome
• 56:19
  extension that I could just hit the button.
• 56:21
  I would just suck them all up and then I had to read them all.
• 56:25
  But at least I wasn't copy and pasting 700 times and getting
• 56:28
  repetitive stress injuries, right.
• 56:30
  But at the end of the day, you know, I just, I just kept doing
• 56:33
  it and kept doing it right, Like Forrest Gump, right.
• 56:36
  I've always been AI like to run far.
• 56:38
  I used to run like 35 miles. Yeah.
• 56:40
  And you don't realize what you're doing until it's past.
• 56:43
  So one day I looked and I was like, wow, I'm in.
• 56:45
  I'm in like 20,000 links that I've hand organized.
• 56:48
  Yeah. And now I think I'm just shy of
• 56:50
  36,000 hand organized looks. And I use a CSV now.
• 56:55
  So I just have a CSV with 45 categories and I take all the
• 56:59
  headlines and I do still click on them all and I put them into
• 57:02
  the categories. And it's everything from
• 57:03
  retrieval augmented generation to Quinn to could be embodied
• 57:08
  robotics or XAI or autonomous vehicles.
• 57:12
  And then I have a Python script, thankfully, that takes that CSV
• 57:16
  and just publishes it for me. Wow.
• 57:17
  Like I've been doing that now. It's been 96 weeks.
• 57:22
  Wow, that's incredible. And, and I think the thing I
• 57:25
  would say for the, the the viewers is I'm always been
• 57:27
  self-conscious that I'm an English major.
• 57:30
  I'm not a technologist, I don't code.
• 57:33
  And thankfully I've got friends who are coders who gave me the
• 57:37
  confidence boost to say no. You have a really cool Spidey
• 57:40
  sense now because you're you're wide.
• 57:42
  Not everyone. Most people are specialists and
• 57:45
  by having a wide view it gives me a more holistic idea of
• 57:49
  what's happening. Well, actually it's a it's a
• 57:52
  great example of the fact that you have the power of some of
• 57:59
  the best coders now with AI, right?
• 58:03
  Cheating a little bit, but yes, I can.
• 58:05
  I know. That's cheating.
• 58:06
  It's it's actually been the arc of all human history, starting
• 58:10
  with the discovery of fire, that we should use technology to
• 58:13
  create easier and more successful lives.
• 58:16
  I should, I should mention because you mentioned I'm a good
• 58:18
  dad, I, I do this sometimes life works out.
• 58:23
  So both of my daughters are competitive dancers and they
• 58:25
  dance every day, six days a week until about 9:00 or 10:00 PM.
• 58:29
  So I'm a bachelor and my wife is with them at the dance studio.
• 58:32
  So I'm alone all day. Anyway, So after work I I work
• 58:36
  out, I go to the gym, try to not become a couch potato, and then
• 58:39
  every time I'm doing. This Yeah, OK, so with the AI
• 58:43
  newsletter, there's very few human beings on earth that I
• 58:48
  could ask this question of that had like really charted it since
• 58:52
  the beginning of ChatGPT, you know, almost 36,000 links,
• 58:58
  right? And and, and, and also, you
• 59:00
  know, you are extremely intelligent.
• 59:02
  I've always thought that about you.
• 59:05
  And So what, what are the, you know, 5 to 10 biggest standout
• 59:10
  moments where you're like, OK, this really changes everything.
• 59:14
  So the first one that is pretty popular would be agents, and
• 59:18
  that was something that I felt intuitively was coming and it
• 59:22
  made me very happy to know I was on the right track when agents
• 59:25
  became the buzzword. So agents for sure, I don't want
• 59:28
  to dismiss the obvious because that's great, but that is really
• 59:31
  where the language model is no more.
• 59:33
  No longer riffing or hallucinating.
• 59:36
  It's got chain of thought, it has reasoning and you can ask it
• 59:39
  to do it, do a task and then it will it will go do something for
• 59:42
  you and bring it back. So it really was the beginning
• 59:45
  of the language model as translator.
• 59:48
  So I could ask it something in plain English and then it would
• 59:50
  go execute and come back as opposed to in the context window
• 59:54
  just predicting the next token. So that agency was a huge,
• 59:59
  that's a big difference than ABS artist.
• 1:00:02
  That's just predicting tokens. Agency was a big one.
• 1:00:05
  And then. And we're still at the very
• 1:00:08
  beginning of agent, totally beginning, very beginning.
• 1:00:11
  I mean, it's, you know, Sequoia capital last year, it kind of
• 1:00:16
  went under the radar. But when the O1 model came out,
• 1:00:19
  that was when they put out a research report as Pat Grady and
• 1:00:23
  a few other people at Sequoia to put this out where they
• 1:00:29
  announced it in this white paper as a $10 trillion opportunity.
• 1:00:33
  Because people don't realize, like how big the software market
• 1:00:36
  is, but more importantly, how big the services market is.
• 1:00:40
  And now you actually see these Frontier models going into
• 1:00:43
  services. You know, Ethan Moloch, I think,
• 1:00:46
  who's one of my favorites, said something like, even if we
• 1:00:48
  stopped now, right, it's going to take a decade to even realize
• 1:00:52
  what we've done right. And so there's there's a there
• 1:00:54
  is one. I believe it's simple and I
• 1:00:56
  think they're get simple dot AI. Hope I didn't mess that up, but
• 1:01:00
  they're they're, I think Y Combinator, but they were they
• 1:01:04
  were an example of a turnkey plug and play agent for
• 1:01:08
  consumers to delegate phone calls to get a reservation or do
• 1:01:13
  mundane tasks like wait to call your hotel and basically take
• 1:01:17
  the man, stick it to the man with your own version of of you
• 1:01:20
  hold for me type thing. And so that that was a very
• 1:01:23
  interesting evolution. I think Yelp then came out with
• 1:01:26
  the version where small businesses can then reply.
• 1:01:29
  So the idea that, you know, my agent calls your agent, right
• 1:01:32
  is, is very powerful. Right, right.
• 1:01:34
  Then I think more so than agents, embodied robots has been
• 1:01:39
  the thing that consumes me. Your audience is probably more
• 1:01:43
  open to that topic than perhaps a layperson.
• 1:01:46
  But but that that is where I focus on is what's happening
• 1:01:49
  with embodied robots. And that's where the conversions
• 1:01:53
  of all these things that I've been seeing through my
• 1:01:55
  newsletter really start to run together.
• 1:01:58
  And that's been the most fun thing to watch.
• 1:02:01
  So before I learned about embodied robots, there was a
• 1:02:04
  guy, I believe his name is B Lawal, and he started talking
• 1:02:09
  about AR, which is a total pivot from from language models as you
• 1:02:13
  think of that being virtual. That's right.
• 1:02:15
  And so artificial, you know, or augmented reality.
• 1:02:18
  And in particular, he talked about neural radiance fields and
• 1:02:21
  something called Gaussian splatting.
• 1:02:23
  And I remember Gaussian filters as a kid with, with, you know,
• 1:02:26
  with, with the blurs in Photoshop if you're an old
• 1:02:29
  school Photoshop. Right, right.
• 1:02:30
  Yeah. And what became super clear to
• 1:02:32
  me was just how much everything is connected with all of these
• 1:02:37
  seemingly separate products. So for example, Gaussian, I
• 1:02:42
  think it's Gauss, I think his name is Gauss.
• 1:02:44
  But Gaussian blurs were something you could apply in
• 1:02:46
  Photoshop to add noise. So if you wanted to make it
• 1:02:50
  more, you know, more interesting, you could blur a
• 1:02:53
  photo. And when I learned about
• 1:02:55
  diffusion models, I found out that the way diffusion models
• 1:02:58
  were trained, this was denoising.
• 1:03:01
  So you, you would train a model on 1000 pictures of a cat, just
• 1:03:06
  just one thing. And then you'd add some noise
• 1:03:09
  and you'd say, what do you think this is?
• 1:03:10
  And it would denoise. So this very simple concept of
• 1:03:14
  denoising, they just kept adding noise.
• 1:03:17
  So you've got easy to easy to pick out a cat if it's 10%
• 1:03:20
  noise. Then they had 203040506070.
• 1:03:24
  And whoever thought of this, we got to Google this person.
• 1:03:27
  Somebody had that aha moment to say, what if we just give it all
• 1:03:30
  noise and say, what if this was a cat?
• 1:03:33
  And then you just denoise this random static and it's a cat.
• 1:03:37
  And then because of that, you could have more than one object
• 1:03:41
  that would never appear in real life, like a cat riding a shark
• 1:03:44
  on top of a tank. And you just give it random
• 1:03:47
  noise and it denoises it into this magical concept.
• 1:03:51
  And so demystifying what is pretty mysterious concept of
• 1:03:57
  image diffusion and then realizing it's Gaussian
• 1:04:00
  splatting and then realizing there are these, you know,
• 1:04:02
  sorry, that's not Gaussian splatting.
• 1:04:04
  But then there's these other things like Gaussian splatting
• 1:04:07
  where you're for for people who don't know, that's where you can
• 1:04:10
  stitch together images and then diffuse what's missing.
• 1:04:15
  And it's a very different technology, right?
• 1:04:17
  So that's why, as you can tell, I get excited.
• 1:04:22
  CJI has been around forever, right?
• 1:04:24
  So like replacing people's jobs with CGI orcs has already
• 1:04:28
  happened? So Lord of the rings, there's
• 1:04:30
  not a million extras CBI and now if it's diffused extra.
• 1:04:34
  You just invoked my favorite movies of all.
• 1:04:36
  Time, different types of ethics, right?
• 1:04:37
  I understand the ethics of it all, but taking taking the
• 1:04:41
  ethics aside and just thinking about how this stuff works, just
• 1:04:44
  like CGI is a different technology, Gaussian splatting
• 1:04:47
  is 1. There's also something, there's
• 1:04:50
  a very cool company called Live Portrait that I think it's from
• 1:04:52
  Byte Dance, and they were basically mapping behaviors on
• 1:04:57
  to stick figures. So basically you could face
• 1:05:00
  swap. So your face is now on me.
• 1:05:03
  And this idea of, I believe it's called latent consistency, where
• 1:05:07
  you take an image and then map something on top of it.
• 1:05:10
  That's AI, but it's very different than chat bots.
• 1:05:13
  Right, right. So to kind of recap, you know,
• 1:05:15
  hearing about how diffusion works and knowing that it's
• 1:05:19
  separate, but it's coming together.
• 1:05:21
  So you just ask your chat bot, Hey make me being a cat surfing
• 1:05:25
  on top of a shark riding a tank. I use plain English, no
• 1:05:29
  Photoshop. It then sends it to the
• 1:05:31
  diffusion model and the prompt turns into my image.
• 1:05:33
  Same thing then, where now I can swap myself to be you.
• 1:05:39
  And then all of a sudden you get a guy named Jim Fan.
• 1:05:43
  And he was the one who, he works at NVIDIA, and he's in charge of
• 1:05:47
  the robotics lab. And he started talking about
• 1:05:50
  training embodied robots in simulation.
• 1:05:53
  So now you're thinking, OK, artificial, you know, augmented
• 1:05:56
  reality, virtual reality, things that you don't necessarily think
• 1:05:59
  of as being applicable to the real world are being used as
• 1:06:03
  proxies for physics models. And Jim Fan, if you haven't seen
• 1:06:07
  it, you need to Google it. They trained a robot.
• 1:06:10
  And I learned about him from you and his and his talks are
• 1:06:13
  extraordinary. He's the man.
• 1:06:14
  If I had to say, like, you know, you have a few people in your
• 1:06:17
  brain. We're just like anything that
• 1:06:18
  guy says. All right, I'm I cosine sight
• 1:06:22
  unseen? Well, and NVIDIA is now worth
• 1:06:24
  over $4 trillion most valuable company in the world.
• 1:06:27
  So. So I'm sure that Jim Fan is like
• 1:06:29
  one of the key people at NVIDIA that makes the company worth
• 1:06:32
  done. He also, in the spirit of your
• 1:06:34
  podcast, strikes me as incredibly authentic and kind.
• 1:06:37
  A lot of agrees are congratulating his Co workers
• 1:06:40
  for their new job. He shares other people's
• 1:06:43
  research. He's a very authentic guy.
• 1:06:45
  Yeah, actually just as a quick aside, because you know, there's
• 1:06:48
  a lot of leaders that will watch this podcast.
• 1:06:53
  One of the key things you can tell on whether or not a leader
• 1:06:58
  is a truly like good person, like really cares about their
• 1:07:02
  employees is how much they talk about themself versus the rest
• 1:07:07
  of the team. And I mean, I am for sure a
• 1:07:11
  product of many, many, many people who chose to go on the
• 1:07:14
  journey with me, whether you as a customer back in the day or
• 1:07:18
  whether is, you know, my Co founders or people.
• 1:07:21
  And, you know, I, I, I really pay attention to that.
• 1:07:24
  And there's certain, you know, very high profile leaders in the
• 1:07:27
  world where they use I, I, I, I, I all the time.
• 1:07:30
  And it really shows you it's a window into how they think that
• 1:07:33
  they really think they're the one on the hero's journey and
• 1:07:37
  everybody else kind of it's like just there to support them.
• 1:07:40
  And it's just not the case. Like societies are created by
• 1:07:45
  people coming together. What to me is the Internet would
• 1:07:48
  say that. Companies are little societies.
• 1:07:50
  That's a red flag, as the Internet would.
• 1:07:52
  Say right, it is a red flag. Yeah, so Jim is all green flags.
• 1:07:55
  But the thing that really woke me up with Jim and embodied
• 1:07:58
  robots was he trained a dog in simulation to it was a robot
• 1:08:03
  dog, like one of those, you know, Boston Dynamics type dogs.
• 1:08:06
  And he trained it to walk on a yoga ball using only simulated
• 1:08:13
  training. And so he basically, I think in
• 1:08:15
  you have to, I hope no one quotes me if I'm wrong, but my
• 1:08:18
  impression was essentially, and if I do this right, you'll get
• 1:08:22
  that spark. He trained it in let's let's say
• 1:08:25
  20,000 simulations where each simulation at different physics.
• 1:08:30
  And that was one of the fun things he did, is he changed
• 1:08:32
  gravity, he changed the factors, whatever it is, just gave this
• 1:08:35
  dog. And the dog would train through
• 1:08:37
  reinforcement learning until they could walk on a ball.
• 1:08:40
  Yeah. And now, much like humans, where
• 1:08:42
  our brains are in our skulls and they've never seen the light of
• 1:08:44
  day, this dog is in a simulation, has never seen the
• 1:08:47
  real, real world. Right.
• 1:08:49
  And then one day he gave it the 20 thousandth and one
• 1:08:54
  simulation, which wasn't simulated, which wasn't, it was
• 1:08:57
  real life, right? And so we were just that plus
• 1:08:59
  one world. And the dog walked on a ball,
• 1:09:02
  right? Right.
• 1:09:03
  So when I saw that, I was like, OK, this is this is science.
• 1:09:05
  Fiction. And this actually is the bowl
• 1:09:08
  case for NVIDIA because you know that that that DeepSeek moment
• 1:09:13
  that I was mentioning earlier, like the week after President
• 1:09:16
  Trump announced NVIDIA lost 600 billion a market cap and the
• 1:09:20
  NASDAQ lost a trillion dollars a market cap because people
• 1:09:23
  thought this is kind of game over.
• 1:09:24
  You can just steal the weights. But the bull case for NVIDIA is
• 1:09:28
  that there's going to be NVIDIA chips in every robot.
• 1:09:32
  They've got the supply. I think they're they're not.
• 1:09:35
  I would wonder this is where the free market and the dynamics
• 1:09:38
  work because they have customers.
• 1:09:40
  You can't get rid of your customers, right?
• 1:09:41
  You need them for revenue too. But on the other hand, they're
• 1:09:44
  releasing open source models. I know they're incredibly open.
• 1:09:48
  And then conversely or in addition, you've got Amazon
• 1:09:51
  making models. That's right.
• 1:09:52
  And so that's been and. Actually Open AI is now
• 1:09:56
  releasing open source as well. So I think NVIDIA.
• 1:09:59
  Like finally. So yeah, that sort of thing.
• 1:10:01
  So is that you? You asked me just what I've
• 1:10:03
  noticed. I think, you know, diffusion,
• 1:10:05
  the idea of agents, the embodied robots, multi modality, context,
• 1:10:09
  understanding. There was a very famous example
• 1:10:11
  from 2023 or maybe early 2024 where they gave GPT actually, I
• 1:10:17
  believe, sorry, it was Google. They gave Google a picture and
• 1:10:21
  it's a famous picture of Obama in the locker room and he's got
• 1:10:24
  his foot on the scale and someone's weighing himself.
• 1:10:26
  And obviously Obama's messing with him because he's got his
• 1:10:28
  foot on the scale. And they asked the the AI what's
• 1:10:32
  happening in this picture? And they asked how many mirrors
• 1:10:35
  are in this photo? What is the man doing?
• 1:10:37
  And it was able to say the context, this man is putting his
• 1:10:41
  foot on the scale. Why is he doing that?
• 1:10:43
  He's trying to probably play a trick on this other guy.
• 1:10:45
  And so that idea, I believe, is called theory of mind, where you
• 1:10:49
  can understand emotional context in addition to just objects, the
• 1:10:55
  ability for computers. And again, I feel like if it
• 1:10:58
  wasn't your audience I would be shy to say this, but I believe
• 1:11:02
  it's been pretty clearly benchmarked that the the top of
• 1:11:06
  the line frontier models can beat most humans at things we
• 1:11:11
  think are exclusively human traits.
• 1:11:13
  Yeah, and that's that that is true and not is with less than a
• 1:11:19
  than 1000 days since the launch of ChatGPT.
• 1:11:22
  OK, so so people really need to mark this moment in time that
• 1:11:27
  humanity again has never prioritized anything with the
• 1:11:30
  amount of money intensity, you know, and and and just labor as
• 1:11:37
  this. And you know, that's coming back
• 1:11:39
  to Chad because I want to I want to come back to that for a
• 1:11:42
  moment to make this really practical for people in terms of
• 1:11:46
  all these mind blowing moments you've experienced and I've
• 1:11:48
  experienced in part through you thanks to thanks to your
• 1:11:52
  newsletter, which is a must follow, is that robots are the
• 1:11:59
  physical manifestation of AI and Chad.
• 1:12:03
  What they really need more than AI right now is they need
• 1:12:06
  robots. They need armies of robots to go
• 1:12:09
  in there and build the pipelining for the electrical
• 1:12:14
  supply they need for the clean water supply they need for the
• 1:12:16
  energy projects they need for the vaccines they need for the
• 1:12:20
  healthcare robots that'll help, you know, cure cleft palate and
• 1:12:25
  all of the other diseases that plagued a country like that,
• 1:12:28
  which largely is just not even thought about by Americans.
• 1:12:32
  I mean, when, when, when that Uber driver picked me up at 4:30
• 1:12:35
  in the morning on the way to MIT to give that talk that, you
• 1:12:39
  know, I shared with you about the hard problem of
• 1:12:43
  consciousness in the accelerating age of AI, quantum
• 1:12:46
  computing, robotics. Like I was kind of embarrassed.
• 1:12:49
  I knew nothing about his country.
• 1:12:50
  And so I was so and he was such an awesome guy.
• 1:12:53
  And so I just started immediately using Chachi PT to
• 1:12:56
  learn as we talked and I was asking him questions and you
• 1:13:00
  know, he he's saving up for the next six to seven years to move
• 1:13:04
  his family to America. But what a beautiful world it
• 1:13:08
  would be if Chad could actually be the Oasis that he actually is
• 1:13:13
  happy to be in. I mean I love America, but but
• 1:13:16
  what if Chad could also be the Oasis?
• 1:13:18
  So I have a question, sort of rhetorical question for me and
• 1:13:21
  you. Yeah.
• 1:13:22
  As far as the, the, the dual, there's a huge split.
• 1:13:25
  Yeah, because when I'm, you know, I'm in the weeds and I'm
• 1:13:28
  drinking the kool-aid. I'm reading.
• 1:13:30
  It's about 600 a week that I read now these headlines.
• 1:13:33
  All right, So you know, there's a great concept from Steven
• 1:13:36
  Johnson's book where good ideas come from.
• 1:13:38
  Oh yeah, I love Steven Johnson. The. 1st chapter is about the
• 1:13:42
  adjacent possible and so I can only bring you one concept at a
• 1:13:48
  time, right? So if you bring, if you talk
• 1:13:50
  about multi channel attribution and I don't know what that is, I
• 1:13:53
  can't, you got to tell me a lot and walk me there and bring me
• 1:13:58
  to you, right? And so with AI, because I'm so
• 1:14:01
  far into it, my adjacent possible, I feel alone most of
• 1:14:05
  the time, right? Because most of the people I
• 1:14:07
  talk to are not as into it. And so this past newsletter, I
• 1:14:12
  was reading a stat that said 2/3 of Americans have not even used
• 1:14:17
  it. And I think back to COVID in the
• 1:14:20
  conflicts with masks and vaccines.
• 1:14:22
  And, you know, even with AI, there are valid criticisms of
• 1:14:26
  the, the, the resource usage or the training data.
• 1:14:29
  And so often I have to split my brain into two pieces where on
• 1:14:33
  one hand I'm talking about like, look, I'm not saying it's right
• 1:14:35
  or wrong, but they can find cancer five years earlier.
• 1:14:38
  And it's hard to tell, you know, if I could get a mammogram that
• 1:14:42
  would know cancer five years early, I'm going to take it.
• 1:14:45
  So there's this tremendous pressure for things that very
• 1:14:49
  few people would say, well, I'm not going to use mammogram AI
• 1:14:52
  because it's unethical. I think most people would just
• 1:14:54
  say, yeah, I want to not die. Exactly.
• 1:14:57
  But then on the other hand, there's there's there's probably
• 1:15:00
  a large disconnect between early adopters, right?
• 1:15:04
  And I would I'd love your thoughts about do you think are
• 1:15:07
  people, you know, you see what happens with the Waymo vandals
• 1:15:09
  and different things, right? You know, is, is, would AI not
• 1:15:12
  only does humanity have a chance, but do you think we're
• 1:15:14
  just going to like kill it, you know, in rebel before it has a
• 1:15:18
  chance? Well, show what it can do.
• 1:15:20
  I think if we did, first of all, I think the answer is no, we're
• 1:15:24
  not going to just kill AI. But I think if we did, that
• 1:15:29
  would mean that fear one, and if fear wins, it's game over for
• 1:15:33
  humanity. Because if fear wins, we already
• 1:15:37
  like if you were an alien that came to planet Earth right now
• 1:15:40
  and you knew nothing about us at all, and you looked at the
• 1:15:46
  weaponry that we have with nuclear weapons alone, you would
• 1:15:50
  be like, OK, this species has created enough nuclear weapons
• 1:15:55
  to destroy the planet many times over.
• 1:15:58
  And by the way, they're still building them today.
• 1:16:01
  Like you would be like, OK, this is insanity, right?
• 1:16:04
  And you know, two things can be true at the same time.
• 1:16:07
  We could be both insane to do that and, or, and it's really
• 1:16:11
  just a handful of human beings that are making the decision to
• 1:16:13
  do that. I don't think 8.2 billion humans
• 1:16:15
  would collectively make the decision to do that.
• 1:16:17
  We're talking about maybe 1000 to 2000 human beings, which
• 1:16:21
  pretty much rule the rest of us. I mean, whether or not we want
• 1:16:25
  to call it rule, like in some countries, like Putin really is
• 1:16:28
  ruling, you know, Kim Jong Un really is ruling in our country.
• 1:16:33
  It is a democracy, but there's various aspects in a various
• 1:16:36
  opinions on that. So, you know, fear is the
• 1:16:43
  biggest problem. And, and, and, and one of the
• 1:16:45
  things I'm being very clear eyed in my book about is there's a
• 1:16:50
  book called The Alchemist, which I love.
• 1:16:52
  And there's a line in The Alchemist where it says it's
• 1:16:55
  always darkest before the dawn. And that line just, I can't tell
• 1:16:59
  you how many times that line rings through my head.
• 1:17:03
  And I do think for humanity that it will be dark and we'll be
• 1:17:10
  able to peer over the frontier almost like you're peering past
• 1:17:14
  the Earth on the on the edge of the curvature and you're seeing
• 1:17:18
  the moon and you're like the moon.
• 1:17:21
  That's, that's abundance, right? And that we'll be able to peer
• 1:17:26
  over that at some point collectively for humanity.
• 1:17:29
  And we have the free will to make the choice at that point.
• 1:17:33
  Do we want to do we want to become the species in the
• 1:17:39
  universe that allowed love to conquer fear?
• 1:17:42
  And then in an exponential curve that'll only get more
• 1:17:45
  exponential if we collectively wake up and collectively we've
• 1:17:49
  decided to really embrace the divinity in each other.
• 1:17:55
  The fact that we're almost genetically identical.
• 1:17:58
  I mean, you take a, you know, 7 foot tall black, you know,
• 1:18:02
  basketball player and you take a four foot tall Chinese monk and
• 1:18:07
  genetically they are almost exactly identical.
• 1:18:10
  I mean, it's, it's, it's, it's like 299.999 keeps going.
• 1:18:15
  And, and if you, if you take that collective wake up, then at
• 1:18:21
  that point, I think humanity is only bounded by the laws of
• 1:18:25
  physics, most of which we don't actually know.
• 1:18:27
  And quantum physics gets really weird really fast.
• 1:18:31
  And I actually think in that in that branch of the universe, if
• 1:18:36
  you will, that we will solve things like faster and
• 1:18:40
  lightspeed trouble. And, you know, part of the
• 1:18:43
  pushback I get it's kind of funny because like, even in a
• 1:18:46
  optimistic age of abundance for all scenario, people are like,
• 1:18:49
  oh, Brett, what are people are going to do on planet earth?
• 1:18:52
  Is it just going to be like people meditating, etcetera?
• 1:18:55
  And I'm like it. Look, there are going to be
• 1:18:57
  people on earth that measure the world in terms of gross domestic
• 1:19:02
  happiness as opposed to GDP, But there's going to be other people
• 1:19:06
  like me that are like super ambitious.
• 1:19:08
  And my sense of view is you're you're also ambitious.
• 1:19:11
  And like, if I can say, Ethan, hey, let's go and go check out
• 1:19:15
  this quasar or, you know, let's go to the edge of a black hole,
• 1:19:19
  not close enough to get sucked in.
• 1:19:20
  Or let's go to another planet, restart life on that planet or
• 1:19:25
  maybe start life on that planet for the first time.
• 1:19:28
  Those sound like really fun. I mean, that's the Star Trek
• 1:19:30
  feature, right? And and I think that we'll have
• 1:19:34
  that possibility because this, the exponential curve we're in.
• 1:19:40
  I mean, you really understand this, but it's far more
• 1:19:43
  exponential than people can grasp.
• 1:19:45
  The only way, I mean, I love your opinion on this.
• 1:19:47
  The only way I can kind of grok it is a week of tech
• 1:19:52
  development. To me, you know, part part of
• 1:19:55
  the partially like buffeted by your newsletter to me feels like
• 1:20:00
  a year of tech progress 10 years ago.
• 1:20:02
  Well, what's it going to feel like 10 years from now?
• 1:20:04
  Is an hour. Going to feel like a week today,
• 1:20:06
  which feels like a year 10 years ago.
• 1:20:09
  I mean that's that's that's what exponential means.
• 1:20:14
  One large harbinger of change that got my attention was I have
• 1:20:17
  a friend who's a world class programmer and he's in that edge
• 1:20:22
  case of people that AI was not good enough to help it.
• 1:20:25
  And he, he is does medium trade, medium frequency trading and
• 1:20:30
  stat ARB, you know, finance quant type stuff.
• 1:20:34
  And he, he just couldn't find a use case for AI given, given
• 1:20:38
  just what he was up to in, in his proprietary models and all
• 1:20:41
  these things. And, and one day, probably two
• 1:20:45
  months ago, three months ago, after Claude released their
• 1:20:47
  latest version in the command line, he texted me and said, OK,
• 1:20:51
  I think he had some cursing in there.
• 1:20:53
  It was like, holy crap, this is it.
• 1:20:57
  And his productivity in the command line, he's like, I can't
• 1:21:00
  stop. I'm so addicted.
• 1:21:01
  I'm just running. I've got multiple computers,
• 1:21:04
  multiple servers. They're able to do PhD level
• 1:21:06
  work, like legit stat PhD level work for him.
• 1:21:11
  Again, the thing that always freaks me out is when the money
• 1:21:14
  lines up with the passion, because this is not a guy who
• 1:21:16
  would just do this for fun like me.
• 1:21:18
  Like, hey, I made a Chrome extension.
• 1:21:20
  He's like, no, I'm making a lot of money faster and easier than
• 1:21:23
  I would with a PhD. And so that was a huge wake up
• 1:21:27
  call. I I don't know if did you ever
• 1:21:31
  read Why the Future Doesn't Need Us by Bill Joy from. 20 I did.
• 1:21:35
  That was a very dystopian article, but yes, I read it,
• 1:21:38
  yeah. That that was one of the.
• 1:21:40
  I ran the circuits that that was kind of like viral before.
• 1:21:42
  Viral weird the crap out of me, especially for those who didn't
• 1:21:45
  read it. Why?
• 1:21:46
  As a spoiler alert, if you want to read it, stop.
• 1:21:48
  Yeah, but. Yeah, why in the future Doesn't
• 1:21:50
  Need Us by Bill Joy. Why This Future Doesn't Need Us
• 1:21:53
  by Bill Joy. He tricks you and you read the
• 1:21:56
  first introduction and it's so amazing.
• 1:21:59
  And then he says that was the Unibomber manifesto.
• 1:22:02
  Oh no. So I remember reading this thing
• 1:22:06
  thinking oh I. Forgot about South bars.
• 1:22:08
  Yeah. And then you're like, holy crap,
• 1:22:09
  is the Unibomber. Right?
• 1:22:10
  Right. But one of the things he did not
• 1:22:12
  say explicitly, but it's something I picked up from, and
• 1:22:16
  I use this as a rhetorical challenge of how do you get off
• 1:22:18
  the train, Right. And this is going to come back
• 1:22:20
  to fear versus love. Try to buy an orange without
• 1:22:24
  scanning it into a computer. Yeah, like everything has a code
• 1:22:27
  on it. Like it's hard.
• 1:22:29
  You know, you can use cash maybe in Delmarva, there's some
• 1:22:31
  places. But if you try to buy an orange,
• 1:22:32
  most places beep. You've told the, you've told the
• 1:22:36
  the POS system to buy another orange.
• 1:22:38
  And I go to work and there's a, there's a good, you know, we
• 1:22:43
  basically commute to our computers to do data entry.
• 1:22:47
  And so Max Tegmark, who's AI think he's Swedish, she's a big
• 1:22:50
  AI researcher. He was talking about how hard
• 1:22:53
  it's going to be to pause. Yeah, he's written a number of
• 1:22:55
  books and given lots of Ted talks.
• 1:22:57
  And that that concept to me is sort of where you can't stop
• 1:23:01
  because, you know, whether you want to like it or not, the
• 1:23:03
  train's going right. And that goes back to Bill Joy,
• 1:23:06
  which is you can't get off this train, right?
• 1:23:08
  Because if I don't do it, you will.
• 1:23:09
  If, if, if Sam does doesn't do Stargate, China will, if, if you
• 1:23:13
  know so. So humanity in, in itself going
• 1:23:15
  back to the Bay of Pigs, it is, it is a prisoner's dilemma and
• 1:23:19
  it is going to happen. Someone's going to build the
• 1:23:20
  data center. Where is Elon or whoever you're
• 1:23:22
  rooting for at the moment? And so the best bet you can do,
• 1:23:26
  and this is what motivates me to know as much as humanly possible
• 1:23:30
  is I never think of myself as an AI evangelist.
• 1:23:34
  I want to, I want to make it. And so if you ask me what am I
• 1:23:37
  doing to make sure love wins, is I want to be the most informed
• 1:23:41
  person at all times so that when the opportunities come, I can
• 1:23:46
  explain the best I can to someone who hasn't had, right,
• 1:23:50
  25 hours a week spare time, right?
• 1:23:52
  And try to be aware of this idea that the only choice is to be
• 1:23:57
  optimistic and have love because you can't stop the train.
• 1:24:00
  It isn't like the trolley problem, right?
• 1:24:02
  Right. You know.
• 1:24:03
  And that that is in fact what you are doing to outside of the
• 1:24:07
  newsletter. You are speaking on this and I'm
• 1:24:12
  quite sure you're one of the best speakers in the world on
• 1:24:14
  this right now because because again, how many people have that
• 1:24:16
  type of knowledge that they've done that amount of Labor to
• 1:24:22
  really chart the progress and then also have the brain you do
• 1:24:25
  going back to the beginning of the Internet revolution well.
• 1:24:28
  It's nice to have a quantitative metric like 36,000 links because
• 1:24:32
  everyone's an AI expert right now.
• 1:24:34
  And that's like my fear. I don't like posting.
• 1:24:36
  When I posted LinkedIn and think I'm like Oh my gosh.
• 1:24:38
  And actually I would. Sit and agree with you.
• 1:24:40
  I think very few you are AI experts right now.
• 1:24:42
  I think a lot of people want to be an AI expert but would not
• 1:24:46
  put in the amount of work. So I figured you can't shortcut
• 1:24:49
  the work. You can't short it's, that's
• 1:24:51
  true of everything. I mean, that's true of if you're
• 1:24:53
  trying to change your physical body or you're trying to change
• 1:24:55
  your diet or you're trying to actually change, you know,
• 1:24:59
  something like a relationship. I mean, you've got to put in the
• 1:25:02
  work it takes. It takes a lot of reps to make
• 1:25:06
  yourself better at anything. And that's very much true about
• 1:25:09
  entrepreneurship. I've always said
• 1:25:11
  entrepreneurship is, is an art of grit and luck.
• 1:25:16
  And I always say, look on the grit side, There's a lot of
• 1:25:18
  people in America working multiple manual labor jobs.
• 1:25:21
  So it's not just grit. You know, luck actually does
• 1:25:25
  play a role, you know, being at the right place in the right
• 1:25:27
  time, being able to forecast the future or manifest that future
• 1:25:32
  and then you're on the right curve.
• 1:25:33
  I mean, that's really, that's really the reason I'm actually
• 1:25:37
  here as an entrepreneur is that I've been on the right part of
• 1:25:40
  that curve. Sometimes been a little bit too
• 1:25:42
  early, sometimes I've been a little bit too late, but I've
• 1:25:46
  always been somewhere on the right part of that curve.
• 1:25:49
  I think that the the moral of the story, no matter what topic
• 1:25:52
  it is, is you should never be embarrassed to be obsessed.
• 1:25:55
  Yeah. And if you want to be, if you
• 1:25:57
  want to be the best you can be, you have to be a little weird.
• 1:26:00
  And I tucked my personality in my entire life.
• 1:26:03
  And so this is the chapter where I'm going to let it fly.
• 1:26:06
  Yeah, yeah, yeah. Awesome.
• 1:26:08
  Well, it's good to have you out there.
• 1:26:11
  I mean, you're adding a lot of light and love into this world.
• 1:26:14
  Well, thank you for including me on your your your faith in
• 1:26:18
  people. Yeah, absolutely.
• 1:26:20
  Well, cool. This is this is a lot of fun.
• 1:26:22
  I really, really appreciate our friendship.
• 1:26:25
  And it's it's one I I really cherish.
• 1:26:28
  And, you know, maybe we'll do this again in the future and get
• 1:26:31
  into other topics. I mean, I'm pretty sure, you
• 1:26:34
  know, a year from now it'll be, I don't know, 50,000 links or
• 1:26:38
  something counting. I mean, it's only getting more
• 1:26:40
  exponential, so it's going to be harder to keep up.
• 1:26:43
  But you have the machine. Andrews Ironically, my my weekly
• 1:26:45
  number is going up, Yeah. And you do have the machines to
• 1:26:48
  help you help you keep up. Well, I hope everyone in your
• 1:26:50
  audience can find a brat because it's really special.
• 1:26:52
  I appreciate. It thanks man.
• 1:26:54
  I appreciate it. Awesome buddy.
• 1:26:56
  Well, thanks everybody for tuning in to Love Conquers Fear.
• 1:26:59
  I absolutely loved this episode. Ethan and I are about to go out
• 1:27:04
  on this beautiful boat and this incredible weather out here.
• 1:27:08
  Feel really lucky to be with him here and to have a friend that
• 1:27:11
  that has meant a lot to me for, gosh, decades now, and someone
• 1:27:16
  that can really jam on all these topics with.
• 1:27:20
  Stay tuned for more episodes. I have a number of just
• 1:27:24
  incredible investors, entrepreneurs, scientists,
• 1:27:28
  spiritual teachers lined up. And so it's going to be, it's
• 1:27:32
  going to be a blast. And I'm loving this and I hope
• 1:27:34
  you're loving this. And we'll learn together and
• 1:27:37
  hopefully inspire each other. Thanks again everybody.