[00:01.640 --> 00:02.200]  Okay.
[00:02.200 --> 00:02.900]  All right.
[00:03.760 --> 00:04.860]  Hello, everybody.
[00:05.180 --> 00:06.420]  How is everyone doing?
[00:07.140 --> 00:08.780]  Is everyone excited here?
[00:08.960 --> 00:10.060]  This is going to be a great talk.
[00:10.200 --> 00:25.980]  This is very exciting for me to introduce to you what I think is going to be a wonderful and very interesting and entertaining speaker with some very incisive comments on our culture and our ethos, I'm sure, which will be following these comments.
[00:27.920 --> 00:29.540]  So it is my great pleasure.
[00:29.760 --> 00:31.180]  I'm Alex Urbellis, by the way.
[00:31.320 --> 00:32.380]  I'm with Off the Hook.
[00:32.460 --> 00:33.340]  I was on the last panel.
[00:33.480 --> 00:36.160]  I'm a lawyer and various other things.
[00:36.360 --> 00:45.860]  I should probably introduce myself for a moment, but now I can introduce to you your wonderful speaker here, Harper Reed, who is, I think, a man with many, many talents.
[00:46.060 --> 00:47.580]  He creates rather interesting things.
[00:47.800 --> 01:01.580]  He wanted me to mention that first and foremost, the most important aspect of his bio is that he's happy to be here, which I think speaks volumes about his personality, but I guess the most interesting aspect of his background.
[01:01.740 --> 01:04.140]  And his background, I think, will also be part of the presentation.
[01:04.140 --> 01:11.260]  So I'm not going to go too far into the very interesting and rich and varied and complex life this man next to me has had.
[01:11.260 --> 01:29.320]  But starting in 2011 and I think 2012, Harper was the chief technology officer of the Barack Obama campaign and was one of the people responsible for the election, the re-election of Barack Obama to president of the United States.
[01:29.680 --> 01:32.160]  So with that, I will hand it off.
[01:32.280 --> 01:34.180]  And please, some applause for Harper Reed.
[01:35.260 --> 01:36.820]  Thank you so much for having me.
[01:37.380 --> 01:38.680]  I'm so excited to be here.
[01:39.200 --> 01:39.920]  Hi, I'm Harper.
[01:40.080 --> 01:43.180]  You can, if there's any questions or comments, please reach out on Blue Sky.
[01:43.700 --> 01:45.240]  I'm super excited to talk.
[01:45.860 --> 01:47.860]  This is, I'm so excited to be here.
[01:47.940 --> 01:51.120]  I have never been more nervous or excited about a talk in my whole life.
[01:51.520 --> 01:54.040]  Typically, I'm speaking to people who I'm like, oh, this is very nice.
[01:54.040 --> 01:55.260]  But here, I love this.
[01:56.040 --> 01:57.980]  This is me in like 1995.
[01:59.800 --> 02:04.060]  If you notice that shirt there, that's a very happy about that shirt.
[02:04.840 --> 02:06.320]  I wanted to go to the original HOPE.
[02:06.460 --> 02:07.040]  I love this.
[02:07.120 --> 02:07.920]  I grew up in Colorado.
[02:07.920 --> 02:09.260]  I did not have an opportunity.
[02:09.480 --> 02:11.140]  My town didn't have a 2600 meeting.
[02:11.460 --> 02:14.700]  I would just look at the back of that magazine wishing I lived in a bigger city.
[02:14.960 --> 02:17.280]  And so being here is just so exciting.
[02:17.660 --> 02:19.880]  Thank you to Greg and Aesthetics for helping me get here.
[02:20.260 --> 02:21.780]  And yeah, I'm very happy to be here.
[02:25.260 --> 02:30.560]  So today, I'm going to talk about something that I really like to talk about.
[02:30.560 --> 02:31.540]  It's very uncomfortable.
[02:31.760 --> 02:33.140]  And it's kind of fun.
[02:33.420 --> 02:34.560]  So bear with me.
[02:34.720 --> 02:36.620]  But basically, it's like this.
[02:37.000 --> 02:37.440]  Technology.
[02:37.780 --> 02:39.500]  In the end, it will all work out.
[02:42.080 --> 02:42.880]  I don't know.
[02:42.960 --> 02:43.320]  We'll see.
[02:43.480 --> 02:46.700]  But first, let me talk about my favorite topic in the whole world, which is myself.
[02:47.860 --> 02:49.240]  You guys know who these guys are?
[02:49.960 --> 02:50.400]  Right?
[02:50.640 --> 02:50.780]  Okay.
[02:50.780 --> 02:53.200]  So it's Steve Jobs, Steve Wozniak, and then some other guy.
[02:54.680 --> 02:55.920]  That was my first computer.
[02:56.080 --> 02:57.740]  This is a book I read when I was very young.
[02:57.780 --> 02:59.760]  And it changed my entire outlook on life.
[03:00.440 --> 03:10.060]  And then, you know, one of the things I really like about that book, it has set a tone for my entire career, which I've spent a lot of time in open-source, a lot of time building teams around this technology, and really enjoying this stuff.
[03:10.520 --> 03:12.000]  I've always called myself a hacker.
[03:12.640 --> 03:13.520]  I'm a programmer.
[03:13.680 --> 03:15.040]  I spend a lot of time touching code.
[03:15.100 --> 03:18.200]  I really believe in the words of Dr. Dre, which is touch code every day.
[03:18.360 --> 03:19.540]  I think it's really important.
[03:20.580 --> 03:22.200]  I was at this company called Threadless.
[03:22.300 --> 03:23.920]  Does anyone remember Threadless from back in the day?
[03:24.160 --> 03:25.660]  Really great company, really fun.
[03:25.820 --> 03:27.820]  We accidentally invented crowdsourcing.
[03:28.220 --> 03:29.260]  We did not mean to.
[03:29.480 --> 03:31.020]  We didn't know what crowdsourcing was.
[03:31.040 --> 03:33.500]  And we went to MIT for a conference, and they said, this is Threadless.
[03:33.540 --> 03:34.820]  They invented crowdsourcing.
[03:34.920 --> 03:35.700]  And we were like, what?
[03:35.920 --> 03:37.760]  We just take this and turn it into that.
[03:37.820 --> 03:39.240]  It's a very simple thing that we did.
[03:39.400 --> 03:41.160]  We obviously knew a lot about the crowd.
[03:41.220 --> 03:43.320]  We spent a lot of time thinking about the crowd.
[03:43.440 --> 03:47.760]  And we really engaged in these four steps, which was get designs from people like y'all.
[03:47.860 --> 03:48.660]  You would upload them.
[03:48.660 --> 03:50.720]  And then you would score them.
[03:50.880 --> 03:52.340]  And then we would get all this money.
[03:52.400 --> 03:52.960]  And it was perfect.
[03:54.460 --> 03:55.200]  It was fun.
[03:55.300 --> 03:56.600]  This was a really nice experience.
[03:57.000 --> 04:01.140]  One of the key things I learned, and this is a theme throughout my life, is we did none of the hard work.
[04:01.240 --> 04:02.140]  I don't want to do any work.
[04:02.260 --> 04:03.200]  Does anyone here want to work?
[04:04.340 --> 04:04.660]  Okay.
[04:04.820 --> 04:06.120]  Some of you probably like work.
[04:06.380 --> 04:08.040]  Liking work is different than wanting to work.
[04:08.040 --> 04:10.260]  But around 2009, I accomplished all my goals.
[04:10.360 --> 04:13.860]  And so I did the one thing that I love doing over and over again, which is I quit my job.
[04:14.220 --> 04:16.260]  And I highly recommend everyone do that today.
[04:16.500 --> 04:20.360]  We can do a party in the back where we just pull out our phones and email our bosses.
[04:21.020 --> 04:28.720]  So I went on a little bit of a vision quest at that moment where I kind of wandered the earth and then was dragged into this amazing organization, the Obama campaign, by this guy.
[04:28.720 --> 04:30.040]  This is Michael Slaby.
[04:30.160 --> 04:31.420]  He was a CTO in 2008.
[04:31.560 --> 04:34.560]  And Michael Slaby was a pretty normal political guy.
[04:34.700 --> 04:35.480]  And they grabbed me.
[04:36.260 --> 04:37.800]  This is a good transition, I think.
[04:40.100 --> 04:41.580]  This was a very fun time.
[04:41.580 --> 04:42.840]  I was a CTO in 2012.
[04:43.080 --> 04:46.880]  And by fun, I mean, it was a horrible experience that was great.
[04:46.900 --> 04:48.500]  And I learned the words rewarding.
[04:48.720 --> 04:50.660]  Whenever someone says, you say to someone, how was that?
[04:50.720 --> 04:51.980]  And they say, that was a rewarding experience.
[04:51.980 --> 04:52.880]  That just means it sucks.
[04:53.740 --> 04:55.500]  It's just they don't know how to describe it.
[04:55.600 --> 04:56.680]  Politics is politics.
[04:56.940 --> 05:00.160]  The first part of House of Cards until they kill someone is exactly what it was like.
[05:02.980 --> 05:04.060]  Anyway, why me?
[05:04.300 --> 05:06.420]  This is the first question that was asked to me by my mother.
[05:06.520 --> 05:07.000]  Why you?
[05:07.360 --> 05:08.400]  And I was like, come on, Mom.
[05:08.700 --> 05:12.680]  Well, the reason is, is they were looking for someone that was different than 2008.
[05:12.800 --> 05:16.280]  2008, it was a startup.
[05:16.680 --> 05:17.740]  2012, it was enterprise.
[05:17.960 --> 05:20.280]  And for some reason, I represented enterprise technology.
[05:20.280 --> 05:24.080]  And we had a very simple mantra, which was just hiring engineers to do engineering.
[05:24.220 --> 05:25.420]  This is a very bizarre thing.
[05:25.620 --> 05:28.440]  We had a lot of engineers, mostly from these great companies.
[05:28.600 --> 05:29.420]  This is the most important one.
[05:29.760 --> 05:31.240]  And then we had all this tech staff.
[05:31.280 --> 05:33.160]  And it was a really fun thing to do.
[05:33.400 --> 05:34.120]  Pretty wild.
[05:34.240 --> 05:35.360]  We started from absolutely zero.
[05:35.600 --> 05:38.560]  And then we just had to go and build all of this great stuff.
[05:38.780 --> 05:39.300]  This was great.
[05:39.480 --> 05:42.660]  And on election day, it was super chill, calm, relaxing.
[05:43.600 --> 05:44.560]  And luckily we won.
[05:44.680 --> 05:45.860]  Or I wouldn't be here today.
[05:46.160 --> 05:46.920]  That's my theory.
[05:46.920 --> 05:50.360]  Anyway, so I went on another Vision Quest after that.
[05:50.440 --> 05:51.680]  We started a company called Modest.
[05:51.820 --> 05:53.480]  This company we named after ourselves.
[05:53.660 --> 05:54.360]  We're very modest.
[05:54.560 --> 05:59.280]  And we had this very simple goal to build e-commerce apps that were very exciting.
[05:59.500 --> 06:01.300]  And it worked out pretty well.
[06:01.300 --> 06:05.060]  And we had this really good cheat code for scaling a business, which is to sell it to a bigger company.
[06:05.460 --> 06:08.860]  It turns out that we can build cool things, but I don't know if we can build big things.
[06:08.900 --> 06:10.900]  And they can build big things, but not cool things.
[06:10.980 --> 06:11.900]  And so it worked out pretty well.
[06:12.220 --> 06:14.520]  And this was a lot of fun, but I didn't last very long.
[06:14.520 --> 06:15.920]  So I went on another vision quest.
[06:16.260 --> 06:18.980]  And this is the time where I installed a payphone in my backyard.
[06:20.140 --> 06:21.240]  Dreams can come true.
[06:21.940 --> 06:22.980]  Even you can do this.
[06:23.040 --> 06:23.680]  You can just buy those.
[06:24.120 --> 06:25.100]  And this is 2019.
[06:25.100 --> 06:26.240]  You guys remember 2019.
[06:26.240 --> 06:27.280]  That was a great time.
[06:27.480 --> 06:28.680]  A time of dreams.
[06:29.000 --> 06:30.140]  An exciting time.
[06:30.300 --> 06:31.220]  We were focusing on robots.
[06:31.520 --> 06:33.600]  And then, you know what's coming.
[06:34.260 --> 06:35.520]  That year sucked.
[06:35.740 --> 06:36.920]  That was a horrible year.
[06:37.140 --> 06:45.900]  And so my plan of just kind of fucking off and doing this fun stuff turned into this really nightmarish time of just sitting in my basement like we all did and trying to figure out what the new world was.
[06:46.300 --> 06:48.080]  And this was a very strange time.
[06:48.080 --> 06:53.420]  And there was a lot of things that happened there that I think I appreciate today and a lot of things that happened there that scared the shit out of me today.
[06:53.700 --> 06:54.560]  But change happened.
[06:54.860 --> 06:55.400]  This was great.
[06:55.700 --> 06:57.240]  So after that, I started a company called Galactic.
[06:57.720 --> 06:58.780]  This was a funny company.
[06:58.980 --> 07:02.200]  We were trying to build things for the new Internet, which was kind of this Web 3 thing.
[07:02.340 --> 07:03.440]  And we thought this was going to be great.
[07:03.600 --> 07:04.720]  We really like stable coins.
[07:04.880 --> 07:07.980]  Our view is crypto isn't real until you can pay for rent or child care.
[07:08.360 --> 07:10.820]  The moment you can pay for rent and child care, I'll believe it is real.
[07:10.820 --> 07:13.780]  So we were trying to do that, and that did not work.
[07:14.780 --> 07:18.680]  So that was kind of a... that was a bummer, but now I'm on another vision quest.
[07:18.800 --> 07:19.620]  We're doing some cool stuff.
[07:19.880 --> 07:22.500]  Every time I go on these things, it is weirder each time.
[07:22.660 --> 07:26.200]  And so this is the beginning of my presentation.
[07:26.820 --> 07:29.360]  I think a lot about my career, as you can just see.
[07:29.860 --> 07:31.240]  I loved every one of these things.
[07:31.340 --> 07:32.840]  Even the ones that failed, the team was beautiful.
[07:32.960 --> 07:33.800]  It was a beautiful time.
[07:33.920 --> 07:36.800]  I spent all this time working with people that I loved.
[07:37.260 --> 07:41.480]  And in every case, you know, somehow out of this, I became known as this innovator.
[07:41.560 --> 07:45.100]  And it was this beautiful thing where I would just sit there, and we were building all these experiences.
[07:45.100 --> 07:48.380]  And we spent a lot of time building things, being thoughtful, thinking about how to be careful.
[07:48.800 --> 07:53.440]  We did the thing called practicing failure, where we would do a lot of experiences and say, how could this go wrong?
[07:53.700 --> 07:55.820]  How could this be used in a way that we don't intend it?
[07:56.620 --> 07:59.700]  We had constant discussions about how do we make this not creepy?
[08:00.020 --> 08:02.120]  How do we make sure that this is safe?
[08:02.420 --> 08:03.560]  How do we make sure it's legal?
[08:04.700 --> 08:07.680]  And we also asked this question of, should we do this?
[08:07.760 --> 08:10.500]  And this was sometimes a complicated conversation.
[08:10.740 --> 08:14.440]  This started back in Threadless, through Obama World, through all of these startups.
[08:15.100 --> 08:16.860]  And the thing is, we also set these checkpoints.
[08:16.920 --> 08:18.160]  We spent a lot of time doing this.
[08:18.180 --> 08:26.520]  And we did everything we could to try and help enable safety in spaces that, for the most part, people had not yet invented, had not yet built inside of.
[08:27.260 --> 08:29.460]  And it was wonderful in 2013.
[08:29.460 --> 08:30.700]  Like, I was on the top of the world.
[08:30.760 --> 08:31.080]  This was great.
[08:31.140 --> 08:31.700]  I was famous.
[08:31.920 --> 08:32.340]  Look at this.
[08:32.520 --> 08:35.280]  I was on a cover of a rock and roll magazine in Italy.
[08:36.080 --> 08:37.620]  And then in Spain.
[08:37.780 --> 08:38.820]  And then in Germany.
[08:39.040 --> 08:39.980]  This was exciting.
[08:40.300 --> 08:41.220]  And then what happened?
[08:41.460 --> 08:42.620]  Oh yeah, 2014.
[08:43.020 --> 08:43.460]  Ukraine.
[08:43.680 --> 08:44.120]  Misinformation.
[08:46.160 --> 08:46.600]  Right?
[08:47.500 --> 08:49.360]  Disinformation was becoming this really big deal.
[08:49.740 --> 08:50.180]  2016.
[08:50.240 --> 08:52.500]  The same misinformation, disinformation cannons.
[08:52.760 --> 08:55.160]  The problem was, this was our playbook that we invented.
[08:55.800 --> 08:57.840]  We spent a lot of time building this technology.
[08:57.840 --> 09:00.120]  But they did it better than we did.
[09:00.280 --> 09:02.140]  In a way that was way better.
[09:02.420 --> 09:03.080]  Way scarier.
[09:03.240 --> 09:04.200]  We built these foundations.
[09:04.720 --> 09:11.620]  These foundations were foundations that we were proud of, that we spent a lot of time and then they were weaponized against us in many ways.
[09:12.240 --> 09:17.860]  And so there is this thing happening where it felt a little bit like the things that I was proud of.
[09:17.940 --> 09:28.440]  The things that I had achieved what I thought That was the answer, this kind of, I'm on cover of a magazine, that type of thing, were now used in a way where every time someone talked to me about it, Harper, what happened here?
[09:28.680 --> 09:29.360]  What was this?
[09:29.420 --> 09:31.520]  I would just be like, oh yeah, I remember when we built that.
[09:32.780 --> 09:34.020]  It continues today, right?
[09:34.120 --> 09:35.060]  So we see this already.
[09:35.200 --> 09:36.260]  This is not a new thing.
[09:36.380 --> 09:37.840]  This is happening over and over again.
[09:37.940 --> 09:41.320]  And so it's happening just with this social media technology that we built.
[09:41.820 --> 09:42.980]  And we were really careful.
[09:43.220 --> 09:50.000]  Like I was in a lot of meetings with a lot of people from lawyers to UX people to ethics type people talking about how this could be used against us.
[09:50.940 --> 09:52.720]  But the thing is, is this is not where it ends.
[09:53.120 --> 09:55.300]  We also invented crowdsourcing, right?
[09:55.480 --> 09:57.460]  Like this is also weaponized.
[09:57.880 --> 09:59.040]  This is really exciting.
[09:59.180 --> 10:06.220]  It's kind of fun to look back at my career and think of all these great things we built and see them used in a way that we didn't intend.
[10:07.540 --> 10:10.100]  A lot of the stuff that we've done was this, right?
[10:10.200 --> 10:11.440]  We did not intend for this to happen.
[10:12.160 --> 10:19.880]  I remember being interviewed by a reporter in Germany and she just kept pushing on like, but you're using data in a wrong way.
[10:20.020 --> 10:20.900]  And I was just like, no, no, no, no.
[10:20.980 --> 10:21.480]  It's beautiful.
[10:22.140 --> 10:23.240]  This is going to work great.
[10:23.400 --> 10:28.460]  And for some reason, who knows why they had a historical context for what was coming.
[10:29.240 --> 10:34.380]  And she was pushing on this and I just was not listening because our intention was to build something that was right.
[10:35.360 --> 10:37.620]  Everything we did was rooted in this hope.
[10:38.020 --> 10:42.240]  And yet, even with all the steps we did to be careful, we weren't careful enough.
[10:42.520 --> 10:46.080]  And the unintended consequences of those very decisions of hope were used against us.
[10:47.240 --> 10:51.280]  One thing that I've come to believe is that technology is a very, very sharp sword.
[10:51.620 --> 10:53.580]  And I don't know if it's because we're creatives.
[10:53.580 --> 10:57.420]  I don't know if it's because we're building something with hope and excitement, but we only think of one side.
[10:57.540 --> 10:58.840]  We never think about the other side.
[10:58.900 --> 11:00.540]  We have to figure out how to think about both sides.
[11:01.060 --> 11:02.800]  And so now we have AI.
[11:03.700 --> 11:04.480]  I love AI.
[11:04.600 --> 11:05.560]  Anyone here like it?
[11:06.480 --> 11:07.200]  Am I the only... Okay, good.
[11:07.300 --> 11:10.700]  I was a little worried that I was going to be the only person that was going to be like, AI is incredible.
[11:11.120 --> 11:12.920]  But I really, really like AI.
[11:13.060 --> 11:14.020]  So there's only two of us.
[11:14.100 --> 11:15.340]  We'll go quit our jobs in the back.
[11:17.220 --> 11:19.460]  It's some of the coolest tech I've ever used.
[11:19.640 --> 11:20.860]  And I used AIX.
[11:24.280 --> 11:24.840]  Thank you.
[11:24.980 --> 11:25.440]  Thank you.
[11:25.600 --> 11:26.820]  The two people in the back.
[11:26.920 --> 11:31.300]  But I really believe that we're in the middle of this really unprecedented shift with AI.
[11:32.240 --> 11:38.820]  And the thing that's funny about this is in 2017, I don't know if you all remember this, all the articles about AI and machine learning.
[11:38.980 --> 11:39.580]  We're like, oh, cool.
[11:39.700 --> 11:42.600]  We're going to replace truckers, warehouse workers, manual laborers.
[11:42.760 --> 11:45.040]  Most of those people were conveniently not us.
[11:45.820 --> 11:47.760]  It was very easy for us to have this conversation.
[11:47.900 --> 11:49.480]  Oh, let's go replace the warehouse workers.
[11:49.600 --> 11:50.100]  It'll be great.
[11:50.800 --> 11:51.120]  Robots.
[11:51.320 --> 11:52.760]  And we're like, what will happen to the warehouse workers?
[11:52.860 --> 11:54.440]  Oh, no, they'll just get another job.
[11:54.540 --> 11:54.980]  It'll be fine.
[11:55.520 --> 11:56.800]  They will re-educate.
[11:58.020 --> 11:59.560]  Then Chad should be T-Lonce in 2022.
[12:00.660 --> 12:02.960]  2025, you know, you just replace half your workforce.
[12:03.700 --> 12:05.080]  Knowledge workers, people like me.
[12:06.060 --> 12:12.080]  This is an uncomfortable thing that I think suddenly all of, you know, the people like me are like, whoa, we got to think of safety.
[12:12.180 --> 12:14.900]  But when you were back here, no one was thinking about safety.
[12:15.800 --> 12:17.440]  And so why is this?
[12:17.640 --> 12:21.000]  And there's a couple things that are happening right now, specifically within tech.
[12:21.140 --> 12:21.900]  There's a lot of layoffs.
[12:22.100 --> 12:23.720]  I don't know if this is all AI.
[12:24.700 --> 12:29.400]  All I know is that in the past, startups and companies I've started started with 50% more engineers.
[12:30.260 --> 12:33.180]  We have half the number of engineers now and we're 200% more productive.
[12:33.480 --> 12:34.560]  We're way faster.
[12:34.880 --> 12:37.240]  The team is drastically, drastically changed.
[12:37.380 --> 12:42.180]  For instance, we do 90% code generated, like generative AI code.
[12:42.560 --> 12:43.380]  That's nuts.
[12:43.640 --> 12:44.500]  I'm a programmer.
[12:44.620 --> 12:45.820]  My entire life was programming.
[12:45.980 --> 12:46.740]  I do it for fun.
[12:46.820 --> 12:48.540]  When I open my laptop, there's always a terminal.
[12:48.840 --> 12:50.420]  I don't even use an IDE anymore.
[12:50.940 --> 12:51.740]  It's insane.
[12:51.960 --> 12:54.220]  I don't know how to handle this emotionally.
[12:54.640 --> 12:57.760]  But one of the thing I do notice is we hired seniors and we hired juniors.
[12:58.180 --> 12:59.380]  The seniors are coaches.
[12:59.920 --> 13:02.200]  We heard the comment about people with gray hair.
[13:02.300 --> 13:03.320]  We hire a lot of those folks.
[13:03.700 --> 13:04.440]  That's me.
[13:04.720 --> 13:08.200]  Then we hire juniors, people who are just fresh out of college, who don't have any experience.
[13:08.580 --> 13:15.560]  The juniors end up hitting this button that says continue or magic or whatever it is to get the AIs to do things while the seniors coach.
[13:15.880 --> 13:17.020]  This is working really well.
[13:17.160 --> 13:21.680]  I am very happy with how this works from a capitalism frame.
[13:22.320 --> 13:23.980]  But the mid-level is gone.
[13:24.220 --> 13:25.420]  There's no mid-level anymore.
[13:25.700 --> 13:26.700]  We don't hire them.
[13:26.960 --> 13:33.140]  You're either very senior and strategic, thoughtful, thinking about things, or you're very junior, where you're focused only on execution.
[13:33.460 --> 13:34.840]  The middle just has disappeared.
[13:35.280 --> 13:42.020]  And I don't know what happens to those folks, because for those folks to suddenly become a senior in this timeframe is very, very, very hard.
[13:42.120 --> 13:43.080]  Some people will be able to do it.
[13:43.660 --> 13:48.660]  For them to become a junior and maybe drop a pay scale or what have you is going to be very, very painful.
[13:49.300 --> 13:55.460]  So I think probably they're just going to leave the workforce, find a job somewhere else, do something that isn't, isn't great.
[13:56.120 --> 13:58.000]  But I don't think this is just the middle.
[13:58.100 --> 14:00.040]  I think the seniors and juniors will be replaced.
[14:00.320 --> 14:04.320]  I was talking to a friend that does a startup that does this, this type of code generation stuff.
[14:04.340 --> 14:07.020]  And he was saying, seniors will be replaced in three years.
[14:07.860 --> 14:11.100]  So I have three years before I'm completely replaced as a programmer.
[14:11.860 --> 14:12.820]  Now, now, wait a minute.
[14:13.000 --> 14:15.380]  Like, let's, let's just chill out for a second.
[14:15.560 --> 14:16.420]  Y'all remember this book?
[14:17.740 --> 14:19.520]  This is one of the worst books ever made, I think.
[14:19.840 --> 14:20.200]  I don't know.
[14:20.360 --> 14:21.280]  Does anyone remember it?
[14:21.460 --> 14:22.160]  Y'all remember this book?
[14:22.300 --> 14:24.860]  This was the book they would hand you when they were doing layoffs.
[14:25.000 --> 14:25.960]  They're like, we laid off your team.
[14:26.100 --> 14:26.820]  Here, read this book.
[14:27.080 --> 14:28.000]  And you're just like, what?
[14:28.080 --> 14:28.640]  What do you mean?
[14:29.040 --> 14:30.420]  Why am I reading a book about cheese?
[14:31.680 --> 14:43.860]  One of the things I think a lot about is like, as a CEO, as a head of a project, as a head of a whatever, why is the only option we have, especially in this time where, from a technology standpoint, we have some abundance?
[14:44.260 --> 14:45.740]  Why do we only think about cutting?
[14:47.000 --> 14:50.320]  And I think we need to really lean into the shift and design the world we want.
[14:50.940 --> 14:52.420]  But I don't think that's it.
[14:52.640 --> 14:55.900]  I think there's all these other questions of like, what do we do as technology people?
[14:56.160 --> 14:57.340]  Do we go and become bakers?
[14:57.580 --> 14:58.360]  Move to the woods?
[14:58.500 --> 15:01.620]  Like, I know every programmer I've ever met has a fantasy about this.
[15:02.020 --> 15:04.640]  They want to, you know, it's like, they want to go work at Home Depot.
[15:04.780 --> 15:05.260]  And it's like, why?
[15:05.360 --> 15:06.660]  And it's like, just because, you know, I don't know.
[15:07.760 --> 15:08.780]  That's a different talk maybe.
[15:08.920 --> 15:09.960]  But I can't do this.
[15:10.040 --> 15:11.600]  That's not what, that's not what I do.
[15:11.800 --> 15:13.940]  Like, I'm a technology.
[15:14.100 --> 15:15.040]  My tools are technology.
[15:15.220 --> 15:18.060]  When I, like I said, when I open my laptop, there's a, there's a terminal there.
[15:18.760 --> 15:23.620]  I'm trying to build a, like, my dream is a, is a, is like a little, I don't know, a cyber deck or whatever.
[15:23.720 --> 15:25.240]  But we have to be really careful here.
[15:25.440 --> 15:26.900]  Like, this is something I keep thinking about.
[15:27.020 --> 15:29.680]  And I keep thinking about this with regards to AI of making things worse.
[15:30.740 --> 15:33.080]  Because AI does it exactly what we tell it.
[15:33.140 --> 15:35.400]  And I was talking to my team about this, a good friend.
[15:35.740 --> 15:37.060]  And he was just like, oh, no, no, no.
[15:37.200 --> 15:38.140]  AI is not like a genie.
[15:38.200 --> 15:38.940]  It's like the monkey paw.
[15:39.880 --> 15:41.440]  It does do whatever you want.
[15:42.220 --> 15:43.320]  And I was like, oh, okay, cool.
[15:43.320 --> 15:45.280]  But I cannot figure out what happened at the end of that book.
[15:45.380 --> 15:45.820]  Does anyone know?
[15:46.860 --> 15:47.900]  If you do know, tell me.
[15:47.980 --> 15:49.240]  Because I have not read it and I want to know.
[15:49.320 --> 15:51.400]  Because I feel like using it as an example, I should know the end.
[15:51.660 --> 15:52.520]  But here we are.
[15:54.100 --> 16:04.380]  So with all that said, I was trying to think of how, what are some tips that we use, obviously very successful, at building things without destroying the world.
[16:04.540 --> 16:08.540]  And so let me go through these things and then we can talk about them later.
[16:08.720 --> 16:10.420]  And I'm not super confident on this.
[16:10.880 --> 16:12.000]  I have a good track record.
[16:12.640 --> 16:13.800]  So let's see how it goes.
[16:14.680 --> 16:18.280]  Also, the caveat here is some of these things are intensely not popular.
[16:18.460 --> 16:21.880]  I think someone would describe some of them as woke, which I'm not worried about.
[16:22.160 --> 16:22.720]  Fuck it.
[16:23.260 --> 16:25.460]  So the first one is think of the people.
[16:26.160 --> 16:27.680]  I know this is a weird thing.
[16:28.280 --> 16:30.840]  And does anyone know this picture, what this is from?
[16:32.140 --> 16:33.380]  Yeah, it's people.
[16:35.120 --> 16:38.680]  But the thing is, is I think we need to really engage in humanist ideas.
[16:39.180 --> 16:46.420]  Like when we are thinking, when I start a company, I raise venture backed, I raise venture backed money for startups from San Francisco VCs.
[16:46.940 --> 16:52.040]  And when we pitch them, we say things like, we're going to interrogate systems of power we do not like.
[16:52.760 --> 16:53.800]  And they give us money.
[16:54.640 --> 16:57.540]  You don't have to, you can think of humans.
[16:57.540 --> 16:59.640]  You don't have to only think of capitalism.
[17:00.320 --> 17:02.840]  Now, we don't get as much money as the people who are thinking of capitalism.
[17:02.840 --> 17:04.760]  But here we are, we're still doing it.
[17:05.320 --> 17:14.820]  Another thing that we truly value and has saved our lives many times, and I think the work we would have done on Obama would have been way worse if we did not have this, this is over and over again is hyper diverse teams.
[17:16.200 --> 17:17.260]  This is very simple.
[17:17.800 --> 17:20.140]  You know, other perspectives will help you.
[17:20.320 --> 17:21.600]  There's a great example of this.
[17:21.740 --> 17:23.400]  This is one of my favorite examples of machine learning.
[17:24.220 --> 17:26.240]  Self-driving cars hit black people more.
[17:26.520 --> 17:28.560]  You can imagine what Google's team looked like.
[17:29.680 --> 17:32.080]  I have a better example, maybe not better, different example.
[17:32.280 --> 17:34.420]  I had a team of people I was advising.
[17:34.720 --> 17:37.240]  They were an incredible machine learning team doing computer vision.
[17:37.400 --> 17:38.160]  They were incredible.
[17:38.360 --> 17:39.980]  They were the smartest people I knew in this stuff.
[17:40.080 --> 17:43.560]  They ended up selling the company to Amazon for gobs of money.
[17:44.860 --> 17:45.800]  They were Chinese.
[17:45.860 --> 17:46.580]  They came in.
[17:46.640 --> 17:47.400]  They did this great demo.
[17:47.760 --> 17:50.820]  They had a thing where you take a selfie and it would guess your age.
[17:50.960 --> 17:52.240]  And I was like, that's pretty impressive.
[17:52.520 --> 17:53.300]  I should do it.
[17:53.440 --> 17:54.040]  I did it.
[17:54.080 --> 17:55.400]  I was 240 years old.
[17:55.920 --> 17:57.840]  And then, and then my buddy Clint did it.
[17:58.980 --> 18:02.540]  Also, side note, Clint and I were supposed to come here together, but we didn't really communicate.
[18:02.660 --> 18:03.540]  So, Clint arrives tomorrow.
[18:03.580 --> 18:04.360]  I speak today.
[18:04.560 --> 18:04.960]  Who knew?
[18:05.400 --> 18:10.500]  But Clint does it, 200 years old.
[18:10.620 --> 18:11.460]  And they're like, what?
[18:11.580 --> 18:13.340]  These are like the smartest ML people I know.
[18:13.420 --> 18:14.320]  They are so smart.
[18:14.520 --> 18:16.020]  They get in a little huddle.
[18:16.240 --> 18:20.240]  They're talking and they come back and they're like, oh, we just never tested it on anyone who wasn't Chinese.
[18:21.080 --> 18:22.220]  Their network was Chinese.
[18:22.360 --> 18:23.100]  Their team was Chinese.
[18:23.420 --> 18:24.600]  They had never touched talk.
[18:24.740 --> 18:26.920]  They had never tested on a white person, a black person, whatever.
[18:27.340 --> 18:29.600]  They just, their network was so tight.
[18:30.000 --> 18:30.920]  Their team was so tight.
[18:31.120 --> 18:34.140]  It doesn't matter who you are, where you come from, the diversity will help.
[18:34.700 --> 18:37.920]  Because homogeneous teams will only think of one side of the sword no matter what.
[18:38.120 --> 18:41.460]  And in these types of situation, there are so many sides of the sword that we need to really think about.
[18:42.600 --> 18:44.060]  This is one that I think is really hard.
[18:45.740 --> 18:48.240]  Oftentimes, as creative people, we are inventing things.
[18:48.340 --> 18:50.780]  We think that no one has come before us.
[18:51.080 --> 18:52.240]  I think that's often been wrong.
[18:53.540 --> 18:56.740]  I find that my peers are reading only hacker news.
[18:57.220 --> 18:58.240]  That's a lot of bias.
[18:59.980 --> 19:00.760]  It's fine.
[19:00.920 --> 19:03.080]  But like, there's a lot of other things out there, right?
[19:03.620 --> 19:09.960]  I find that when I prioritize voices who are underrepresented and often ignored, that the outcomes that I have are better.
[19:10.700 --> 19:12.800]  Because many times they have been here before.
[19:13.220 --> 19:17.860]  Many times they have experienced these things before, not with AI, not with crowdsourcing, but with other things.
[19:18.220 --> 19:19.660]  Interrogating those systems of power.
[19:19.880 --> 19:26.980]  And that leads me to this one, which I find very, very hard for me, which is, how do you build a threat model for your product, for your project, etc.?
[19:26.980 --> 19:35.260]  In every case that I put up there about the work that we did, there were signs that kind of suggested maybe this was the path.
[19:35.420 --> 19:39.300]  Whether it was the German person telling me, hey, why don't you think a little harder about this, etc.?
[19:39.300 --> 19:44.660]  And every time I was just like, ah, maybe, you know, like I just, I didn't really believe them because I wasn't listening.
[19:45.420 --> 19:50.840]  But the thing is, is humans are really clever and they're really good at taking something and building upon it.
[19:51.600 --> 19:53.460]  We know this from the hacker community.
[19:53.560 --> 19:54.940]  We know this from the open-source community.
[19:55.080 --> 19:56.840]  Why wouldn't we know this from an adversary?
[19:57.640 --> 20:00.600]  Why don't we trust them to be good at their job?
[20:01.540 --> 20:05.000]  And that leads me to this other thing, which we've done a lot of, which is tabletop exercises.
[20:07.960 --> 20:09.180]  Invite that diverse team.
[20:09.440 --> 20:10.640]  Invite people you wouldn't expect.
[20:10.960 --> 20:13.540]  Artists, cops for some reason, other people.
[20:14.460 --> 20:15.060]  Diversity, right?
[20:15.280 --> 20:19.480]  And we invite them to game out all the eventualities.
[20:19.680 --> 20:21.880]  Invite them to really think about what goes wrong.
[20:22.040 --> 20:26.400]  There's a reason why the defense companies do this, because it's very good at understanding that.
[20:26.860 --> 20:28.040]  Fail safety is another one.
[20:28.120 --> 20:32.140]  I love this because if you go to like the NASA, you go to Orlando or whatever, you see all those NASA things.
[20:32.200 --> 20:33.600]  They always have this thing about fail safety.
[20:33.780 --> 20:35.220]  We got to bring them back alive.
[20:35.520 --> 20:40.720]  And it's like, yeah, but there's all these things about like just making sure that you're thinking about the humans.
[20:42.180 --> 20:44.600]  This is something that I think we don't do when we're building software.
[20:44.720 --> 20:45.300]  It's very strange.
[20:45.780 --> 20:47.460]  I mean, another thing is basically do what I'm doing.
[20:48.080 --> 20:48.760]  Launder your guilt.
[20:48.900 --> 20:49.840]  Talk about your failures.
[20:50.160 --> 20:54.440]  Get up here and tell a story about how you fucked up, what was successful, what worked really well.
[20:55.180 --> 20:57.540]  And then I think this is the second to last thing.
[20:57.720 --> 20:58.360]  It's like, should.
[20:58.520 --> 21:03.300]  I think this is the question that we all should ask ourselves for almost everything just because we can.
[21:03.460 --> 21:03.780]  Should we?
[21:05.560 --> 21:07.680]  Maybe no, but maybe yes.
[21:07.920 --> 21:11.900]  Like, but actually going through that process is really helpful and do it with someone else.
[21:12.200 --> 21:15.600]  But the final one is design and creativity because like, here's the thing.
[21:17.200 --> 21:21.060]  With a diverse team, you get stuff that isn't necessarily your world.
[21:21.880 --> 21:29.200]  If you ask people what they think and you go through that exercise of should, then you get things that aren't necessarily maybe what you think.
[21:29.340 --> 21:31.320]  And this is about reacting to dominant culture.
[21:32.080 --> 21:37.620]  Sometimes we're building with inside of a dominant culture that we don't even realize is dominant.
[21:37.700 --> 21:41.380]  And we have to reach outside of that dominant culture to understand why we're fucking this up.
[21:42.020 --> 21:44.240]  But the key here is just designing the world we want.
[21:44.360 --> 21:47.340]  We have to be building that thing in the future that we want, that we're looking for.
[21:47.420 --> 21:48.280]  And it is up to us.
[21:49.200 --> 21:51.260]  And so do I think this will work, these things?
[21:52.340 --> 21:52.740]  Maybe.
[21:53.260 --> 21:55.440]  Previously, this said no, but I thought that was a little negative.
[21:56.480 --> 21:57.820]  Do I think it's better than the alternative?
[21:57.980 --> 21:58.640]  Yeah, absolutely.
[21:58.860 --> 21:59.420]  Like this is the thing.
[21:59.560 --> 22:09.500]  I think that that all of this innovation that we're seeing, whether it's, you know, crowdsourcing from back in the early 2000s to the social media cannons that we build on Obama land to the AI displacing everything, I think it's inevitable.
[22:09.920 --> 22:10.700]  Someone is going to do it.
[22:10.820 --> 22:11.500]  I misspelled inevitable.
[22:11.720 --> 22:12.300]  Don't tell anyone.
[22:13.420 --> 22:16.700]  But if it's not me, then doing this, then it's going to be y'all, I hope.
[22:16.880 --> 22:18.580]  And if it's not us, then it's them.
[22:19.320 --> 22:24.340]  And I would prefer that y'all do it with safety and being thoughtful about it than them doing it without safety.
[22:24.920 --> 22:26.500]  So all that said, I'm back.
[22:26.600 --> 22:27.380]  I started a new company.
[22:27.540 --> 22:28.260]  It's an AI company.
[22:29.140 --> 22:30.720]  So, yeah, with that, I think we're all doomed.
[22:32.200 --> 22:32.680]  That's it.
[22:32.760 --> 22:34.120]  Thank you so much for having me.
[22:34.200 --> 22:35.660]  Hopefully this was a good thing.
[22:40.950 --> 22:42.870]  I have some time for questions, I think.
[22:47.150 --> 22:48.450]  And you don't have to as well.
[22:48.610 --> 22:51.110]  I'm totally fine just slinking out of here.
[22:58.880 --> 22:59.240]  Hello.
[23:00.040 --> 23:03.220]  So you said you don't need middle-level engineers, right?
[23:03.220 --> 23:04.020]  Excuse me?
[23:04.100 --> 23:04.420]  Say that again.
[23:04.500 --> 23:05.900]  You don't need middle-level engineers right now?
[23:06.800 --> 23:09.240]  Need is a complicated word, but yeah, I think that's the right one.
[23:09.460 --> 23:11.420]  But then why are you hiring the low-level engineers?
[23:11.720 --> 23:12.620]  So that's a great question.
[23:13.220 --> 23:14.560]  Actually, that's a really good question.
[23:14.680 --> 23:17.120]  The reason we hire the low engineers is because they're cheap.
[23:18.360 --> 23:20.000]  So if you think about the tiers, right?
[23:20.080 --> 23:23.160]  You have a low engineer in Chicago, where I live.
[23:23.260 --> 23:23.900]  Anyone from Chicago?
[23:24.480 --> 23:25.940]  Yeah, it's a great place to live.
[23:26.040 --> 23:26.820]  You all should move there.
[23:27.560 --> 23:28.000]  Incredible.
[23:28.580 --> 23:29.440]  Actually, never mind.
[23:29.540 --> 23:30.240]  I meant it sucks.
[23:30.320 --> 23:30.880]  It's horrible.
[23:31.360 --> 23:32.200]  Very violent.
[23:32.340 --> 23:33.160]  The weather is terrible.
[23:33.360 --> 23:34.020]  Don't move there.
[23:34.580 --> 23:35.420]  Talk to me later.
[23:35.680 --> 23:41.280]  But in Chicago, at entry-level, fresh grad engineers, like 100K, 120K, right?
[23:41.380 --> 23:42.760]  A mid-level, 250K.
[23:42.880 --> 23:45.040]  A senior, 250 to 500K.
[23:45.480 --> 23:47.760]  If you're trying to save money, who do you lay off?
[23:48.340 --> 23:49.980]  The seniors have all the knowledge.
[23:50.380 --> 23:53.340]  The middles don't yet have all the knowledge and are kind of a pain in the ass.
[23:53.580 --> 23:55.340]  And the juniors are just a pain in the ass.
[23:56.300 --> 24:00.440]  And so you look at that thing there and you're like, okay, where do I save money?
[24:00.840 --> 24:02.320]  And you lay off the juniors.
[24:02.440 --> 24:02.900]  They're cheap.
[24:03.080 --> 24:04.000]  They cost nothing.
[24:04.200 --> 24:08.200]  And then you can train them to do whatever work you want.
[24:08.520 --> 24:15.660]  The middle there is the really complicated ones, because we know that the middle has to be the middle and the juniors have to be the middle to become the senior.
[24:16.660 --> 24:19.440]  But capitalism doesn't really care about the future.
[24:19.480 --> 24:20.820]  It only cares about today, right?
[24:21.140 --> 24:27.600]  So when you look at that from like the view of like a financial model, it's very easy to say, oh, we should lay off the middle.
[24:28.400 --> 24:35.980]  And I think what we're seeing now in the industry is there's a lot of, there's a lot, especially in programming, there's a pretty high, like a 7% unemployment rate.
[24:36.540 --> 24:39.880]  And I think what we're seeing is that this has not yet hit as hard.
[24:40.380 --> 24:45.740]  But every CEO I talk to about this and ask about this agrees that the middle is the one that they're not hiring for.
[24:46.700 --> 24:52.300]  And that as someone who has been in the middle is complicated.
[24:52.740 --> 24:55.820]  And so I don't know how to solve that.
[24:55.940 --> 25:00.300]  I actually think I was talking to a friend of mine at a giant, you know, Fortune 100 company yesterday.
[25:00.520 --> 25:02.180]  And he said something that I thought was really good.
[25:02.300 --> 25:13.540]  He was like, we should just be building the products in ways that actually employ people who are at the middle of their career, like change how we work to support those careers.
[25:14.920 --> 25:18.040]  But that seems a little too hopeful for the United States of America at this moment.
[25:19.100 --> 25:19.640]  Thank you.
[25:21.600 --> 25:26.780]  My question is, what do you, what advice would you give to recent college graduates that are trying to get into this field?
[25:26.920 --> 25:29.420]  Because it just kind of looks like the hope is fading away.
[25:30.580 --> 25:32.360]  Well, I think the question is, which field?
[25:33.340 --> 25:42.620]  Like if you're trying to get into programming, would you prefer, like, would it be better to just go start your own thing and figure it out from there or work at one of these companies and just be a junior executor, I guess?
[25:42.840 --> 25:43.540]  Great question.
[25:43.760 --> 25:48.800]  I think being an entrepreneur is a compulsion and probably could be cured with medicine.
[25:50.400 --> 26:01.320]  And so I wish that I had that intervention as a youth because I probably would have been much happier and would have more money if I just would have gotten a job somewhere and did that for the same amount of time that I struggled and was participating in risk.
[26:02.020 --> 26:03.500]  With that said, it's really fun.
[26:04.320 --> 26:08.100]  If you have an idea and you feel like you can start something, then I'd highly recommend starting something.
[26:08.360 --> 26:11.260]  If you feel like you want to try it out, then just participate on a project.
[26:11.580 --> 26:16.360]  Like walking around the hall up here, there's so many projects that would love help and just participating there.
[26:17.040 --> 26:21.180]  If you are a programmer, like, I mean, getting a job isn't horrible.
[26:21.700 --> 26:22.620]  They do pay.
[26:24.020 --> 26:32.000]  I personally find that finding an employer that matches my values to be very difficult, which is why I am typically my employer.
[26:33.500 --> 26:35.620]  And I sometimes don't match my values either.
[26:36.120 --> 26:37.500]  And so it's very complicated.
[26:38.440 --> 26:39.680]  But talk to me after.
[26:39.800 --> 26:41.500]  I'd love to help connect you or what have you.
[26:41.820 --> 26:43.280]  My email is up here for a reason.
[26:43.420 --> 26:44.200]  Please email me.
[26:44.780 --> 26:49.540]  I would love to talk more about some of this stuff as needed or if there's questions or if anyone's looking for connections.
[26:50.240 --> 26:51.620]  And if you're middle, I'm sorry.
[26:51.800 --> 26:52.900]  We can just commiserate.
[26:56.310 --> 26:56.750]  Hi.
[26:56.930 --> 26:57.810]  Thanks so much for this.
[26:59.270 --> 27:03.330]  How would you measure success at fighting or combating disinformation?
[27:05.150 --> 27:09.070]  Personally, probably poor because we made that disinformation cannon.
[27:11.390 --> 27:12.850]  I find it really interesting.
[27:13.050 --> 27:23.950]  I had this very interesting experience, which I didn't really talk about, but it's kind of insinuated, which was we built all this technology that did the exact same thing that the Trump team did in 2016.
[27:24.130 --> 27:31.630]  But we were lauded as doing a good thing because we were on the Obama side, whereas Trump to us was did a bad thing because they were on the Trump side.
[27:32.010 --> 27:33.730]  It was pretty much the same tool.
[27:33.950 --> 27:38.930]  Like we didn't do the psychometric stuff and we didn't, we had access to Facebook's graph legally.
[27:38.930 --> 27:39.670]  They did not.
[27:39.810 --> 27:44.310]  Like there's some like nuances there that I think are probably not worth it because the end was pretty much the same.
[27:44.790 --> 27:50.010]  And so I think this is the same question as like, you know, who owns the truth at the moment?
[27:50.710 --> 27:51.030]  Right?
[27:51.270 --> 27:55.250]  Like in 2013, it certainly seemed like we did magic.
[27:56.150 --> 27:59.430]  2025, I'm like, oh yeah, we did other things too.
[27:59.810 --> 28:01.170]  We don't, you know, so I don't know.
[28:01.270 --> 28:02.090]  I don't know the right answer.
[28:02.490 --> 28:14.090]  I actually don't have no idea how to counter it either because I think it's the same kind of problem is that, you know, I'm awkwardly for this group on the board of a school for the Rand Corporation.
[28:14.850 --> 28:20.170]  And they had this really interesting conversation about like what happens when facts become political.
[28:21.150 --> 28:28.390]  And someone said, well, the facts have a really kind of annoying way of coming true over time.
[28:29.050 --> 28:39.730]  And I think about that a lot when some misinformation happens where it's like, this is going to be a really hard time for the future, but maybe eventually, hopefully, I guess we just, we don't die before that happens is the hope.
[28:40.130 --> 28:40.590]  Yeah.
[28:43.090 --> 28:43.570]  Okay.
[28:43.690 --> 28:45.050]  Any more questions or no?
[28:45.310 --> 28:45.850]  Oh, one more.
[28:45.990 --> 28:46.210]  Yes.
[28:46.550 --> 28:47.010]  Thank you.
[28:47.350 --> 28:47.850]  Two more.
[28:48.070 --> 28:48.170]  Whoa.
[28:48.770 --> 28:50.050]  Let's do the red shirt first.
[28:51.670 --> 28:52.350]  Thank you.
[28:57.880 --> 28:59.120]  A text question.
[29:02.300 --> 29:03.080]  No problem.
[29:03.400 --> 29:07.720]  How do you think the company loyalty among seniors is affecting junior choices for career paths?
[29:07.920 --> 29:08.660]  Great question.
[29:10.340 --> 29:11.680]  Company culture is everything.
[29:12.360 --> 29:15.560]  One of the things that I've been talking a lot of people about is apprenticeships.
[29:16.220 --> 29:23.940]  I think the thing that tech has, one of the things that tech has is this idea that no one has come here before, which is obviously not true.
[29:24.580 --> 29:28.220]  Like automation has hit factories, automation has hit accounting, automation has hit designers.
[29:28.660 --> 29:33.080]  And if you look at even designers today, they often talk about apprenticeships.
[29:33.220 --> 29:38.820]  You look at, you know, accountants, they always have these kind of apprenticeship-like programs, internships, et cetera, even at the big force.
[29:39.340 --> 29:50.180]  And I think that's what we need to do as technologists is be open to training and not just certs, like actually helping someone shadow you and spend time.
[29:50.320 --> 29:56.200]  And if you're a middle employee somewhere and you're worried, like, you know, what does that look like?
[29:56.260 --> 29:57.740]  What world can we create for that?
[29:58.080 --> 30:00.380]  Like, how do you advocate for the next middle employee?
[30:00.560 --> 30:01.700]  We have a little bit of time, I think.
[30:01.720 --> 30:11.700]  I don't know how long until this is all true, but I do think that the key is really about trying to help them through it.
[30:12.280 --> 30:13.980]  But I do think culture is the key.
[30:14.240 --> 30:17.760]  Like, you're at a place that shits on middle developers and junior engineers.
[30:18.080 --> 30:20.440]  Like, it's going to be shittier.
[30:21.140 --> 30:23.300]  Thank you for your question.
[30:26.370 --> 30:30.410]  My concern is AI-powered troll farms.
[30:30.590 --> 30:32.230]  We're seeing that more now.
[30:32.450 --> 30:46.030]  And it's particularly scary because it profiles people according to their interests or fears and basically gets greater penetration now more than the simple, like, Russian troll farms did in 2016.
[30:46.550 --> 30:47.550]  You know, it's...
[30:47.550 --> 30:52.590]  they're harder to spot and they're designed to really get into people's heads.
[30:52.790 --> 30:54.690]  Yeah, this is a really hard one.
[30:54.830 --> 30:56.770]  And has everyone here played with LLMs?
[30:59.970 --> 31:02.030]  It's going to be very hard.
[31:02.510 --> 31:07.750]  And I have a few friends, and I'm sure we all do, of people who are trying to think about how to solve this problem.
[31:08.610 --> 31:13.110]  And I was talking to a, interestingly, Japanese politician.
[31:13.430 --> 31:18.510]  And he said, we're spending all this time to tag LLM-generated text.
[31:18.750 --> 31:21.490]  Well, maybe we should spend more time about verifying humans.
[31:22.270 --> 31:24.890]  Because the LLM text is just... it's just out there.
[31:25.030 --> 31:26.290]  We're not going to... we can't pull that back.
[31:26.670 --> 31:38.450]  But verifying humanity, verifying, you know, whether it's, you know, verifiable credentials, whether it's, you know, I mean, even in the crypto space, there's all these things that people are doing to verify humanity that I think it will eventually come.
[31:38.610 --> 31:41.610]  And it's up to us to make that... to hasten that, to make that come faster.
[31:42.430 --> 31:45.450]  But you're... I mean, what your question is really onto something.
[31:45.570 --> 31:51.230]  I have friends that work in this space, and they're saying they could detect the networks in 2016, they could detect them in 2018.
[31:52.030 --> 31:54.450]  2023, no more detection.
[31:55.730 --> 32:00.390]  Just... I'm probably a minority opinion, but I don't consider LLM to be true AI.
[32:00.710 --> 32:02.350]  It has no mind.
[32:02.710 --> 32:04.030]  It has no concept.
[32:04.030 --> 32:21.530]  It's simply manipulating you know, symbols in order to put things together and making certain connections, but it's not... and I say that as somebody who's working in the AI field, if going back to the 80s, and this is not true AI, but it's industrially useful,
[32:21.770 --> 32:22.910]  but it has no mind.
[32:23.110 --> 32:24.790]  It has no theory of mind.
[32:24.930 --> 32:26.450]  I think that's pretty true.
[32:26.930 --> 32:29.510]  I think it's going to be hard to convince people of that.
[32:29.950 --> 32:32.370]  I say that as someone who studies this quite a bit.
[32:32.490 --> 32:35.310]  And one of the things I've noticed is all my friends that aren't in the space have named it.
[32:36.430 --> 32:46.950]  And so then you're like, okay, if you name it... and then I also have this theory, which I'm sure you can see very quickly, which is when you're chatting with your Excel spreadsheet or you're chatting with your email, as soon as you get bad information from that,
[32:47.170 --> 32:49.870]  they're going to be like, you got an email and it says you're fired or whatever.
[32:49.970 --> 32:51.610]  I'm just going to type, that sucks.
[32:52.050 --> 32:54.070]  And it's just going to be like, how does that make you feel?
[32:54.150 --> 32:56.230]  And suddenly, I have a therapist email client.
[32:57.130 --> 33:01.530]  And I think that it's very hard to stop these very human interactions.
[33:02.550 --> 33:05.570]  And I don't think we're thinking about it enough.
[33:06.270 --> 33:08.110]  We're spending a lot of time thinking about this.
[33:08.210 --> 33:13.250]  And to be honest, when I talk to VCs, sometimes us even thinking about the outcomes of that turns them off.
[33:14.350 --> 33:16.690]  But then there are some that are really good that don't.
[33:17.030 --> 33:26.350]  So, and hilariously, we raise money from like Bloomberg and Betaworks and a few others that are like really interested in these, like solving these humanist problems within the ML space.
[33:26.790 --> 33:30.130]  And this is also something that is deeply poisonous within our industry.
[33:30.350 --> 33:33.530]  Because like, if you work in ML, now you're no longer hip.
[33:34.370 --> 33:35.730]  It's literally the same.
[33:36.190 --> 33:39.790]  But you just have to change it to AI, and then you raise your salary a bunch.
[33:41.070 --> 33:52.790]  Yeah, I was going to say, just like when Eliza was invented in the AI lab at MIT, they started throwing people out of the lab so they could be alone to talk.
[33:52.790 --> 33:53.670]  Yeah, yeah.
[33:54.190 --> 33:55.830]  I do the same with ChatGPT.
[33:56.150 --> 33:57.910]  And that was just manipulating symbols.
[33:58.350 --> 33:58.830]  ChatGPT.
[33:59.150 --> 34:01.770]  Well, LLM is just that on steroids.
[34:02.090 --> 34:02.610]  Yeah, it is.
[34:02.750 --> 34:03.350]  Thank you so much.
[34:05.970 --> 34:08.570]  So, I am a high school teacher.
[34:08.570 --> 34:18.270]  And so I think a lot about how to inspire students and like young people who are looking into getting into CS and tech in general.
[34:18.270 --> 34:35.810]  And then as someone with like a philosophy and ethics background, I also think about like what kinds of ethical and like what, what like, besides just teaching someone how to code, what things to really bring up and try to build into a classroom.
[34:35.810 --> 34:57.330]  So I would be curious about your thoughts on kind of advice to students and things to focus on or think about as I'm discussing things with, you know, 14, 15, 16, 17 year old kids who are really trying to figure out what their future looks like and what this world is that feels really unstable.
[34:58.650 --> 35:04.870]  I, I don't, I have a young child, so I'm not, I'm thinking about this, but he's too young to like actually think about this a lot.
[35:05.030 --> 35:06.850]  But I've been talking to some teacher friends.
[35:07.010 --> 35:08.630]  I mean, there's two things that strike me.
[35:08.750 --> 35:12.850]  One is, how many of y'all got started as a hacker when you were that age?
[35:14.570 --> 35:16.850]  Like, there's obviously something happening.
[35:17.030 --> 35:21.830]  And one of the things that does happen is systems that are trying to stop the kids from doing something.
[35:22.830 --> 35:29.370]  And so like having a world in which you can try and figure out what's behind the door.
[35:30.430 --> 35:35.250]  All of my friends who, like I have so many friends whose Genesis point, you know, we're all old.
[35:35.410 --> 35:40.250]  So like the Genesis point is the nineties, but the Genesis point was like their parents taking away their keyboard.
[35:40.790 --> 35:49.170]  And so in AOL chats, they got stolen credit cards, bought a new keyboard and continued their whatever reign of terror on AOL chats.
[35:50.330 --> 35:54.950]  But like that kind of story is I think how a lot of us get started is trying to get around it.
[35:55.190 --> 36:04.930]  And I think we have such a world specifically within this stuff of only talking about the walls and the barriers, instead of talking about, okay, so this is a barrier.
[36:05.150 --> 36:09.250]  What do you need to do to get over the barrier, to get through the barrier, to get around the barrier?
[36:09.890 --> 36:15.630]  And I love when any of my friends are like, I finally put like, I locked down my kid's phone and I'm just like another hacker born.
[36:17.230 --> 36:18.230]  And I love this.
[36:18.270 --> 36:19.470]  And I think about this a lot.
[36:19.570 --> 36:28.250]  But then I think there's the other side of that, of just creating an environment where this, all this stuff, isn't just bad ends.
[36:28.710 --> 36:31.730]  So like Chachapiti, they obviously have.
[36:31.930 --> 36:38.330]  I did a talk at Northwestern in Chicago and I was just trying to get them to admit that they use Chachapiti and none of them would say anything in the class.
[36:38.510 --> 36:41.190]  And then afterwards, the teacher left and they're all like, we all do.
[36:41.370 --> 36:42.490]  Every one of them was like, yeah, I do.
[36:42.830 --> 36:43.650]  Do you have any hints?
[36:43.890 --> 36:48.830]  You know, and it's like this thing of like, I think it's like, it's like public health.
[36:48.970 --> 36:50.050]  It's like, they're going to do it.
[36:50.170 --> 36:52.710]  So let's, how do we, how do we help them do it safely?
[36:53.810 --> 36:55.110]  None of this is helpful, probably.
[36:55.230 --> 36:56.150]  You probably know all this.
[36:57.530 --> 36:59.090]  But I think that's kind of the idea.
[36:59.090 --> 37:00.510]  I don't know.
[37:01.190 --> 37:06.530]  Maybe we need a new hackers movie to get a new generation of people into Angelina Jolie and that other guy.
[37:06.610 --> 37:07.230]  You know, I don't know.
[37:07.850 --> 37:10.130]  I'm trying to think of all the things that like really got me in.
[37:10.150 --> 37:11.030]  It was such a different time.
[37:11.150 --> 37:13.250]  There was hope and everything was beautiful.
[37:15.350 --> 37:17.630]  I don't know, like tricky.
[37:18.190 --> 37:19.650]  Where are you a teacher at?
[37:19.910 --> 37:20.910]  Like in the city or?
[37:21.290 --> 37:21.650]  Massachusetts.
[37:22.090 --> 37:22.650]  Okay, Massachusetts.
[37:23.010 --> 37:23.130]  Yeah.
[37:23.290 --> 37:25.410]  I don't think I can show hackers to my kids.
[37:25.670 --> 37:26.150]  I don't think.
[37:26.150 --> 37:29.950]  I, I recently showed it to, to my team at work and it was a little awkward.
[37:30.130 --> 37:31.190]  I was like, I gotta go.
[37:31.330 --> 37:32.730]  You guys have to do this on your own time.
[37:33.250 --> 37:35.630]  But, but, but everyone looked really cool still.
[37:35.930 --> 37:36.730]  I was happy.
[37:37.670 --> 37:38.290]  Thank you.
[37:42.700 --> 37:43.000]  Hello.
[37:43.480 --> 37:47.780]  I would love to see hackers too with the original cast older now and just chilling.
[37:48.000 --> 37:48.680]  They'll have kids.
[37:48.940 --> 37:49.300]  That'd be a great.
[37:49.400 --> 37:50.020]  Yeah, that'd be great.
[37:50.340 --> 37:52.180]  They're just like, where you're not on life 360.
[37:52.840 --> 37:53.240]  Yes.
[37:54.420 --> 37:55.960]  So excellent presentation.
[37:56.340 --> 38:05.880]  Uh, you know, I'm reflecting on the topic of AI and the fact that I feel like a lot of things are becoming a slop factory right now in terms of social media and everything else.
[38:06.420 --> 38:08.100]  Let's say we can't stop it.
[38:08.220 --> 38:13.960]  Let's say every social media platform, newspaper, et cetera, it's all AI slop coming out.
[38:14.180 --> 38:19.640]  What do you see happening to the Internet then and how we interface with it as non slop lovers?
[38:19.980 --> 38:27.500]  I'm kind of not worried because in some regards, the Internet has always been slop.
[38:28.300 --> 38:29.520]  It's just slower.
[38:30.080 --> 38:35.980]  And so we, we routed around that by building communities that we appreciated and people that we love to interact with.
[38:36.540 --> 38:46.300]  And I mean, in the beginning, like when you first saw a corporate website, you were like, oh no, the corpos are here, right?
[38:46.400 --> 38:46.940]  You remember this?
[38:47.040 --> 38:48.780]  Like, like that was a horrible idea.
[38:48.960 --> 38:49.740]  Like, what are we going to wear?
[38:49.900 --> 38:51.240]  They're going to take away my mud.
[38:51.580 --> 38:53.360]  You know, like this type of this.
[38:53.520 --> 38:55.720]  I don't think, I think what we have now is an acceleration.
[38:56.500 --> 39:00.340]  And so I think what we have to do is use those same tools to accelerate the communities.
[39:00.960 --> 39:02.720]  So you see this actually a little bit already.
[39:02.880 --> 39:04.040]  Like, this is a blue sky link.
[39:04.160 --> 39:06.260]  There's a lot of really neat stuff that's being built on that proto.
[39:06.500 --> 39:08.700]  It is not being built by hands and fingers.
[39:09.340 --> 39:14.880]  People are vibe coding the shit out of that to build cool stuff so that they can, they can facilitate human to human conversation.
[39:14.880 --> 39:16.760]  Um, I think that's great.
[39:17.360 --> 39:22.780]  I also think that, um, parts of the Internet probably could use a little more slop.
[39:23.700 --> 39:27.840]  And so maybe it's time to fill up Twitter with some more swap, slop.
[39:28.160 --> 39:31.340]  And so we can have some networks of, of safety.
[39:32.220 --> 39:40.640]  Um, there's a great article that was, I think, from yesterday about how a lot of the misinformation farms are actually just from Bangalore, people making money off ads.
[39:41.040 --> 39:43.440]  And they're finding the Republicans to be very susceptible.
[39:44.440 --> 39:45.720]  And I think that's true.
[39:45.940 --> 39:47.180]  Um, I think that's great.
[39:47.560 --> 39:48.500]  Um, fuck them.
[39:48.680 --> 39:50.740]  But then there's the other side, which I think is really interesting.
[39:50.780 --> 39:55.980]  That is some of these LLMs, you can do a lot of work to, uh, uh, generate things that aren't slop.
[39:56.900 --> 40:00.540]  Um, and then there's questions of privacy, et cetera, that are, that go along with that.
[40:00.640 --> 40:05.180]  But I'm not super, super worried about this as a, as a, probably a slop generator myself.
[40:05.440 --> 40:11.920]  Um, but, um, it has accelerated everything in my life from a startups to information stuff.
[40:12.080 --> 40:20.880]  But I am seeing also, as probably we all are, um, I had to put a thing on my blog that says it was written by humans, because I didn't want people to think it wasn't.
[40:21.300 --> 40:24.620]  Um, I said it was 98% written by humans, because I felt bad about spellcheck.
[40:25.140 --> 40:29.060]  Um, but, uh, but like, but it, but it, but I, I find that happening a lot.
[40:29.200 --> 40:32.720]  Like we're starting to, like, like we talked about earlier, like verify humanness.
[40:33.480 --> 40:36.180]  And so like, fuck it, you're waiting through, through slop.
[40:36.760 --> 40:47.320]  But as long as you know, you have a space where you can verify humanness and take advantage of that humanity and also take advantage of the slop generated by those humans, then like, it seems like it's going to be okay.
[40:48.040 --> 40:49.460]  Um, we'll see.
[40:49.640 --> 40:52.540]  I'm not necessarily good at project, you know, predicting as we just saw.
[40:52.600 --> 40:57.060]  So maybe I'm wrong, but, uh, yeah, let's, let's, let's meet up next year and see how it goes.
[41:00.620 --> 41:04.500]  Um, you had a slide that said in 2017, we're going to replace the truckers.
[41:04.620 --> 41:05.760]  We're going to replace the warehouse workers.
[41:05.960 --> 41:11.200]  But last I checked, the guy drove my groceries here and my Amazon warehouses are full of minimum wage employees.
[41:11.380 --> 41:12.420]  This is my favorite thing.
[41:12.420 --> 41:17.780]  I love this so much because people were paid so much money to stand on stages like this and be like, in 2017, we will do.
[41:17.880 --> 41:19.460]  And then they're just like, they were just so wrong.
[41:20.380 --> 41:21.860]  Well, yeah, we get to look back on that.
[41:22.000 --> 41:25.480]  And I get that you said 90% of your code is AI generated.
[41:25.480 --> 41:30.980]  And your friend said in three years, we predict we won't need the senior developers, but where is that optimism?
[41:31.080 --> 41:36.520]  I feel like most of it's overhyped because while I do use LMS all the time, I do find efficiencies grab from it.
[41:36.620 --> 41:43.240]  I don't, I feel like more of the hype is to raise capital so they can keep burning $5 billion a year, uh, and losing it.
[41:43.360 --> 41:46.280]  But where do you think we'll actually have in three years?
[41:46.380 --> 41:48.880]  How do we balance that optimism and where is your head at on that?
[41:48.880 --> 41:51.540]  So my, my canary in the coal mine is a weird one.
[41:51.760 --> 41:55.380]  And it is, uh, the big four consulting firms.
[41:56.320 --> 42:03.520]  And the reason is, is because they're very, I don't know a technology company, maybe y'all know one that's different, that understands the amount of revenue per employee they get.
[42:04.500 --> 42:12.880]  Whereas the big four consulting firms know 100% how much every single employee from an admin all the way to their partners generates for their organization.
[42:13.420 --> 42:18.280]  And the moment that they stop hiring or layoff mid-level, that's when I know I'm right.
[42:19.120 --> 42:27.200]  So about six months ago, if you went to the Deloitte subreddit, or you went to the ENY subreddit or any of these big consulting firms, it was new hires being like, where should I work?
[42:27.280 --> 42:28.260]  It's super exciting to be here.
[42:28.440 --> 42:31.900]  People getting promoted to more senior positions and mid-level getting laid off.
[42:32.860 --> 42:40.500]  And I think, I think the reason for that is because, well, Doge took away a huge contract for, you know, all of them.
[42:40.800 --> 42:49.000]  But the second reason is I think that they have found that they can get the same work done with better tooling and, um, less people.
[42:49.200 --> 42:50.980]  But this didn't start with AI, right?
[42:51.100 --> 43:00.240]  This started, I mean, I remember in 2001, when I got my first Java job, I had a, there was like a sea of cubicles as far as the eye could see, you know, and Solaris boxes stacked to the ceiling.
[43:01.020 --> 43:04.280]  And, um, that then turned into 20 people.
[43:04.280 --> 43:05.480]  And less boxes.
[43:06.040 --> 43:08.300]  And then that turned into 10 people and Linux.
[43:08.660 --> 43:09.940]  And then that, you know, so on and so forth.
[43:10.040 --> 43:20.520]  So it's not, I don't think this is just like a suddenly drop, but, but when you look at people who understand the revenue per employee, and when they start getting rid of employees, that means that the revenue story is there.
[43:20.900 --> 43:22.200]  And that's what scares me.
[43:22.700 --> 43:26.580]  The story of like, we can, like I can code with my robots.
[43:26.740 --> 43:27.140]  That's cool.
[43:27.860 --> 43:29.100]  Like, I think that's neat.
[43:29.280 --> 43:32.080]  Um, but that's not, like a startup doesn't have that same story.
[43:32.180 --> 43:35.100]  Like we could hire 10 people and have the same output as four.
[43:35.300 --> 43:35.540]  Like, great.
[43:35.700 --> 43:36.100]  Who cares?
[43:36.200 --> 43:38.220]  As a startup, like we, we raised fantasy money.
[43:38.580 --> 43:42.480]  Um, but when you have people who actually have real businesses thinking about this stuff, that's when it gets scary.
[43:42.620 --> 43:44.700]  And I think we're on the cusp of that, but that's why I watch.
[43:44.800 --> 43:46.420]  And I watch it via the subreddits.
[43:46.540 --> 43:47.900]  It's actually a really interesting signal.
[43:48.300 --> 43:48.980]  Is that helpful?
[43:49.180 --> 43:49.540]  Yeah.
[43:50.120 --> 43:50.320]  Thank you.
[43:53.120 --> 44:00.220]  So one thing I've started to see more is, uh, employers mandating the use of AI in workflows.
[44:00.540 --> 44:12.760]  And, um, the thing that I have a hard time with is kind of encountering the, the bullshit asymmetry principle, where it takes an order of magnitude more effort to correct mis or disinformation.
[44:13.160 --> 44:18.700]  Um, and that's ultimately resulted in my withdrawal from online discourse.
[44:19.060 --> 44:24.900]  Um, how do you combat that or see that playing out in workplaces?
[44:25.720 --> 44:29.780]  That is, that is, you've echoed 100% my perspective, sadly.
[44:30.040 --> 44:33.080]  I also have withdrawn, for the most part, from online discourse.
[44:33.080 --> 44:35.940]  All I do is consume memes and share them to private group chats.
[44:36.400 --> 44:38.020]  Um, I no longer participate.
[44:38.180 --> 44:40.460]  And like Twitter, I was such a huge Twitter fan.
[44:40.480 --> 44:41.200]  I love Twitter.
[44:41.340 --> 44:41.780]  I adored it.
[44:41.860 --> 44:46.600]  I had such, I met so many great friends and really, like really, truly good friends in that experience.
[44:46.600 --> 44:47.440]  And it's just gone.
[44:47.720 --> 44:52.900]  I haven't found the same on some of these other blue guys seems too close, but still, I don't know.
[44:53.140 --> 44:54.400]  I've been trying to figure this out.
[44:54.520 --> 45:11.660]  Um, I, I miss the days of yore where you had like public IRC channels, but I can't tell if that's just because I'm a boomer and I feel nostalgic every once in a while for, uh, poorly ran Linux boxes, or if it's just that we, um, we, we need those spaces again.
[45:11.820 --> 45:14.320]  Um, but I really believe and I truly believe this.
[45:14.360 --> 45:27.980]  And I, I'm so happy when I walk through the halls and see some of this happening, whether it's the D web stuff or the mesh New York stuff or whatever it might be of creating some of these experiences that aren't rife with, um, people trying to get views or trying to get whatever.
[45:28.100 --> 45:30.260]  And it doesn't matter your political belief where you are.
[45:30.400 --> 45:38.380]  It just feels so good to be in this space that is like, quote unquote, a safe space to achieve the things that you want, which sometimes for me is just hanging out.
[45:39.180 --> 45:41.560]  Like, I just want to have friends and talk about things that are normal.
[45:41.700 --> 45:44.680]  Like, what is it like, you know, why was a hacker movie so weird?
[45:44.800 --> 45:45.500]  Like, or whatever.
[45:45.500 --> 45:51.300]  I didn't need to, I didn't, I don't need to have something that is like always, I don't want to always have to present online.
[45:51.300 --> 45:52.860]  And I feel like that's part of it as well.
[45:53.120 --> 45:54.940]  Um, but I think we have to create the spaces.
[45:55.520 --> 45:56.440]  And I, I don't know.
[45:56.520 --> 45:58.500]  Like, I, I, I don't know.
[45:58.640 --> 46:01.120]  I'm, I'm, I'm, is everyone stuck in group chats?
[46:02.040 --> 46:02.880]  Is that just me?
[46:03.080 --> 46:03.300]  Yeah.
[46:03.360 --> 46:04.840]  I feel like we're all stuck in group chats.
[46:05.580 --> 46:06.280]  So I don't know.
[46:06.380 --> 46:16.200]  It's, you know, maybe we just need to publish the group chats out, like have a bot that just spits I actually have that.
[46:16.300 --> 46:19.540]  If you can find the Tumblr or the group chat, I'll invite you to it.
[46:19.940 --> 46:20.080]  Okay.
[46:21.020 --> 46:21.460]  Maybe...
[46:21.460 --> 46:24.180]  We've got some questions coming in from online as well.
[46:24.460 --> 46:27.220]  So we can't ignore our virtual attendees, of course.
[46:27.940 --> 46:28.640]  A couple of them.
[46:28.780 --> 46:30.620]  Hopefully we have time for a few.
[46:31.080 --> 46:39.320]  So question is, why is so much effort put into startups like Uber that, quote, disrupt industries that are already struggling, like taxis?
[46:39.800 --> 46:43.860]  And not much effort into startups to disrupt industries like oil companies.
[46:44.220 --> 46:50.200]  It seems like Silicon Valley congratulates themselves a lot for not actually doing anything that challenging.
[46:51.320 --> 46:53.000]  That's a really, really, really good question.
[46:53.120 --> 46:53.540]  I love it.
[46:56.060 --> 47:00.920]  As someone who has participated in Silicon Valley culture for a really long time, I think it's just because they're very boring people.
[47:01.680 --> 47:02.840]  I think that's it.
[47:02.960 --> 47:06.340]  I think that they are often people who don't... Like we live in cities.
[47:06.500 --> 47:07.760]  I'm sure how many people here live in a city?
[47:08.340 --> 47:09.540]  Like cities are pretty cool.
[47:09.660 --> 47:13.520]  And the thing about cities, you kind of have to figure out how to live with one another and interact with one another.
[47:13.680 --> 47:16.340]  And like you're not scared of people on the street.
[47:16.520 --> 47:17.480]  You just are like, yo, what's up?
[47:17.540 --> 47:18.060]  And then that's it.
[47:18.140 --> 47:18.820]  And you go on your way.
[47:19.140 --> 47:27.240]  And so I think that people from cities, because of that density, et cetera, are going to build things that are a little more focused on not disrupting the city and the harmony, et cetera.
[47:27.240 --> 47:28.380]  That's not always true.
[47:28.760 --> 47:29.780]  You have the finance people.
[47:29.960 --> 47:33.740]  But there are, you know, there are a lot of people who are building interesting things there.
[47:34.200 --> 47:39.980]  But I think the other problem is, is that there's just a general lack of imagination and hard things are hard.
[47:40.360 --> 47:43.020]  So I have a few friends that are focusing on energy.
[47:43.480 --> 47:46.140]  And they have been working on the same problem for five or six years.
[47:46.280 --> 47:48.280]  And they struggle to raise money because it's just hard.
[47:48.640 --> 47:48.980]  Science.
[47:49.140 --> 47:49.740]  It's hard science.
[47:49.880 --> 47:50.620]  Deep tech is hard.
[47:51.020 --> 47:57.240]  It's super easy to build an app that like, you know, you sell to taxi drivers or whatever to like create Uber.
[47:57.520 --> 47:59.060]  That seems to be much easier.
[47:59.200 --> 48:00.980]  Now, the hard thing about that is growth and all that other stuff.
[48:01.120 --> 48:01.680]  They're very smart.
[48:02.480 --> 48:04.640]  But I just think it's, I don't know.
[48:05.100 --> 48:12.400]  I think about this a lot when we build companies because we don't want to build something that is small.
[48:13.040 --> 48:13.340]  Right?
[48:13.480 --> 48:15.800]  We still want to do something that feels good to be a part of a big team.
[48:15.940 --> 48:19.000]  But also, I don't want to build something that is just destroying.
[48:19.420 --> 48:20.440]  And it's just a hard thing.
[48:20.520 --> 48:21.020]  I don't know.
[48:21.120 --> 48:24.240]  I think the problem is, and I once talked to this young person about this.
[48:24.620 --> 48:26.840]  He was like, I'm just very disappointed in all the jobs.
[48:26.840 --> 48:30.340]  And I was, we talked for about an hour and I was like, I think your problem is with capitalism.
[48:31.300 --> 48:33.100]  I don't think it's an employment problem.
[48:33.240 --> 48:34.900]  I think the issue here is capitalism.
[48:34.900 --> 48:37.380]  And I think that's, it's like water, right?
[48:38.640 --> 48:39.020]  Yeah.
[48:39.380 --> 48:39.600]  Yeah.
[48:39.780 --> 48:40.480]  Great, great answer.
[48:40.920 --> 48:42.300]  Another question from the audience.
[48:42.440 --> 48:44.260]  In fact, this was the first question that was asked.
[48:45.140 --> 48:46.420]  Are you an LLM?
[48:47.040 --> 48:52.280]  You know, I sometimes wish I was, it would just be easier just to just buy GPUs, increase my capacity.
[48:52.280 --> 48:58.660]  I am pretty sure my context window is much smaller than anything you can run on a standard Mac today.
[48:59.300 --> 49:04.880]  Probably a little under a thousand tokens, because I have a real problem with recency bias.
[49:06.520 --> 49:09.920]  And a more pointed question here.
[49:10.420 --> 49:12.880]  You've spoken about the companies you founded.
[49:12.880 --> 49:19.260]  What are you personally doing as a senior leader to address the issues with junior and mid-level developers within your own startups?
[49:20.900 --> 49:23.080]  So we're, we're kind of experimenting right now.
[49:23.840 --> 49:25.020]  I'm trying to figure this out.
[49:25.120 --> 49:26.860]  I'm trying to do it by actually doing it.
[49:28.020 --> 49:35.360]  We've always ran a little bit of like more of a studio kind of environment in Chicago, where we will, anyone who kind of wants to come hang out, can just come hang out.
[49:35.460 --> 49:41.940]  And if they just happen to accidentally join the team, basically by being there all the time, then they just kind of join the team by being there all the time.
[49:42.500 --> 49:44.660]  And of course, like this is, we pay people, etc.
[49:44.940 --> 49:46.900]  And we, we've been experimenting with interns.
[49:48.000 --> 49:51.040]  Most people that we have with our interns are not necessarily junior, junior, junior.
[49:51.220 --> 49:53.000]  They're not just fresh out of undergrad.
[49:54.060 --> 49:57.620]  And I had an intern when I was mid, I was an intern when I was mid-career.
[49:58.540 --> 50:02.220]  And it kind of changed my entire trajectory of my entire career.
[50:02.560 --> 50:06.600]  And so I think that the language of intern kind of is bad language.
[50:06.800 --> 50:13.220]  But I think we could change it to apprenticeship or something else to try and bring some people who are just trying to figure out how to get into that role, how to get into that thing.
[50:13.320 --> 50:14.440]  The other thing is, we work in startups.
[50:14.620 --> 50:15.700]  No one should ever work in startups.
[50:15.820 --> 50:16.400]  Do not do this.
[50:16.940 --> 50:20.180]  It's like that scene in Oh Brother, We're Out That, where it's like, do not seek the treasure.
[50:20.920 --> 50:21.540]  It's bad.
[50:22.120 --> 50:27.180]  But if you do choose that because it is a compulsion and you have problems, then, you know, come talk to me later.
[50:28.420 --> 50:29.300]  But yeah, I don't know.
[50:29.440 --> 50:34.400]  I don't I really wish I had a better idea of how to solve this because I do worry about it.
[50:34.420 --> 50:40.120]  And I remember the time fondly when I had the senior people who were helping me up through that career ladder.
[50:41.280 --> 50:42.900]  And yeah, I don't know.
[50:46.200 --> 50:51.240]  Well, I think we're just about out of time unless there's perhaps maybe one more question or so.
[50:51.380 --> 50:55.800]  We might have time for since we're a little bit over, but it's a very popular session.
[50:55.800 --> 50:57.080]  Well, thank you so much.
[50:57.160 --> 50:58.780]  This has been really, really nice to chat.
[50:59.040 --> 51:01.040]  Yes, let's hear it for Harper.
[51:01.600 --> 51:02.140]  Thank you.