Subscribe
Share
Share
Embed
NET Society is unraveling the latest in digital art, crypto, AI, and tech. Join us for fresh insights and bold perspectives as we tap into wild, thought-provoking conversations. By: Derek Edwards (glitch marfa / collab+currency), Chris Furlong (starholder, LAO + Flamingo DAO), and Aaaron Wright & Priyanka Desai (Tribute Labs)
00;00;16;06 - 00;00;28;18
Aaron
How cool is that Midjourney device, by the way? Midjourney is such an interesting project. I saw somebody muse that its founder. His name escapes me. Is like the Nikola Tesla of our time. What do you guys think of it?
00;00;28;19 - 00;00;55;28
Chris
Well, a like it's a curve ball from left field. That was I don't think was on anyone's bingo cards. I think in hindsight it it makes sense. But like, wow. You know it's nice to be like genuinely surprised to the upside at someone, you know, applying the hard problems of slop and, you know, turning that into medical device and starting to give us the future I think everyone's excited about.
00;00;55;29 - 00;01;23;19
Derek
Just to maybe just to level set the the kind of just the Midjourney release. And I think people have already demoed it like they had this demo day announcement. But basically it's like a ring. You stand in water, a platform kind of lowers you progressively through the ring, and there's like hundreds of thousands of, like these tiny, tiny sensors that are basically just sending and receiving sound waves from like 360 degrees around your body.
00;01;23;19 - 00;02;04;03
Derek
And then in basically one minute, they're throwing a huge amount of like compute to reconstruct, like all the acoustics that are bouncing back, that data that's bouncing back into kind of like a high resolution 3D map of your entire body. And so, yeah, for anyone that's I'm pretty sure it's like almost everyone at this point. But like for anyone that's ever gone through like an MRI or like a CT scan, basically the idea is that what they're trying to accomplish is like whole body imaging without any radiation, without any magnets, without waiting inside of a tube for 35 minutes while getting a more granular look at like, things that typically wouldn't even appear.
00;02;04;03 - 00;02;21;06
Derek
So like organs, tissues, vessels, bones, muscle, fat, like all of the things that make up your body being mapped in real time in like a very comfortable setting. And I think people were just like, where the fuck did this come from? And I think that's kind of the enthusiasm that we've seen on the internet.
00;02;21;08 - 00;02;44;03
Chris
Right? And so just to like, really, really ground how wild things have come. My dad was a nuclear medicine technician and he worked for the VA hospital. And, you know, this was back in the 80s. And what he did for a living was he injected isotopes in veterans, waited for it to circulate through their body, and then took pictures of them.
00;02;44;03 - 00;03;02;18
Chris
And so that's how they used to do this once upon a time. And like they used to because, like, they were nuclear isotopes and used to like, store it in like they called the ion room, where they literally like, it took the magazine from an old battleship because it was all lead coated and like, that's where all this stuff lived.
00;03;02;18 - 00;03;17;01
Chris
And so now we've got high end spas in which you're getting an acoustic bath that can do this at, like, you know, 10,000 times better, faster. So, you know, great job technology.
00;03;17;03 - 00;03;35;15
Aaron
I just think the story of Midjourney, as best I understand it, maybe you guys know it more is just so interesting, right? It started with a group of friends and a discord chat. Right? They were kind of geeked on using, you know, AI tools to generate images. And, and now they're kind of moving into medical devices. It's almost like that.
00;03;35;16 - 00;03;35;29
Aaron
Like.
00;03;36;00 - 00;03;37;10
Pri
Yeah, it does feel like that.
00;03;37;11 - 00;03;39;10
Aaron
I don't think they raised outside capital.
00;03;39;13 - 00;03;49;24
Pri
That's what I was just going to add. Like didn't aren't they like kind of completely bootstrap. And that's kind of the reason that they're able to do this. I mean I read that on Twitter, but I actually don't know if that's 100% true.
00;03;49;25 - 00;03;57;09
Derek
Yeah. I'm not I'm not actually sure I don't I don't yeah, I don't know the answer to that. I mean, bootstrapped with revenue from a journey. Would that be the idea?
00;03;57;12 - 00;04;12;22
Pri
Yeah. Like, yeah I think I think they kind of bootstrapped the whole thing. And like part of the reason everyone like is kind of excited about this story is that in a way, it's like he's able to kind of just totally do this pivot because of the fact that he wasn't venture backed, I guess, like that's what people are trying to say.
00;04;12;22 - 00;04;23;19
Pri
I don't know if that's actually the case. Like maybe he would have done this anyway with venture backing, but it is a very kind of amazing story because he's just like a CEO of conviction is like doing whatever he wants.
00;04;23;20 - 00;04;46;18
Derek
I definitely think there's something psychologically to like when you raise venture for to point those capital dollars at something very discreet, and then you're not for whatever the market turns or the product doesn't quite work. Or like the revenues can't support a venture style business. It really is kind of like, yeah, it's it's deflated or it's just like you set out to do this mission.
00;04;46;18 - 00;05;06;15
Derek
You didn't accomplish it. And so like the tendency is to, of course, is pivots like founders pivot all the time. But like the tendency is like to wear that capital in a way that makes you like either want to return it back to your venture investors or to maybe take a swing that looks like not like is adjacent enough to like, you know, to use some of the expertise that you've built.
00;05;06;15 - 00;05;27;09
Derek
And I think what you're flagging pre and Aaron is because these guys originated from like more of a tinkerer, a space of tinkering. They felt like they had the mental plasticity to just say like, nope, that the revenues don't support a venture business here. Moving on to the next thing, let's huddle up and like, hack on something fun and new and like it ended up being preventative medicine.
00;05;27;09 - 00;05;38;09
Derek
And I think that's that maybe comes with the freedom of not feeling like you have to respond or answer to a third party, but like you're able to kind of just, you know, follow your, your whimsy.
00;05;38;10 - 00;05;39;05
Pri
Exactly.
00;05;39;05 - 00;05;58;23
Aaron
I wonder if there'll be more stories like this just because it feels like with all these new AI tools, you should be able to kind of replicate the Midjourney story a little bit more, right? If you're geeked on a specific topic, you've got like these powerful tools that your disposal like you can the like the barrier to like build something is a little bit lower.
00;05;58;24 - 00;06;18;17
Aaron
And maybe you do find like, you know, that quick product market fit. And I don't know if it was quick or that quick. And and Midjourney is in major these case, but I don't know. I just think it's an interesting story and definitely like a contrast. Like I think you guys were saying to some of the like hyper venture scale backed AI companies.
00;06;18;19 - 00;06;42;25
Pri
Yeah, I'm personally super excited about it, though. I mean, the fact that it's an alternative to a CT scan and like, it's my mom literally got a CT scan last week, but like, didn't want to do it because she was like, she's like, I don't want to do it too much because of all the radiation. Like, this feels like a very, very I mean, even that innovation alone, like the lack of radiation order to get this full body scan, that alone is like going to motivate so many more people to want to get these scans.
00;06;42;26 - 00;06;53;21
Pri
Like, I think that that is actually like low key, huge innovation. I mean, granted, it's not like per se AI. I don't know how much AI is, actually. I need to look more into that.
00;06;53;21 - 00;07;07;19
Derek
I think it's a lot. I think it's a lot, a lot of the only way you can do the reconstruction is by throwing like a ton of compute at it and, you know, allowing AI to kind of work its way through the data. That makes sense. Reconstruction. Yeah.
00;07;07;22 - 00;07;10;17
Chris
Yeah. I mean, they're capturing 40GB a second.
00;07;10;19 - 00;07;11;22
Pri
Oh wow. Okay.
00;07;11;23 - 00;07;16;23
Aaron
Oh, wow. Yeah. How do you do that? That's what I don't understand.
00;07;16;26 - 00;07;33;19
Chris
Right. I was just about to say I wonder how quickly they're able to process that, you know, is this like, you're in the thing for a minute and then, you know, a server farm is spending 12 hours deconstructing that and turning it into a scan, I don't know.
00;07;33;20 - 00;08;04;24
Aaron
Have you seen those projects where they're like mapping, like brains of animals? I happen to bump into, like a team that's doing it. It's kind of like insane technology where they can, like, literally upload like a brain. At this point, it's like a fruit fly. They're working towards a mouse. But I was talking to the team or one of the teams that's building that space, and they said that actually, you know, file storage is their biggest constraint and not compute like the computer, not the issue, just processing, you know, this massive amount of, you know, image data, right?
00;08;04;25 - 00;08;12;00
Aaron
Which is kind of like a video or a scan. Right? Chris. So I wonder if that's like the next big constraint. It's actually like data storage.
00;08;12;03 - 00;08;13;19
Pri
Data data organization.
00;08;13;19 - 00;08;24;14
Derek
And I just saw I just saw hundreds of thousands of underwater file coin holders just rise up from the graves. When you said that, Aaron. Well.
00;08;24;16 - 00;08;50;24
Aaron
Like sending them an olive branch, like maybe they weren't wrong, right? They got that option when I talked to the team. Yeah. Or even for some of some of these bits, like if everybody's generating, you know, 40 100 gigabyte related images. Right. It's not like the scanning and image quality is not going to get better. Like where does that all get stored, especially if there's different memory or other constraints caused by AI to.
00;08;50;25 - 00;08;52;03
Chris
Hear me out, guys.
00;08;52;04 - 00;08;53;01
Derek
Yeah. Go ahead Chris.
00;08;53;02 - 00;08;57;08
Chris
Orbital file coin.
00;08;57;10 - 00;09;01;13
Chris
That's what's going to save their bag. Just putting file in space.
00;09;01;14 - 00;09;20;26
Derek
Just frontier tech vertical on top of frontier tech so that you create the ultimate frontier tech. I love it I was just going to say the you know, it really does. Like I think maybe part of the through line across all the stuff is that old expression like what gets measured, get managed. And in a lot of ways, a lot of the innovation here is just measurement technology.
00;09;20;26 - 00;09;46;28
Derek
It's like we're actually able to measure things. Previously we weren't able to measure, and then once we're able to measure it, we can make it digitized and put that digital information into an environment that can be referenced by these, these synthetic intelligent systems, and they can get optimized as time goes on. Right? Like storage get solved or like latency get solved, or the computation to apply gets weighted differently to make these things much more tailored for the cost.
00;09;46;28 - 00;10;10;09
Derek
And it's like it just becomes optimization games at that point. Once like the 0 to 1 moment around this, whatever this measurement technology happens. And I think that's the exciting part where it's like you're at the the start of like that hockey stick right now where like people are like, oh, let's measure this thing using this process. And you can apply that for whatever, lots of different categories of value.
00;10;10;14 - 00;10;21;10
Derek
I think health is one that we've just basically been blindly operating in the dark for whatever it feels still like the it still feels like, you know, prehistoric era around like.
00;10;21;12 - 00;10;22;26
Aaron
The time. Right, Derek.
00;10;22;29 - 00;10;24;23
Derek
Since leaving of time. Exactly. Like when.
00;10;24;23 - 00;10;25;18
Aaron
We're just like.
00;10;25;20 - 00;10;47;24
Derek
What exactly? It's all trial and error and like, you know, we're we're bumbling idiots trying to figure out how to, like, heal and cure each other. And we're taking scaffolds to people's bodies to like, you know, for solutions that honestly, probably can be had in much more elegant fashions. And so, yeah, anyway, I'm really, really optimistic about all of this stuff.
00;10;47;24 - 00;10;55;21
Derek
And to me, it's like the common denominator is just like better measurement and being digitized and being put into these intelligent systems.
00;10;55;22 - 00;11;08;02
Chris
Okay. We're going to take a little diversion through history for a second. When do you think the concept of the second was introduced? Talking about time here, people.
00;11;08;05 - 00;11;11;25
Derek
Wow. I have no idea. What's the answer to this one?
00;11;11;27 - 00;11;13;18
Chris
The Babylonians.
00;11;13;20 - 00;11;15;16
Aaron
Oh man. What was going to guess? I was.
00;11;15;16 - 00;11;18;24
Chris
Going to get. Sorry, Aaron. Don't worry. We got a whole history to get through here.
00;11;18;26 - 00;11;21;23
Aaron
Well, I would have guessed wrong, so please go ahead. Chris.
00;11;21;24 - 00;11;38;06
Chris
Okay. The mathematical origin apparently, was the Babylonians in 2000 BCE. The theoretical concept was the Greeks in 150. All right. Now I'll let you play a little guessing. What do you think? We had the first mechanical clocks that could measure seconds.
00;11;38;07 - 00;11;39;19
Aaron
Had it be enlightenment era?
00;11;39;25 - 00;11;46;02
Chris
16th century? Yeah. All right, now, here's the big one. When did we get the millisecond?
00;11;46;03 - 00;11;47;28
Aaron
I'm going to say.
00;11;48;01 - 00;11;49;11
Derek
1800s early.
00;11;49;13 - 00;11;52;07
Aaron
Yeah, early 1800s. That's what I was going to go. Go for.
00;11;52;08 - 00;11;59;28
Chris
Yes. You didn't need the idea of milliseconds until the Telegraph came along and they sorted it out in the 1850s.
00;12;00;07 - 00;12;05;02
Derek
This is a great. I already know where Chris is going with this, and this is an awesome analogy.
00;12;05;03 - 00;12;06;15
Aaron
Where are we going to.
00;12;06;18 - 00;12;08;08
Chris
Take the baton, Derek?
00;12;08;10 - 00;12;40;24
Derek
Chris is basically making the argument that as you bump up into better and better, better and better and better technology, the idea of more granular assigning definitions or concepts or ideas to better measurement just happens naturally and evolves. And so like we're currently in whatever, to use his time analogy, the millisecond era. But what we're quickly going to be bumping up to with better measurement tools is products and services that are going to require new definitions for the things that we don't quite need to quantify at.
00;12;40;25 - 00;12;42;12
Derek
Is that is that right? Chris.
00;12;42;13 - 00;13;03;23
Chris
Yeah. Like what's the term for I get a process 40GB a second. Yeah. I mean, you know, but that's like kind of the exciting stuff that, you know, we're just starting to see here. Right. Is that problems create solution solutions, open new or I guess problems create new language. New language opens new doors.
00;13;03;25 - 00;13;25;15
Derek
Yeah, it's a cool line of thinking for sure. It's a very cool thing. I'll say maybe one other thing on the bath technology before we move on. I really, I think pre was kind of touching on this, but there's something about preventative medicine that just has felt like I don't even know, pseudoscience. Like it's like hasn't really been taken seriously by like the medical profession.
00;13;25;15 - 00;13;54;16
Derek
And I do think so much of what of, of you know, of people of longevity happens before you even enter a hospital room. I don't even think that's a crazy thing to say anymore. There's only so much that can happen once you end up on a hospital bed. And this idea of normalizing technology to help with preventative care inside of day to day experiences, like a sauna or a sound bath or wherever you may be, and just pop in there and get great diagnostics in a setting that's comfortable.
00;13;54;16 - 00;14;16;11
Derek
I feel like we'll do so much culturally around longevity that it's really difficult to even it's it's difficult to see its impact right now, but I think it's going to be such an important tailwind towards human longevity that I, I don't even think I fully appreciate it, but I definitely don't think most people haven't fully appreciated yet. That feels like this product is starting to kind of get at.
00;14;16;12 - 00;14;17;23
Derek
Do you guys feel the same way?
00;14;17;29 - 00;14;47;12
Chris
Yeah, I mean, everything's a hassle, right? When it comes to medical, you know, diagnostics or just interacting with the healthcare system. And so yeah, if you can just be like, hey, let's, let's hit the spa and, you know, for next year or 100 bucks or whatever you're walking out with, you know, something that would have required approvals, this, that and the other thing and can, you know, help your physician, you know, cut a real problem off at the past.
00;14;47;19 - 00;14;49;02
Chris
That's fantastic.
00;14;49;05 - 00;15;19;17
Aaron
It's just deflation, right? Like it's just things get cheaper. I guess the one second order effect of that is, you know, maybe that causes paranoia and some people or like over, over diagnosing certain certain elements, which I guess is always a risk. But I think, I mean, it's amazing. The journey story is amazing. I think that's why I kind of flooded everybody's timeline, you know, and even over the past couple of weeks, and we kind of see it like around eight, and it just feels like space is really hitting like an inflection point too.
00;15;19;18 - 00;15;38;19
Aaron
And obviously space is causing a lot of excitement, however you feel about its leader, Elon. But it does feel like that's like pretty real too, which is not something I would have thought about 2 or 3 years ago. Yeah, there's a lot of really interesting innovation that's happening on that front. Is that capturing your attention?
00;15;38;20 - 00;15;48;13
Chris
Well, I mean, my attention apparently is taken up with tub based technology because this is like the second episode in a row. We've talked about it, but space is the place. What if we put tubs in space? Man.
00;15;48;15 - 00;16;10;09
Pri
We definitely will. A little spatial bubble bath. But one thing I wanted to talk about just before we wrap up the Midjourney section. So thinking about this, I think a lot about just like the AI wealth effect that's happening, and I guess that does feed into the space IPO and the number of like millionaires and billionaires and billionaires and all these people that are newly minted.
00;16;10;09 - 00;16;31;11
Pri
It feels a little bit like the crypto wealth effect, but the scale is just way bigger. There's going to be a tremendous amount of capital and like they're going to want to donate, you know, pursue things of interest. To me, it feels like that cohort feels very interested in like longevity, preventative medicine, health care, novel. Like, you know, I'm talking like an artificial womb company next week.
00;16;31;12 - 00;16;58;07
Pri
Like they're interested in the human existence. Maybe as a counter to AI. I do wonder, you know, typically wealthy patrons like you go back to the robber barons, Carnegie's Rockefeller like they tended to give to like, books, culture, libraries, art, commissioning artists, even the Medicis, like they were giving, you know, money to fund work for the church, or they were giving money to the church or whatever.
00;16;58;07 - 00;17;23;27
Pri
I do wonder if that class of people ends up just deploying their excess capital and things like this, as opposed to culture, art and other things. Or maybe there's just so much capital that it's all going to trickle in to different areas. But I just thought like the Midjourney thing was like emblematic of the fact that, like, okay, I have this capital, I'm going to explore what I'm interested in, which is like preventative medicine and do open up the sauna.
00;17;23;27 - 00;17;33;07
Pri
But like, I'm curious, that sort of applies across the board. I don't know if you guys have thought about that at all, but yeah, I just wanted to put that out there to get thoughts.
00;17;33;11 - 00;18;02;19
Chris
Well, one thing is the robber barons were great builders of sports and so they got everywhere. And, you know, they they loved their Saratoga Springs and, you know, building giant five star hotels out by good thermals. So that is one way I guess this generation is mirroring the the wealth effect of the past. But I don't know where else you think the money is going to go, or do you think it's just going to stay here.
00;18;02;19 - 00;18;14;00
Chris
And then as an aside for for Derek, after you answer that one, the idea of like saunas and cold plunges being venture back categories is absolutely wild to me.
00;18;14;03 - 00;18;19;05
Aaron
Yeah. I mean but it is wild. There's just been some rounds around that, you know, for what it's worth, I think it's going to go everywhere.
00;18;19;07 - 00;18;50;03
Derek
I think Aaron's right. I think I think the, the story around that Chris is like, look, I think the medical sales, the medical business, preventative care, like longevity tech, like this stuff has literally been up only on like a 24 month run. And I do think that it's convergence with analytics and AI and better preventative care and keeping people out of hospitals is going to be such a massive boom for a number of different.
00;18;50;09 - 00;19;20;02
Derek
I mean, on like on both the health dimension for people, on both, you know, there's going to be massive winners for product building around this category. There's going to be software products, there's going to be hardware products. And I think investors are seeing health tech is kind of like just bio broadly and like whatever cancer drugs and clinical trials and wet labs and on top of AI companies, but also kind of consumer tech around preventative care as being one larger trade over the next 5 to 10 years, like modernizing all of this stuff.
00;19;20;02 - 00;19;43;06
Derek
And so, yeah, it's like the whatever the sound bath or the, you know, the hot tub or the cold plunge is really just the loss leader into kind of longevity and longevity data. And yeah, products that people want to engage with that help I think about like the eight sleep or the aura ring or, you know, the Fitbit or, you know, Strava.
00;19;43;07 - 00;20;03;22
Derek
It's like this idea of like digitization of the things that we're doing to keep our bodies from, you know, from erosion and and being able to monitor and manage that is just I feel like we're we're really starting to to get sophisticated around some of this stuff. And I think it's going to just yeah, it's going to continue to march on as being an important category.
00;20;03;22 - 00;20;04;28
Derek
People want to back in fund.
00;20;04;29 - 00;20;11;10
Chris
So you know pre is on like a thought fluency tear. Right now she she's the hottest thing on the timeline.
00;20;11;11 - 00;20;12;17
Pri
And I can't.
00;20;12;17 - 00;20;16;27
Derek
Believe I can't believe she carved out an hour to join us for this call this week.
00;20;17;00 - 00;20;20;21
Pri
My head's my head has gone really big you guys it.
00;20;20;23 - 00;20;30;01
Chris
Well, I'm about to give you even more credit because I'm about to reference one of your tweets where you said you wanted an AI mattress. I want you to pitch Derek right now.
00;20;30;03 - 00;20;49;03
Pri
Okay? Yeah, I was sorry. So I also think every every object is going to have a genetic element. So I was in bed, I was on my phone, I was scrolling endlessly, I couldn't go to bed. And I was like, damn, I wish my bed was able to nudge me to stop doom scrolling so I could go to sleep.
00;20;49;03 - 00;20;59;21
Pri
And then I was thinking about it. I'm like, it would actually be kind of cool to have an AI mattress that could, like, understand what you would. I don't know, it would just be like, it could just like move.
00;20;59;22 - 00;21;00;05
Aaron
Like it would.
00;21;00;05 - 00;21;01;07
Pri
Get bump it.
00;21;01;09 - 00;21;03;14
Aaron
Yeah. It was like, make it uncomfortable.
00;21;03;16 - 00;21;08;27
Chris
Or maybe, maybe it gets softer and softer and she just sinks into the bed till she falls asleep.
00;21;08;28 - 00;21;24;24
Pri
Yeah. It just does something and it just nudges you or like, vibrates and it's just like, get off your phone, get off your phone or something like that. And then, you know, if it felt the movements, it could adjust your temperature. Like, I mean, I know eight sleep exists, but it's very like you set a temperature and go to sleep.
00;21;24;27 - 00;21;45;09
Pri
This would be kind of interesting if it was able to track like your movements and like just really find the most like optimal. Yeah, optimal thing. And then if you had a wake up time, it was able to like nudge you up in like an alarm or whatever, I don't know I don't know how this would work, but I think something that really could contoured to your own body.
00;21;45;09 - 00;21;53;28
Pri
And, you know, if you had sleep apnea or something, it was able and it knew that it could kind of adjust accordingly. Thoughts on this pitch? Derek.
00;21;53;29 - 00;22;19;09
Derek
I actually love this pitch. I will say I, I bought my first eight sleep in 2021 and have been a devoted user of the product for the last, whatever that's been five, five and a half years or so. And the things that you're describing is exactly where a sleep is trying to go. So right now there's so basically that you, you, you either by the mattress, you buy the mattress topper, you lay on top of it, you fill it with water.
00;22;19;10 - 00;22;37;19
Derek
That water can heat or cool depending on your preferences, those your body room or cold at night. What time of the year is it? Can it calibrate based in real time based on how warm its sensing you're getting and or how you're sleeping, and the next day you're provided you get a text message with just like how you slept.
00;22;37;19 - 00;22;55;17
Derek
So like, I mean, I was just looking at mine a couple hours ago. I'll tell you what it said. It said your eight seat metrics of the night, your sleep fitness score is 92. You slept for seven hours and 30 minutes. You had deep sleep for 18% of the time, REM sleep for 30% of the time, 63 beats per minute, HV 54.
00;22;55;17 - 00;23;13;18
Derek
And then it gives me a bunch of like how that benchmarks against previous averages. And so the stuff you're suggesting around like just laying on the mattress and it giving you alerts like maybe your past your bedtime on average or maybe you're haven't woken up you they actually started building this in I think it was a couple years ago.
00;23;13;18 - 00;23;21;29
Derek
And it's like the toppers that were shipping in the matches that were shipping. So right now, like I get, I get woken up from a bug like the mattress will shake, it'll vibrate.
00;23;22;01 - 00;23;23;09
Pri
Oh it does, it doesn't.
00;23;23;09 - 00;23;45;05
Derek
Even know. And so I think the next step is what you're describing, which is like it's now, I don't know, 1140 at night, I'm just scrolling Netflix or whatever, Instagram or TikTok. And it's like you're an hour and a half past your bedtime. It starts rumbling. It starts like an alert comes up on your phone. That definitely feels like where products like this are going to go.
00;23;45;05 - 00;24;15;25
Derek
And then having all this information and data, being able to kind of like mesh with the larger preventative health stack that I have so that like maybe my nutrition in the morning matches the how I slept or like the recommendation that like, hey, I should probably take a power nap at 2:00 between these calls. Like, I think this idea of what you're saying is, is not I don't know if it's like in the next three years, but like, I definitely feel like we're headed to a world where, like, these things are just going to start layering on top of each other.
00;24;15;26 - 00;24;30;28
Pri
Well, Aaron said this really well, he was mentioning and I like, like it. It's like the IoT vision actually might become more alive with AI. Like maybe it's like there's going to be an IoT arc. Probably. Yeah. And that's I think part of it.
00;24;31;02 - 00;24;49;24
Aaron
Yeah. My thought there is just like everybody's like when I think of increasingly when I think of AGI, I've been trying to just wrap my head around it. And I think the right shorthand is like the more and more powerful these systems get, it's really just about robots and making them smaller so they can go into robots. And like, that's the goal.
00;24;50;01 - 00;25;13;05
Aaron
That's why Elon is even building rock, right? Like that's super important for space. It's really what I think if we were internal at OpenAI or anthropic, they're they're probably thinking a lot about like how to embody these things and the way that the, the models for robots tend to work now is that they use a lot of vision information, like a world model, to process things.
00;25;13;05 - 00;25;37;04
Aaron
And I just feel like it's probably more efficient to use like sensors. So I think there's going to be like a corresponding push to put more sensors in different parts of, let's say, like an industrial, like a manufacturing process or, you know, other places where humans work just to make it easier for the robots. So but thinking a lot more about like, IoT, I think the costs have been there for a while.
00;25;37;05 - 00;25;42;06
Aaron
I think the demand hasn't been there. And like, I think these robotic systems will create a ton of demand.
00;25;42;09 - 00;26;12;00
Derek
Yeah, I definitely agree with that. And I think your gut is right. I do feel like I do feel I don't know if it's a new startup or eight sleep nails it or what the story is there, but like, I think we spend whatever half of our lives in bed and it's like starting to. Yeah, starting to digitize and optimize that entire process feels like, you know, something that feels like a low hanging fruit for some of, like, this new technology that we've got.
00;26;12;03 - 00;26;22;23
Pri
Yeah. And that's like AI for good. I actually think people would like that. It's not like taking people's jobs. It's actually optimizing their life in a meaningful way, I don't know. Yeah, I'm looking forward to seeing products like that.
00;26;22;24 - 00;26;43;27
Aaron
I saw my first AI slot marketing in a physical store yesterday, by the way. There was a ice cream shop that I went to, and they were showing their ice cream with like an obvious slop video that had like clear AI, like hallucinations. And it was kind of wild. Yeah. Like it it was showing them make like the ice cream.
00;26;43;27 - 00;27;00;29
Aaron
And then instead of putting down like one slice banana on the ice cream, it put for like instantaneously. It was just like something was clearly off on it. So in the one hand I'm like, oh, look like the store is like leaning into technology. Like that's kind of interesting. At the same time, I'm like, oh, that's not how you do it.
00;27;01;01 - 00;27;01;17
Aaron
Yeah.
00;27;01;22 - 00;27;07;16
Chris
Come on guys, it's an ice cream shop. They're allowed to do any sort of weird ass marketing they want.
00;27;07;18 - 00;27;21;06
Aaron
Yeah, I mean like I appreciated those, those bits, but it's definitely like beginning to diffuse. It wasn't like a super fancy look, like, you know, like a family or business on that one. I mean, I applauded the effort.
00;27;21;08 - 00;27;46;04
Chris
Yeah. Let's let's turn to AI. AI for good, for good for a second. Because obviously, you know, we've seen that in Midjourney as an actual product. I think we're also very clearly seeing the pushback against AI, doom trolling, you know, and anthropic is that had itself quite a run over the last week or two where I think we're getting a collective enough is enough, guys, can you please be normal?
00;27;46;07 - 00;27;49;08
Chris
What are the times in the whole fable?
00;27;49;10 - 00;28;01;11
Pri
The times of all people, the times of all people publish that? That's surprising even. They're like, okay, we're tired of this. This the the boy who called wolf thing like, give it a rest.
00;28;01;12 - 00;28;21;11
Aaron
Well, because it's there's just like a lot of cognitive dissonance around it. Right. Like on the one hand, they're plowing forward with incredible speed. At the same time, they're yelling from the rooftops that all this is going to be horrible, right? And there isn't any data yet to kind of suggest that where if there is data, it's not conclusive one way or another.
00;28;21;14 - 00;28;32;29
Aaron
Maybe that's a kind of a better way to to phrase it. So yeah, it does feel like it kind of is. There is like a concerted effort from a couple of different channels to to try to change that conversation.
00;28;33;00 - 00;29;01;06
Chris
Yeah. I mean, I think maybe the, the meta narrative around all of this is, ironically enough, how unaligned anthropic is with everyone else. And for the lab that, you know, is really, really focused in on alignment, you know, the kind of missed the forest for the trees in terms of aligning with the rest of society. And, you know, it seems that everyone else is having a come on man type of moment.
00;29;01;07 - 00;29;24;11
Chris
Let's let's bring you back into, you know, the rest of the human race here because whatever road you're going down like it ain't, it ain't jiving. Well, with us, there's also like, I think there's so many different issues at play here. There's so many like motivations to form, I'd say temporary alliances around knee capping and dropping, slowing them down.
00;29;24;11 - 00;29;47;05
Chris
It's probably really hard to to read the tea leaves of what's actually going on here. You know, it was really interesting to see that Seattle was so far out in front around this. You know, Andy Jassy being the one who sounded the alarm. You know, Amazon has a stake in anthropic. And so, you know, you're pulling the fire alarm on like your own investment.
00;29;47;05 - 00;30;12;09
Chris
But then, you know, Satya coming out with that blander statement. I mean, you know, like clearly the traditional mag seven, you know, is getting a little nervous here. And maybe they're doing this just to pump the brakes and, you know, help preserve their, their business lines and, you know, kind of prepare them for future. You know, it's almost like we can't let anthropic go lapping people here.
00;30;12;10 - 00;30;15;17
Chris
Let's let's tie a rope around them. No.
00;30;15;24 - 00;30;31;03
Derek
No, I think I think you could be right. I think you could be right. It's so hard to know. I feel like we're. Yeah. I mean, like, there's there's. So these interests are so intertwined, right? Because there's, there's, you know, the the Max seven has a massive amount of exposure in most part to a lot of the success around this technology.
00;30;31;03 - 00;30;57;00
Derek
It's so very clearly the next big revenue unlock for these firms directly and also indirectly via like their 5 to 20% ownership over some like the major labs. Meanwhile there are real risks, you know, around this technology like whatever there's restraints to the First Amendment, right. Like you don't want things like the ability to build a nuclear weapon, you know, being used to this technology.
00;30;57;00 - 00;31;26;11
Derek
You don't want the distribution or dissemination of illicit material for minors being disseminated through this. You don't want people plotting to overtake, you know, parts of the government or kidnap people using this technology. I mean, there are, you know, even and not even to mention just like the competitive threats with just like having US intelligence leaked out to foreign adversaries, like there's a million reasons why in the wrong hands, some of the power of these tools could actually be quite destructive.
00;31;26;11 - 00;31;56;16
Derek
And so, like the First Amendment is not completely universal in its reach. There needs to be tempers. And so, yeah, I think what I think what I'm seeing is just this grappling that's happening between like, okay, between the economics, between the radical benefits, the between like the safety concerns. And it's coming out sideways through these conversations or through Dario fumbling in front of the US government on these phone calls or through Satya trying to figure out like, yeah, we're about to make a ridiculous sum of money.
00;31;56;16 - 00;32;19;26
Derek
But also, you know, can North Korea just like, use this stuff to break encryption on a number of private things that we have? I mean, there's yeah, this is where I feel like as a bystander and informed bystander who's obviously like following this technology very closely, there is a lot there's a lot to unpack here. And there is an argument that, like, we're moving a hair to fast on some of the stuff, right?
00;32;19;27 - 00;32;49;07
Chris
I mean, I this is this came really far really quick. I think the other interesting thing is I get the sense that anthropic walked into a trap. You know, everyone knew this model was coming. They previewed it. They ran, you know, that is a golden wing or whatever. The glass wing, you know, whatever the cybersecurity access part. And so no one was taken by surprise when the model came out.
00;32;49;08 - 00;32;59;26
Chris
Which leads me to believe that, you know, they intentionally let it. You know, no one stopped it then. But, I mean, you know, I just feel like it was a bit of a trap.
00;32;59;27 - 00;33;27;27
Derek
Yeah. I think the difference from my understanding of glass swing and I think I think pretty universally felt that glass swing should not be in public hands, was the sense that I got from a lot of the early partners. But what they did acknowledge was that like a guarded or a guardrail product release around basing that kind of limited the ability for the manipulation of the the model to be used in disadvantageous directions could be permissive.
00;33;27;27 - 00;33;51;06
Derek
And I think that was the piece that people were like were okay with. But what what seems to have come out was an ability to root around that blockade or those guardrails to basically avail yourself to the full benefits of glass swing, which was a big no no on everyone's kind of on everyone's list. When, when these when people were kind of like playing around with these models.
00;33;51;06 - 00;34;11;26
Derek
And I think the argument is stemming from where the guardrails really bypassed was the benefits actually just part of this neutered model that were being received? I think this is where yeah, I mean, this is where some of like this, this I think the trick happens because I yeah. And frankly, we're not in the room. So we can only speculate.
00;34;11;26 - 00;34;19;26
Derek
But I think people may have different interpretations of like what can truly be accessed even when things like stringent guardrails are put in place.
00;34;19;27 - 00;34;40;16
Pri
I agree with that. My one thing, and maybe that is something that is what happened. Like that's like the very clean, non conspiratorial take and probably likely take of what happened. However, what I don't understand is like when that was the case and there was like these like kind of let's call it back doors in order to get to that model.
00;34;40;18 - 00;34;44;21
Pri
Why was there apprehension to shut down then?
00;34;44;22 - 00;34;52;05
Derek
Why was there apprehension to shut down the with. Oh. Bye bye bye anthropic.
00;34;52;08 - 00;34;53;00
Pri
Yeah.
00;34;53;01 - 00;35;20;13
Derek
Yeah I think this is the big question because like I think we're getting into kind of he said she said territory Amazon the US government or Satya from Microsoft. A lot of people are claiming that what was the guardrails were so deprecated that it made basically it made glass swing usable to the public. And so like the fable guardrails weren't actually able to withstand some of the pressure testing.
00;35;20;13 - 00;35;47;02
Derek
But what Dario is saying is that's not true. And the benefits that were being received were similar to benefits that could have been received by any soda model, including like 5.5 and code or Codex and a number of these other kind of best in class soda models that were released at the time. And the reality is, is like, I don't think any of us know like where what's true or what's not true, because we don't have access to some of the intelligence that's circulating around this issue.
00;35;47;02 - 00;36;04;01
Derek
And I think that's where the he said she, she said, is coming into play, where it's just like, I think we're kind of just watching this song and dance happen around some of the largest players in AI and the US government, and, and people are just speculating one way or the other.
00;36;04;04 - 00;36;14;24
Pri
Yeah, I get that. I guess maybe like, I would think like if Amazon set the the US government are all saying that, then it's like maybe it's maybe.
00;36;14;27 - 00;36;27;28
Derek
It's a, it's a really good point. But then I think other people are like, well Dario has been maybe the guy holding the banner for safety AI and like, would he be really the one at this point to.
00;36;28;01 - 00;36;28;05
Pri
To.
00;36;28;06 - 00;36;44;08
Derek
Let it to kind of like to let it run? Yeah. To like just let it ride. Like to just say just yolo into this thing. When he has been the only one of the enterprise partners that has been like championing these slow release rollouts, that has been trying to advocate for policy. And so this is where, like I would say, I'm confused.
00;36;44;08 - 00;36;54;23
Derek
I don't know what to believe or who to believe or where this issue, how this where the lines are around this issue because it doesn't match the previous fact patterns.
00;36;54;24 - 00;37;06;00
Chris
Well also to play like five d confusing chess, we do have to realize that like effective altruists or dubious class of people just based on past history, especially crypto.
00;37;06;01 - 00;37;07;04
Derek
Here, here.
00;37;07;07 - 00;37;29;25
Chris
Yeah, you know, my my, I have a simple solution to this whole problem. And that anthropic needs to hire Roger Sterling from Mad Men to be the public face of the company and to smooth all these things over. They just need a really good client guy who can have a couple martinis, speak the right language, get everyone the same page, and then come back home until Dario look to see what you can do.
00;37;29;26 - 00;37;44;14
Chris
This is what you can't do. I smoothed this over as best I could. You know, I think we've got a lot of people who just aren't really speaking the same language right now. And you know, they just need a really good old school network to work this out.
00;37;44;16 - 00;37;46;28
Pri
The committee is probably working overtime, Chris.
00;37;47;00 - 00;37;57;15
Chris
Oh, I'm sure they are. I'm sure they are. Speaking of things we didn't expect to see on the timeline this week. How about our buddy Eli Cart and Kanye around our Basil?
00;37;57;21 - 00;37;58;09
Pri
That was.
00;37;58;11 - 00;38;19;19
Aaron
I didn't know how to feel about that. I was like, It's Connie here now. I mean, I know I'm sure it was like pretty exciting for for Eli, but I don't know. Connie is such a. I mean, obviously very talented, but just the past couple of years is such a tricky character. So, I mean, and then and then at the same time, I'm like, I don't think he is kind of always early to things.
00;38;19;19 - 00;38;26;29
Aaron
So I found that interesting as well. But I don't know, I like I left very conflicted. Poconos makes me feel a bit conflicted.
00;38;27;00 - 00;38;29;01
Pri
He's not anymore. I don't think.
00;38;29;03 - 00;38;30;24
Derek
He's not. He's been canceled.
00;38;30;25 - 00;38;31;27
Pri
I think so.
00;38;32;00 - 00;38;35;19
Aaron
He may not be canceled, but it doesn't mean he's forgiven, if that makes sense.
00;38;35;20 - 00;39;00;04
Derek
Yeah, I think we're I think this is a obviously a very conflicting story for a lot of people. Very divisive topic, even just within the small, small crypto art timeline. I just feel like people were very quick to jump on one side or the other around it. So it feels like kind of a no no winners around this, this one just kind of a very touchy topic.
00;39;00;05 - 00;39;09;09
Pri
It is this one. Need to see them though. Like card around zero ten. I saw all the pictures of him talking to like William upon. I am wondering if he ended up buying anything.
00;39;09;11 - 00;39;14;05
Chris
There you go. Let's make it about our bags.
00;39;14;07 - 00;39;17;10
Pri
I mean sorry I am curious though. Like I wonder.
00;39;17;12 - 00;39;20;13
Derek
What a show he's there to buy. Art. It's an art show.
00;39;20;14 - 00;39;38;06
Aaron
Yeah, it was less curious about that. But I mean, I think William is a very one of the most talented artists in our space. So I was excited, just even separate apart from Canada to see that people were beginning to enjoy his work. He had a successful sale through Art box, which I thought was great. So that.
00;39;38;06 - 00;39;39;22
Pri
Is work. Looks fantastic.
00;39;39;25 - 00;39;51;17
Aaron
Yeah, and I thought the work was some of the strongest that he's put out since anticyclone, which is great. You know, I think he should have a long career and I think he's notable in so many different ways.
00;39;51;17 - 00;40;06;13
Chris
It was also great to see Infinite Garden so prominent. Andreas Geist there. You know, a lot of good art at that show getting exposure. And so I'm glad that, you know, continental Digital Art had a successful run in Basel.
00;40;06;14 - 00;40;09;21
Derek
Did any of the other Basel stuff stand out to you guys?
00;40;09;23 - 00;40;17;18
Chris
A lot of it. I felt like we, we, we already saw. Right? I mean, you know, that's the thing about these big shows is, you know, some of these works been out for a year.
00;40;17;22 - 00;40;23;13
Derek
The like the John Gerard flags. Yeah, I think that was originally in art blocks work.
00;40;23;16 - 00;40;27;20
Chris
Yeah. The John Gerard stuff doesn't do it for me. Like. Oh great, you made another digital flag.
00;40;27;22 - 00;40;32;25
Aaron
Like I thought that was. Yeah, that was through the pace. Art blocks clad. Right. Derek. If my memory served.
00;40;32;25 - 00;40;34;20
Derek
I think so, yeah I think that's right.
00;40;34;21 - 00;40;43;16
Pri
I'm trying to think what else was there. I'm like literally blanking. All of a sudden I was like looking at like the full. There was like a YouTube video of like the full tour.
00;40;43;19 - 00;41;07;10
Aaron
It looked good. It didn't look like crazy because I don't think there was like a people like moment. But I thought, I thought it was nice to see artists showing their work, working on great things. You know, it just gives, gives a little bit of confidence. And to the extent that Kanye is ahead of the curve, which he's had a history of being, maybe that like foretells some some more excitement around that entire category, which would be great.
00;41;07;12 - 00;41;34;16
Pri
You know which one I liked? I like the hito, the hito stare stereo. I don't know if I'm saying his last name right. I like the def beef quite a bit as well. And obviously the Avery Singer maximalist works huge fan of those. So I had a couple. I had a couple favorites. Of course, the Harold Cohen paintings were sick too, so I was just like looking at the list.
00;41;34;17 - 00;41;40;12
Pri
I'm like, they were I thought, I thought I actually thought Avery Singer's work was really, really good.
00;41;40;14 - 00;41;41;26
Aaron
She's super talented.
00;41;41;29 - 00;41;45;06
Pri
Yeah, he don't work too. That was sick.
00;41;45;07 - 00;42;04;15
Aaron
Well, we talked about AI, you know, going into mattresses, do we think do we think we start seeing more like autonomous agent based art systems soon? I've been thinking a little bit about that. Like I'm surprised we don't see a little bit more of that type of work. Kind of like, well, like Botta was like a really amazing first version of that.
00;42;04;17 - 00;42;09;24
Aaron
But I'm surprised there's not like a V2 or V3, like something pushing the envelope there a little bit.
00;42;09;25 - 00;42;10;22
Pri
Yeah.
00;42;10;25 - 00;42;12;19
Aaron
It's more like a system of some sort.
00;42;12;20 - 00;42;41;25
Derek
I think we will eventually. I think we're yeah, I think we're, I think people are so in like kind of like the corners of the AI community. Like these experiments are actually pretty relentless. I get I, you know, folks from portfolio companies that are building kind of like bleeding edge AI stuff that I was sending. They, hey, check this out or, you know, somebody whatever has figured out how to like, you know, pull Ascii art out of like these really crude early models and is making work around that constraint.
00;42;41;26 - 00;43;03;13
Derek
Like there's there's a lot of kind of at the edges, like playfulness around these tools right now, which I think is really, really fun. I think over time, you know, more and more these, these kind of experiments are going to get. Yeah, like released in a way that feels like their inauthenticity with like an actual practice and not just kind of like experiments.
00;43;03;13 - 00;43;34;09
Derek
And hopefully a couple of these things get packaged or produced in a way that feels like a very compelling artwork to engage with. And I think we'll probably see that soon. I think we'll, you know, this, this, this technology is just like, so it's just. Yeah, the it's just it's so vast and what can be accomplished. So to your point, and maybe the some of the design around it is really focused on like the systems behind it, maybe it's, it's around some of how the experience degrades around some of these systems.
00;43;34;09 - 00;43;59;14
Derek
Maybe it's around. Yeah. It could be objects. I think we have that little debate last week or like Chris and I were riffing on, like, does the ontology of of what's interesting about AI art actually mean it gets produced into an object in the end, or is it really meant to serve a different role in the artistic endeavor, which doesn't have an object which I could fully by and is part of what I'm what I'm seeing right now?
00;43;59;14 - 00;44;06;24
Derek
So, yeah, I think we're I think we're probably going to see some cool stuff soon. If I had to guess, I just feel like we're it's getting we're we're teeing up for that.
00;44;06;25 - 00;44;29;20
Pri
I'm excited. I actually have been seeing some cool stuff because I'm like helping curate, like some of the stuff that's coming through the AI Psychosis Summit in New York. And there's like, they basically like the whole idea is to try to attract AI artists who aren't quote artists. They're just like people messing with AI but producing art. So it's like not, you know, artists you would ever they wouldn't even call themselves artists.
00;44;29;20 - 00;44;49;18
Pri
They're just being artistic with the tools. And so it's definitely more like I'll share as it comes. But it's been kind of cool to see that there's a lot of people who are who. I mean, it's kind of like the NFT space where a lot of people who were maybe graphic designers and had like artistic practices, but that wasn't how they made money.
00;44;49;19 - 00;44;57;13
Pri
There's like a lot of people who are artistic and know how to use these tools, but, you know, are now actually making art with it.
00;44;57;20 - 00;45;03;24
Chris
You're on a heater here. You're just casually dropping that. You're doing a little curation work for AI psychosis.
00;45;04;00 - 00;45;08;12
Pri
You know? Yeah. I mean, it's like I'm helping. I wouldn't call myself a curator, but I'm like helping.
00;45;08;15 - 00;45;15;06
Derek
I think everyone, everyone, everyone's got a little curator in them. The internet makes possible. I would definitely call you a curator pre.
00;45;15;07 - 00;45;16;26
Pri
Thank you, thank you.
00;45;17;00 - 00;45;19;01
Chris
Is she a curator of taste?
00;45;19;04 - 00;45;20;29
Pri
Oh, my God, we didn't.
00;45;21;01 - 00;45;24;08
Aaron
Put you on the minefield.
00;45;24;11 - 00;45;28;02
Pri
You want to talk about taste labs? No, I'm not going to say anything positive.
00;45;28;03 - 00;45;30;26
Derek
God, those comments were. Those comments were so funny.
00;45;30;26 - 00;45;54;10
Aaron
Though muted on my timeline. I was just like. This is too much. Yeah. I mean, it's just even in the name, like if you it's called Taste Labs. It's like the most functional name ever created, right? I don't know, I mean, I don't know whether or not you can't replicate taste using like a more automated system, like, I'm, I'm still not fully willing to say that you can't do that.
00;45;54;12 - 00;45;57;11
Aaron
Like, I can't tell if that's, like, a little bit of coke, but I.
00;45;57;11 - 00;45;58;23
Derek
Think I.
00;45;58;25 - 00;46;19;04
Aaron
I mean, like, Netflix tries to do this, right? I just like the like Netflix tries to do this just in the media. Right. Like they try to predict what people want to see. I know that there's like some early experiments there. So and I don't know if that's taste or not. Right. Like then you get into this like more like a theoretical metaphysical type question, like what is what is taste?
00;46;19;06 - 00;46;40;06
Aaron
But you know, if the extent that you're defining it as like, you know, things people want, I don't see why the predictive capability or forecasting capability of an AI system may not, you know, match like the top quote unquote tastemakers in the future. You know, that being said, I don't know if this this company I don't know anything about this company.
00;46;40;08 - 00;46;46;01
Aaron
Maybe they have the goods, but it just didn't feel like it. They do from like this is all they were showing.
00;46;46;02 - 00;46;55;08
Derek
I think the issue with, with this, with whatever trying to taste is, is I hate to throw my hat in the ring in this debate because I think it's like, whatever.
00;46;55;09 - 00;46;55;29
Chris
You can't.
00;46;55;29 - 00;47;17;29
Derek
Resist yet. I can't resist you. Hey, they gave me a microphone. Chris. Right there. Ask the people are asking for my thoughts every week. Listen, I do. Part of me feels like there's a huge communication gap, which is like people who actually have tastes that culturally set in motion, like trends or ideas or, you know, it could be music, it could be fashion, it could be whatever.
00;47;18;00 - 00;48;02;05
Derek
Understand that the way taste emerges is never from like from being able to kind of like collect more data and then come up with a conclusion. It's so much more softer than that. And these ideas are so what's interesting about things is like sometimes these very, very tiny insights that get pulled out of a whatever, a piece of clothing and then gets refactored and reintroduced somehow in like set design for a music video, which then gets, you know, that tiny little idea that somebody picks up on and perceives gets pulled out and brought into, you know, film which then sets in motion, kind of like Uniqlo trying to, you know, create or commercialize the jacket
00;48;02;05 - 00;48;26;26
Derek
to 50 million people in North America. And so it is like this trickle down process that's it's not what's cool can't be engineered to be cool. It really is really emergent out of this entropy that happens around how people live on the planet. And I think a lot of people who are really into this idea of taste, truly into the idea of taste, not like the classical definition of like, I've got a taste so I can whatever.
00;48;26;26 - 00;48;54;06
Derek
I'm going to be more interesting as it relates to AI. Not that idea, but actual interesting, informed like culture, bleeding edge like, you know, tip of the spear taste. No, that like building a fucking AI lab to kind of like reproduce this for the masses is the exact opposite of how taste originates. And I think they're speaking past each other because like, other people are like, no, we're going to quantify it and it's going to be awesome and everybody's going to have access to great taste.
00;48;54;06 - 00;49;15;13
Derek
But the mere idea of that is the opposite of how emergent the you know, the concept, in my view, like the concept of taste originates that I think it's just I think people are just like kind of like laughing past each other on the concept that like a Taste Labs is going to bring taste to the masses. Is that resonate with any of you guys or spend a lot of time thinking about this?
00;49;15;18 - 00;49;34;20
Pri
Yes. Though I mean, I could say, okay, so there's so much wrong with that. Like, I mean, I don't want to put down anyone who's like starting a company, but like, it was it was embarrassing. And so in like the image of after being like join us like and everyone looked just so not I don't know whatever I have thoughts on that.
00;49;34;20 - 00;49;58;21
Pri
But but the actual concept, I don't know if I fully I agree so so llms like, I think there is an idea where you could have specific categories of certain categories that could be trained on more elite data than others. So and I think taste. Yes, of course people say it's subjective, but sometimes it's just straight up objective.
00;49;58;22 - 00;50;27;24
Pri
And so let's say you wanted to taste I think I was talking like, damn about this to like, if you wanted to create pull like the best makeup designers on YouTube, like the ones the makeup designers for the stars, and they're clearly like the most elite makeup designers on Instagram and YouTube. You probably could pull that and have an LLM that's going to be better than a generalized oh on makeup makeup tutorials like it actually probably would lead to better esthetic output.
00;50;27;24 - 00;50;45;14
Pri
So in that sense, I guess I would say that the taste thing could make sense in like these certain narrow niche categories where like there's just a level of professionalism needed or tasteful professionalism. But I don't think that this is what they were talking about.
00;50;45;15 - 00;50;54;29
Chris
Yeah. I feel like what you're describing is more like the successful application of already established tasteful things.
00;50;55;00 - 00;50;55;29
Pri
Correct.
00;50;56;02 - 00;50;58;11
Chris
Versus the emergence of taste.
00;50;58;13 - 00;51;09;15
Pri
And is that what are they trying to do? I thought they were like trying to do the latter, or like maybe they're trying to create out of the generalized LMS some tasteful outcome, as.
00;51;09;15 - 00;51;11;15
Chris
I have no clue. I haven't looked at this.
00;51;11;17 - 00;51;19;20
Pri
Yeah. I mean, I just kind of was just like, this is like cringe. Also, the name Taste Labs cringe like whole thing. Not interesting to me.
00;51;19;21 - 00;51;38;08
Chris
You're going to look so bad once they successfully invent the unit of taste that once we have the second in the milliseconds for taste, you're going to look terrible here. Like, I've total confidence that these guys are going to be able to to crack that nut. And we're going to be talking about taste units.
00;51;38;11 - 00;51;43;02
Pri
But are they based in San Francisco just to kind of run that out because.
00;51;43;06 - 00;51;47;00
Chris
I don't know Aaron muted them and I'm not paying attention. So you're on your own.
00;51;47;00 - 00;51;51;14
Pri
Here I assume I assume if you're naming it Taste Labs you're probably based in San Francisco.
00;51;51;15 - 00;51;52;13
Chris
But probably.
00;51;52;15 - 00;52;13;22
Aaron
I don't know where they were. I mean I think there's probably an interesting engineering challenge on how to on how to begin to solve some of that, that problem. Right? I mean, a lot of taste is forecasting, right? It's taking something that interesting people or what like interesting people find interesting now and predicting we're forecasting what everybody would like in the future.
00;52;13;24 - 00;52;43;00
Aaron
Right. That could be a pair of X right or Crocs. Right. Or it's a really broad category. So I can't tell if that's what they're focused on. And if that's the case like okay I could see that, right. Like it's kind of like looking for a little bit of signal early signal from like a lot of noise. And that's like a pretty interesting problem if it's more like, you know, on the fine art side, I don't know if it will kind of hit there, but so that's kind of like also just some of my questions related to it.
00;52;43;01 - 00;53;00;21
Aaron
Like I'm not completely willing at this point to kind of foreclose that. That's not a possibility. But I think I think it's really hard engineering challenges. And I think you're going to need a like a pretty good and interesting lens on culture and media, because I imagine that's what it's mostly focused on.
00;53;00;24 - 00;53;02;13
Pri
Yes, you definitely.
00;53;02;16 - 00;53;08;29
Aaron
Like consumer goods to kind of really build a to solve some of those problems and build a product that that does.
00;53;08;29 - 00;53;30;09
Pri
It taste takes time to develop like as when we all have our own separate tastes, like we probably spent several years of our lives like harnessing that and making sense of our own taste, like the notion that this group of people is going to be able to, like, pick something that is going to be universally high taste. I don't know, it feels good luck.
00;53;30;10 - 00;53;32;06
Pri
I guess.
00;53;32;09 - 00;53;48;11
Chris
I'm telling you, for you, once you get up to 40 gigabits per second of taste processing, they're going to solve this problem. Stop going to have little, little speakers just sending out echolocation all around. Just measuring culture at 40GB per second.
00;53;48;13 - 00;53;54;01
Pri
Watch me be wrong. And this is like the like next like neo Lab. That's just like.
00;53;54;02 - 00;54;16;12
Aaron
Well, I don't know. Yeah, I think I think it's more like forecasting. Did you guys remember that AI Dash 2027 kind of right up. It came out about a year ago like May last year. We can include kind of in the show notes. But it was an effort to kind of like predict like how AI how the AI race would, would continue, you know, over the next two years.
00;54;16;12 - 00;54;40;12
Aaron
And, you know, part of that and they have it kind of on the side is like a little prediction related to forecasting. And they're they're saying, you know, by this time next year, maybe a little bit later, let's say, like summer next year, that across a number of domains, like the ability to forecast will improve where the an AI system will be better than the best human.
00;54;40;18 - 00;55;01;26
Aaron
So that's by like June of 2026. So the top performing AI in the range of a, you know, is is pretty much in the range of a professional in that domain. So that's by this time next year, and then by the end of the year it would hit like a superhuman capability. So let's say like, you know, July, December.
00;55;01;28 - 00;55;08;11
Aaron
And if that if that's true and it may not be true right then I don't know. Like maybe, maybe we start seeing a bit more.
00;55;08;16 - 00;55;10;15
Pri
2027 is going to be insane.
00;55;10;16 - 00;55;37;11
Aaron
Yeah. And I think, you know, I kind of remember reading that and thought that was a really well done work and it kind of goes all the way through the end of 2027, where it basically is leading to like significant government oversight, which is just going to be kind of fascinating. And I think your predictions pretty or your forecasting where I think you are superhuman already, I think that, you know, that's going to align with like an election of some sort.
00;55;37;13 - 00;55;39;27
Aaron
That's going to be really complicated.
00;55;39;29 - 00;56;03;25
Pri
Yeah, yeah. 2027 plus then you like the 2028 election. It's going to be pretty crazy. And then like there's gonna be like, people are going to space like, I don't know next couple of years, but we like need to spend. I mean we don't have that much time today, but we should spend some time talking about like, space and like the broader implications of, like, orbital travel and like what the world looks like in ten years.
00;56;03;25 - 00;56;21;28
Pri
Because I do think that's like going to be crazy. And it's moving very quickly that that is the space stuff has potential, the potential to move as quickly as some of the AI stuff potentially like I wouldn't be surprised in like 3 to 5 years. I don't know when to ChatGPT get released like three and a half years ago, or was it four and a half years ago?
00;56;21;28 - 00;56;30;15
Pri
I can't remember now, but and it's moved pretty quickly. I wouldn't be surprised in like 4 or 5 years of the space stuff pretty accelerates quite a bit.
00;56;30;21 - 00;56;56;04
Chris
It could also serve as a forcing function in which, you know, we get a lot of great technology back back down here on Earth, the same way, you know, that NASA produced a ton of stuff back in the day. You know, especially if space is a because it's so net new in a sense, you know, that AI and automation, you know, might be like very, very central once they get this whole rocket thing sorted out.
00;56;56;05 - 00;57;16;07
Chris
Right? Actual doing things in space is, you know, enough of a blank canvas that you can really be using, you know, just brand new blueprints that can drive innovation even quicker. It's not like they have to interrupt with anything, you know, down here on Earth. The same way you still got to plug you get your AI mattress still needs to fit in the bed frame.
00;57;16;07 - 00;57;20;06
Chris
That's true. All right, gang, we have we run our mile here.
00;57;20;06 - 00;57;20;29
Aaron
We've done it.
00;57;21;00 - 00;57;42;18
Pri
We've done it all. Introduced the podcast. So hey everyone, welcome to Net Society to the. Today you have me, Chris, Derek and Aaron talking all things tech, crypto, AI, art, culture and more. Just as a quick reminder, these thoughts and opinion are our own and not of our employer. And none of this is financial advice.