KP Unpacked

Why Robots Spark More Outrage Than Digital AI

• KP Reddy

Share episode

Use Left/Right to seek, Home/End to jump to start or end. Hold shift to jump forward or backward.

0:00 | 56:15

What is it about watching a machine tape drywall that creates visceral discomfort in ways software automation never did?

In this episode of KP Unpacked, KP Reddy and Nick dissect the emotional response to physical AI versus digital AI. Nick's Okibo robotics video got 300K views and sparked a firestorm: half celebrating reduced construction costs, half horrified that "they're coming for the physical jobs too." The backlash reveals something deeper. People feel guilt about blue-collar displacement in ways they never did about white-collar knowledge work. Why? Because physical labor was supposed to be the fallback when AI took everything else.

KP counters with the mop thought experiment: would you pay your house cleaner more to scrub floors by hand without tools? Of course not. So why do we romanticize construction labor that breaks backs when better tools exist? The conversation moves from a software engineer quitting over AI coding adoption (identity crisis around lost craft) to whether nostalgia will create retro coding communities the way vinyl and Japanese stationery stores preserve analog experiences. Then they pivot to the scarcity flip: intelligence is now abundant and cheap, but transformers have 18-month backlogs. A startup building next-gen transformers would have been laughed out of Shadow Ventures three years ago. Today? Immediate funding.

Key questions answered:

  • Why does watching robots do drywall create more outrage than software writing code?
  • What happened when Nick posted an Okibo video that got 300K views?
  • Would you pay your house cleaner more to scrub floors by hand without a mop?
  • Why did a software engineer quit when his company adopted AI coding tools?
  • What's the nostalgia equivalent for coding: vinyl, retro Game Boys, or Japanese stationery?
  • Why do people feel more guilt about blue-collar job displacement than white-collar?
  • What's scarce now: intelligence or physical materials like transformers and turbines?
  • Why would a transformer startup get funded today but not three years ago?
  • Will graphic designers be forced to monetize art on Substack instead of corporate gigs?
  • Is there craftsmanship left in software engineering, or is that identity dead?
  • Are we going to be arrested for driving cars in 20 years?
  • What happens when physical labor stops being the economic fallback plan?

If you're grappling with why automation feels different when it's visible, wondering whether nostalgia creates business opportunities in a post-scarcity world, or trying to understand why transformer companies suddenly matter more than SaaS startups, this episode will challenge how you think about the emotional response to technology displacing human work.

Listen now.

Catching Up On TV Habits

SPEAKER_01

Hey Nick, how's it going?

SPEAKER_04

KP, how we doing?

SPEAKER_01

Go ahead. What's what's new in your world?

SPEAKER_04

Man, I've got the house to myself tonight. Stairs out of town. And uh Gustu stayed with me this week. He is at my parents' house, so the world is my oyster, and I've chosen to hang out with you on the Zoom call. So I don't know about I don't know about I don't know about my life decisions right now, but um hey, this could be fun.

SPEAKER_01

You gotta catch up on some Netflix?

SPEAKER_04

I haven't watched a show in feels like a year. Like I just completely gave up on TV.

SPEAKER_01

It's weird, right?

SPEAKER_04

It's work, kids running around outside. That's all I got time for. Yeah. Yeah.

SPEAKER_01

It's kind of weird. Like I think generally too. I talk to more people, they're just not watching as many TVs, TV shows, much of anything. Um, yeah, I don't know. That's net Netflix stock seems to be doing okay.

SPEAKER_04

Yeah, I um, you know, my brother, so my brother is three years younger than me, and he's in LA, married, but no kids. And yeah, I feel like so just there's a there's a generation of of people that I think they still like they they use shows as a way to relax. That's what he tells me. Like they're you know, there's like a window of time every night, maybe eight, you know, eight to nine, nine to ten, where they get to indulge in shows. And it's funny. I heard you know, you know Paul Tudor Jones is the famous trader.

SPEAKER_02

Yeah, yeah, yeah.

SPEAKER_04

He was on this uh invest like the best podcast this week, which is like one of my favorite investment podcasts. And by the way, great episode. Totally everyone should listen if you if you joy enjoy hearing from good investors. He's he's the ticket. It's got a nice southern accent, so I like him for that. And but he he watches shows every night with his wife.

SPEAKER_03

Yeah.

SPEAKER_04

I watch Netflix every night, it's how I come down from the stresses of trading. Interesting. And anyway, that's that's like the that's what my brother tells me, and that's like what I get, that's the vibe I get from that generation. Is like instead of reading a book before bed or listening to the radio or turning on c cable television, you just flip on Netflix. And I don't know if you're spending your whole time like searching for the show or you're actually watching, but that's what that's what they do.

SPEAKER_01

But don't all the wellness people tell us all these gurus tell us about like the blue light and it's bad and sleep hygiene.

SPEAKER_03

Just gotta get the blue light, blue light blocking glasses.

SPEAKER_01

Yeah, like all the things like we're not supposed to be watching TV before bed, yeah, and all that. I mean, I just don't I I just don't watch much TV. I listen every once in a while. Instead of like if I'm I'm I've been kind of getting bored with podcasts lately. Really? Yeah, I've been getting really bored with podcasts. Yeah, um, and so I just yeah, so I've been like listening to shows and it's kind of funny because when I was a kid, when I was like 12, my parents bought me an alarm clock that actually tuned into TV channels.

SPEAKER_03

Okay.

SPEAKER_01

So I would fall asleep listening to TV shows through my clock radio. So it kind of feels makes me feel like doing the same thing.

SPEAKER_03

Like as a kid in bed, yeah.

A Peptide For Sleep And Focus

SPEAKER_01

Yeah. Speaking of wellness, you've been pretty active on the Twitters lately.

SPEAKER_04

Are you are you insinuating that my wellness is not not good because I'm active on social media, or are you are you reading my reading my peptide posts?

SPEAKER_01

Your peptide posts. I didn't know you were like, I didn't know you were such a wellness guru these days.

SPEAKER_04

We have a we should just designate at least uh the first five to ten minutes of this episode to talk about wellness. I feel like we hit on some esoteric wellness topic every time, but that just just you know, you're talking to me. I'm I can talk about that stuff all day. Uh yeah, so I had a post this week to jump into Nick Unpacked. I had a post about a peptide that I've been trying that I've had an amazing experience with. You know, not medical advice, not financial advice, all the disclaimers.

SPEAKER_01

But please don't please don't like any anything we say is any kind of advice.

SPEAKER_04

This none of this is advice, just two friends chatting. And but anyway, so first of all, I had so I posted about a peptide. I got a bunch of VC follows on Twitter because of it. Notably Mark Andreessen, Joe Lonsdale liked the tweet. Like there's this obscure investment community that I didn't realize was that interested in peptides. I expected the only people that interact with the tweet would be, you know, maybe some biohackers that I've come across over the years and that are that are friends of mine. But all right, so let's get into the but anyway, like sub subset of there's a group of investors out there that are very interested in optimizing their cognitive abilities via peptides. That's one takeaway. Now, onto the peptide. I think it's pronounced penial on the peptide that I that I've taken. I don't like it's like so it's it's it's something that affects the pineal gland, which is something that sets your circadian rhythm, sets you know, your the melatonin that's released essentially when you're you know about to go to sleep, it it it basically modulates that system and makes sure it makes sure that this peptide makes sure that it's run that circadian system is running at at full strength and operating correctly. So it like if you're not optimized, it's adjusting it for you and modulating in the background. The effects of that downstream are amazing sleep. So like I posted my sleep score and best sleep score I've ever had in my life. So like if anyone is tracking the Ora ring sleep scores, their Apple Watch sleep scores, I use a Whoop. I've been collecting data for about two years. Yeah, like I had a heart rate variability of over 100, which was very high for me, and a resting heart rate of 41, which is very low for me. I'm usually like a 45 to like a 70 on the heart rate variability. So, like, not slight improvement. I mean, like like you know, 20 to 30 percent improvement over I've been taking the peptide for about 14 days. Uh it's an oral peptide, so I don't mess with needles, like I don't do the injections, you know, maybe down the road when I'm old and want to take care of myself and do whatever's necessary. But yeah, so oral, it's it's literally just a couple milligram pill. And okay, now let me talk to you two about the effects that I care about.

unknown

Okay.

SPEAKER_04

Penal penalon. I've had I never I've had difficulty waking up. Like I never enjoy waking up. Like it's like probably my least favorite time of the day. I like to sleep in. And so, you know, usually in the morning I'll I'll lay in bed for you know 20, 30 minutes, hit hit snooze on my alarm. Probably something that a lot of people listening to this do. This thing makes you jump out of bed.

SPEAKER_01

Okay.

SPEAKER_04

Like, so that's the first thing. It's just like a noticeable thing of like, oh, I'm ready to go. Interesting. I haven't felt this in a long time. The other thing is it completely clears out brain fog. So, like, kind of something you don't really realize that exists until you know, after you eliminate brain fog, like maybe there's many mechanisms mechanisms to do it. And so through me taking this peptide, I realized that actually was carrying just a lot of brain fog, just accumulated stuff over the years. And I would how I would reduce it is I would meditate. I meditate daily, and or I'd go for you know, I'd work out. I'd I work out a lot, and so I've I've hacked my way around eliminating the brain fog and optimizing my cognitive ability during the day. And but over time, like that's like it's time consuming, and like you can't, like I can't, I got kids, I can't meditate for an hour every day, and I'm not gonna be optimal if I'm not. It completely eliminated the brain fog. Like I can't describe to you how much lighter my brain feels. And it's almost like like it's so much lighter that I'm searching for, hey, where'd that other stuff go that I'm used to feeling? Like it's un, it's like unsettling, unsettlingly nice. And yeah, it's just like the brain frog is gone, and I have like you know, just a lot more ability to focus because of that. Like when your brain fog is gone, you're not in between multiple things, you're not getting tired easily with cognitive tasks, and it's been great. So I'll stop there. I've you know, probably surpassed the five minutes that were allowed in this health segment. But if anyone is curious, reach out. I'm happy to chat with you about it. I'm I'm obviously very intrigued, and I probably will over time experiment with more peptides, but this this particular one has been amazing.

Finding The Right Peptide Stack

SPEAKER_01

Why did you like how did you stumble into the peptide sphere? And how like because there's a lot of them, right? If you if you're on the internet, you see there's there's so many, and they're either weird names or numbers, right? How did you decide that this was the one and this is what like you started? Because a lot of people, it's not just one peptide, they're taking all kinds of peptides, right?

SPEAKER_04

Totally, totally. Yeah, there's like the weight loss peptides that there's like the injectables, like technically some of the GLP ones are peptides, right? So yeah, I um well, first of all, I'm like in these niche communities. So I like you know, I follow Andrew Huberman and that that crowd. So I I I read about it, I'm curious. I I've taken coaching calls with people that are deep in the weeds of this stuff, like nutrition, you know, nutritionists, integrated medicine. So like I've like I've I've dabbled in you know in and around this space for years, and I'm generally like aware of just the research that's coming out of them on them. And then as a stepping stone, you know, the the most notable peptide that probably a lot of people have heard about is called BPC 157, which is a peptide that reduces inflammation, acute injuries. Like if athletes use it when they're recovering from like a torn ACL or any injury, they're you know, injecting it. Um, but there's an oral version that essentially just reduces inflammation and actually helps gut issues. Like I don't eat gluten and I don't eat I don't eat eggs just due to allergies. And so just yeah, gut issues, it shores up. And I've so I tried that about a year ago, and then I had some pretty good and I had some good effects. Yeah. And it just let so that was my my gateway, my gateway drug into the uh the cogn the cognitive peptides.

SPEAKER_01

I've been I've been having the classic 50-something year old rotator cuff issue. I don't know if you know about this. It's a very common thing. Yeah, it's a very common thing. So I've been dealing with that for a while, right? But of course, I clicked on some things, right? When you click on some things, there's all kinds of stuff, right? There's rotator cuff stabilizers, there's wave therapy. I actually bought this thing, you spin this little ball, it's supposed to help you build the muscles up. It helped a little bit, but that BPC 157 that shows up on my stream constantly. That shows up.

SPEAKER_04

I I think that like whenever you have an acute injury, from what I've read, like you can do, yeah, if you do injectables, like it's pretty effective at at solving the problem quickly. So if you're open to experimenting, I mean, I don't think it's like again, not medical advice, but I don't think it's that risky. I think it's like pretty pretty proven. And actually, I have I have a friend who's an orthopedic surgeon, and uh I I started talking to him about this one day. I just was curious on his opinion, and he basically was like, All of my clients that are coming to me are asking about this stuff. Yeah, and they he can't have a position on it, like he it's just he can't.

SPEAKER_01

So you know, I was thinking if we keep talking about wellness stuff, we might get a sponsorship from like AG1. Yeah, yeah. What's the other one? The eighth sleep, the pod stuff. I mean, I don't know, man. But if we if we spend like 10 minutes an episode talking about wellness, maybe we get flagged and somebody sends us money at the end.

SPEAKER_04

The po the po the peptide post I had had like 200k views, and like 20 people were asking, where did you buy it? Like, I want I want to buy this, where'd you buy it? And I was reluctant to share the link. One because I want it for myself, right? And then two because I, you know, I don't I don't want to promote influencer, yeah. Well, I don't want to I don't want to promote yeah you know random website that I found, but maybe it's not effective for someone else. Uh and I also want, you're like, hey, whoa, where's my yeah, where's my affiliate check? Where's my influencer check? Like, I'm gonna send thousands of people to your site. That's not fair. No, all right. Well, that's cool.

SPEAKER_01

Yeah, appreciate it.

SPEAKER_03

All right, let's show you.

SPEAKER_01

The first time I saw that post, I was like, I better call Nick. His Twitter account got hacked. I was unsure that it was you.

SPEAKER_04

Yeah, I've been experimenting out there with talking about like Frank Lloyd Wright.

SPEAKER_00

You're talking about like Frank Lloyd Wright and robotics and then peptides. I was like, what happened?

Hardware Investing Gets Serious Again

SPEAKER_04

That's right. We have a we can go deep on many topics here at Shout Aventures. Yeah. Anyway, on to the usual topics of the show, built environment, technology. We want to talk about physical physical AI. Does everyone know? Yeah, do you think our listeners know what that that term means?

SPEAKER_01

Yeah, I think you know, popularized by NVIDIA. Once again, if you if you only have so much time to listen to anyone out there talking about AI, the only person you should make time for is Jensen. The guy's just good. It's interesting. He's just he's right. Yes, is he trying to sell more NVIDIA stuff? Yes, but I'm also unsure who he's competing against. Like I don't think it's purely just promotional, I think they're doing okay. However, I think he tends to have the most concise ideas around where we're at and where things are going. And he's been he's been right a lot over the last few years. But I think I think what stood out to me today, right? I mean, I think we've spent so much time in the world of software for decades, and we've done some hardware investments, right? We've done, I mean, honestly, my first and my first investment was icon 3D, right?

SPEAKER_04

Which is hardware, which was in twenty eight 2018, right?

SPEAKER_01

2018. So before it was cool, so to speak. And we've made other hardware investments along the way, so it's not like we've been afraid of hardware, but a lot of VCs just kind of say, like, oh, we don't do hardware, we don't do hardware. Yeah, and it was interesting. I was at lunch with a well-known VC today, and he was kind of going through some of the latest deals he's done in hardware, and we got in this conversation. He was like, Yeah, because he was talking about building his team, he's got three new hires starting, blah, blah, blah. And he was like, Yeah, you know, I've thought about hiring like a PhD type to help me with technical vetting. And I said, like, well, I'm don't you just want to hire that as a consultant? Like, one PhD can't know everything about everything, right? And he said, like, that's not the problem. He's like, the problem is most of these people will tell you it won't work. And he's what, but the deals that we want to fund are the ones that everybody thinks won't work. The ones that are like a guarantee will work may not be venture fundable.

SPEAKER_03

Totally. The price is too high, yeah.

SPEAKER_01

Right. So um so that was kind of interesting. But the thing that stood out to me today in the world of hardware was caterpillars at an all-time high and Salesforce, ticker CRM not doing so high.

SPEAKER_02

Right.

SPEAKER_01

So if you would have said two years ago, man, bet on caterpillar, forget about these SaaS companies, go bet on caterpillar. You people would have probably thought you were a lunatic. But what I think is interesting, you know, being in venture, we've always, you know, we always talk about can this SaaS startup disrupt Salesforce, Procord, whoever, right? Are they going to be able to dethrone the incumbents? Which I think is also like really fascinating. I was I spoke at a conference last week called Building Transformations. It was over at Hensel Phelps Innovation Center in Phoenix, and I kind of gave a little keynote, some insights into what we were building at zero R5 from a tech stack perspective, and I got just a ton of feedback, which was cool. Yeah. But you know, one of the things I brought up is Autodesk, who everybody loves to hate on, right? It's just easy. I was like, you know, they made a$250 million investment into world labs. Like all these startups should be worried now, right? Procore is making some moves. So these people, these firms, these old incumbents in software that we've always thought about our startups disrupting, they're actually making some really good moves. Yeah, which should concern you because a lot of startups are like, I mean, Procore, they're the worst. We're gonna disrupt them because they're so terrible, blah, blah, blah. I mean, maybe that's not the case, right? But there's there's a formula to disrupting the software companies, and a lot of that is the capital moat to go build these things over the years has been not so much money. Right? So if you say, okay, well, we're out of this cycle, software has become a commodity, blah, blah, blah, right? If we believe in that narrative, and we're saying, like, hey, let's go invest in hardware. How do you disrupt an incumbent like Caterpillar? One, it's capital intense, it's manufacturing from an in, you know, there's RD, but there's also manufacturing, right? Think about what happened with software. The minute I stopped having to ship CDs and put it on a server and do all those things, all these things, and I could just spin it up in the cloud. That was a big unlock for the SaaS economy, was just, oh, you need a hundred more customers, spin up some more virtual servers in AWS, right? So the unlock there was versus in the 90s when I was having to make my own servers and install them in rack in data centers and pay for rack space in data centers. All of a sudden AWS shows up and you don't have to do any of those things, right? So we de-risked and took the cost of entry way down with AWS. And then now with AI, we've taken down the software development cycles way down, right? So, but in hardware, you got to go design and make sure the thing works, right?

SPEAKER_02

Totally.

SPEAKER_01

Yeah, you gotta get it like certified, and you have to build a manufacturer. And so, by the way, you know, RD is very different than manufacturing. Like you have to go to a warehouse, like our friends at Andrew. I think what they're up to two million square feet now, manufacturing stuff. That's that's a different thing, right? It's a different thing. When the minute you have loading docks, I mean, like me and Barry, Barry Clark, you know, we worked at software automation and we our office space was like a lab, right? We had loading dock doors and stuff, it's a different business. And so when you look at the moat that companies like Caterpillar or John Deere or any of the big the big hardware companies have, you know, how do we even look at how is a startup gonna disrupt them?

SPEAKER_03

Yeah, super hard to displace.

SPEAKER_01

Right. And then the only one that we have a really good benchmark against is Tesla, which was not easy either. That was not that was not a that was not a straight line either.

SPEAKER_03

That was not a layup, no.

SPEAKER_01

Yeah, it was not a layup. I think it's been well documented. Elon Musk almost went bankrupt.

SPEAKER_04

You could also say SpaceX is there too. So in both cases, you know, they were competing with Tesla, you know, competing with traditional auto, you know, uh automakers, and then SpaceX, obviously, with anyone who was building rockets before, and the the only way to beat them was considerably lower costs. And it wasn't like, hey, I can I need to be cost competitive, it was like I have to be much, much cheaper in my manufacturing processes. It's the only way they won. But it required it required an entire re-architecture of both systems, right?

SPEAKER_01

Rockets that's the thing, it's two different levels of innovation, right? You have to innovate on the RD side and you have to innovate on the manufacturing side.

SPEAKER_03

Yep.

SPEAKER_01

And then I get so much to overcome.

SPEAKER_03

Yep.

SPEAKER_01

I totally agree with capital needs, right? Capital for RD, capital, you know, got to get a real estate loan, right? Got to get a mortgage to go or get lease space. Maybe put up a letter of credit to be able to be worthy of leasing the space. And then we have to build inventory, which means we have to have parts and fund inventory. You know, by the way, we probably have supply chain dependencies. All of a sudden you're fighting, you know, you're you're you're Tesla and you're fighting for batteries with everyone else, with Hyundai and everyone else, right? Like it's it's not for the faint of heart. So I think when VCs, including us, say like, hey yeah, we're done with SaaS, we're moving to physical AI. I don't know that most of the VCs understand it. Like, I don't think I don't think a lot of them understand themselves. The capital stack and how to pull this stuff off. Yeah. I'm sure clearly there's some that have been successful at it. But I think that's also why classically most VCs have said, oh, we don't do hardware.

SPEAKER_04

100%. Yeah, there's a different underwriting process entirely. There's a different capital structure that should be there to support it. There is a different skill set required from your VCs in a lot of ways. Yeah. Like what do you need from your investors? It's probably not just software strategy advice and product advice and higher hiring recruiting, right? I mean, those are all still true of hardware, but you need financial sophistication to pull off the financial engineering to make it possible. You need the technical sophistication to make the right, you know, if you if you want to, you know, if you want to double check that you're making the right decision with someone or in your circle, ideally you can call an investor and ask the, you know, ask them, ask them questions. Now, I don't think Elon's a unique, unique individual. I don't think he was relying on his investors to make the right technical decisions. But, you know, when you look at the scale of founders who are going to need to make really important technical, technical calls and they're building hardware systems that haven't existed before, I don't know. I think it's kind of a baseline, it's a baseline need.

SPEAKER_01

Yeah, I mean, I think we've always said classically to be in venture capital, having an MBA is not a requirement. In fact, it could be a problem because you could over, you know, investing in a seed stage stage SaaS company. I'm not convinced, you know, how much MBA work and experience is needed. Right. Um, how important is that? Probably not. I mean, how much time do we spend doing discount discounted cash flow analysis and any of those things? None. And by the way, Claude can do all that stuff for you now. But I think in the world of hardware, you do need to have a level of expertise, you know, to understand truly, you know, what does the balance sheet have to look like? How do you think about RD spend? How do you think about equity? How do you think about debt, you know, whether it's permanent debt, you know, whether it's mezzanine debt. I mean, it's just it's really complicated, right? But I think, you know, and and we've been spending a lot of time specifically with one of our portcos to really, I would say, help them. But I would also argue we're kind of learning along the way, too, right? Um in terms of how they're thinking about acquisitions and things like that. I feel like we've had to learn along the way, just as the founder was learning along the way.

SPEAKER_04

Yeah, 100%. Yeah, and I think I think partly we're learning, and this is a new space that a lot of VCs are gonna, I think, struggle to find their footing. Because it's never like it's never really been done in venture before. There's never been a a class of hardware investors beyond the people that were early SpaceX, Induel, and Tesla. Like those are the those with the those are the winners.

SPEAKER_01

I would slightly disagree with you. Go on, mostly being that Silicon Valley got its name from the Intels and the Fairchild Center.

SPEAKER_02

Sure.

SPEAKER_01

Yeah, right. I mean, rooted in hardware. In fact, many of the classic venture folks believe that the whole point of venture capital is to put that risk capital really around technical risk, right? Less so about market risk was kind of the old, you know, it's like if you can actually build a processor that's 10 times faster than the other guy's processor, we don't have to do any market analysis. We know it'll sell, right? The question is, can you actually build it and keep it build it for a price point that you can actually make? Is there enough market for it to make a margin on it, right? It's not that there's not demand. Yeah, it's kind of like biotech, right? If I mean if you can get those two molecules to bind and find a cure for cancer, I mean, you probably don't have to worry about the market sizing, right? It's more the technical risk. And I think somewhere along the way, venture shifted from being 100% about technical risk back in the day, right? The Don Valentine days, to being about business risk and like go-to-market risk. And I think now you're seeing maybe a shift otherwise to start saying, like, well, even what I'm doing with zero around these AI roll-up strategies is you know, as venture, we look at this downside risk, right? Like they might go to zero. And private equity doesn't optimize for downside risk, they look up, they optimize for upside opportunity. Yeah, right. And so no, so I think it's it I think in many ways we're getting back to basics, which is we might invest in something where they never even get it to market because they couldn't get it to work.

What Physical AI Actually Means

SPEAKER_04

Yeah, which has happened before for sure. Yeah, the I've seen wide-ranging definitions for physical AI. You mentioned that Jensen kind of created the the word and the label. What's what's your definition for it? Like when you think of that as an investment category, what does that span for you?

SPEAKER_01

I think it's uh so I don't think you can throw just hardware and you know if you're gonna go build a new type of race car, I don't think that's physical AI, right? I think it is the idea of the intelligence interface being AI, but manifesting its value creation in a physical space. Yep. Right. So I think it it's not like, oh, it helped me write code fast. Like that's that's different. I think it's really this idea that AI, the AI aspect of it and the value creation aspect is is in the physical space, not in the soft space. Um it's interesting, I was talking to someone the other day, and they were talking about cognitive cities and cognitive buildings, right? Which I thought was kind of okay, you people are just like coming up with new words.

SPEAKER_04

Sounds sounds like an evolution from smart cities and smart buildings.

SPEAKER_01

That's exactly what it is.

SPEAKER_04

Smart to cognitive. Good uh, yeah.

SPEAKER_01

I'm sure it was IBM. Someone at IBM, because you know, they launched the first like smart cities. I think they owned that for a while. That's very funny. But no, I I think that's I I think that's what we think about like physical AI, you know, because I think if you're just gonna go build a better bicycle, cool. That's just that's just physical as a product. Yeah.

SPEAKER_04

Yeah, I think that when I first heard of physical AI, the first this was a couple years ago when he coined it, but the first thought that came to mind for me was humanoids, because they're essentially in intelligence embedded into you know human form factor, physical form factor. You know, another world for word for that is embodied AI. So I think like those are kind of interchangeable with that word word. But I think for us, the the applic applicability for the built environment is actually much, you know, it's it goes well beyond just just humanoids and just you know, and and and just into AI embedded into the into into the physical form. It actually can, you know, it's AI potentially embedded in into industrial processes. So if you had a specialized robotic system, like an automated architecture or an icon, right, they've invented new robotic systems or they have specialized components on the robotics, and you're embedding new layers of intelligence into that industrial system, that is also physical AI. Like if you're doing things that you could not do before, or there's like a, yeah, there's a there's a new there's a new data set that you're now able to to leverage, to simulate, to get a lot more accurate. That that that to me, that's also like included in the definition. And so that that physical AI umbrella, I think actually covers a lot of what we're gonna be investing in over the next 10 years, is you know, anything that's that's embedding intelligence into a physical forum, whether it's an industrial robotic system, Okibo, our icon, those types of products, and this just gonna continue to layer more and more data and intelligence into their systems. And yeah, and just a lot of lot of a lot of other autonomous equipment. So Lumina would be another good example because they're gonna embed intelligence into their into their heavy machinery. So full fully autonomous dozers, fully autonomous excavators, but it's the it's the it's the AI component that allows them to be autonomous, and that's you know, obviously embedded into the into the hardware.

SPEAKER_01

Yeah.

SPEAKER_04

What do you think?

SPEAKER_01

So I was listening to this interview from John Fort with CNBC to this yogi named Sadhaguru, who who I've met several times over the years. He's fairly well known and mainstreamed. And I think the last time uh I had an audience with him, he he made the comment like, I don't know why everybody's wearing all these wearables and smartphones. If you actually listen to your body, it'll tell you everything you ever needed to know about what's going on with your body. But none of us paused long enough to really look within the so he's kind of a spiritual guru. And John Ford, who's CNBC, right? He's not uh he's not in the spiritual business, right? He asked him at this conference, like, you know, hey, what do you think about AI? And you know, he had a couple funny comments, you know, because he's kind of funny, like he's pretty well attuned to what's going on in the world. And he's like, look, you know, it used to be to be the most powerful person, you had to be big and strong and muscular. And he's like, now we we call that a security card, right? Like, that's not the the power dynamic of society has changed because of these technologies. I think it was him or was someone else that said, look, if as humans we've been trying to fly and we invented airplanes, but we still can't fly as humans, right? We still can't do it. Like nobody can flap their arms and fly. Sure, we can just build all the things, like we'll never replace a bird, was kind of you know the the idea. But you know, the the commentary is like, what are we all gonna do? Right, what are we gonna do when AI does everything? And it was funny. I just started I'd just been listening to that, and then I went out to some restaurant tea, and of course I get there five minutes before they're closing, and there's like a woman pushing a mop, right? And I was like watching this. I'm like, okay, like would because you know, I lived in India for a little bit, they don't have mops. People get on the ground with a cloth and wipe with their hands, right? Which is backbreaking. And I just was sitting there watching her mop, and I was like, could she imagine a world where she didn't have a mop? Could she imagine a world where she got a bucket and had a rag and had to sit there and get on her hands and knees, you know, Cinderella style, and like scrub the floors. And so I think it's it's interesting because when we think about some of this physical AI stuff, we're acting like uh as humans, we've never created tools. Totally. Yeah, we've constantly created tools, and in many ways, I think some of those tools created an abundance, which has been problematic, right? In other words, if you can mass manufacture food and get you know, mass, you know, essentially mass manufacture calories, right? You know, we talked about AI factories, you know, the food industry is calorie factories, right? They're generating calories for us. That got over optimized, right? I don't think any of us believe that you know how fruity pebbles are created and how food is manufactured these days is a good thing. How mass farming works is a good thing, pesticides are a good thing, but those were all innovations. If you think about a pesticide, it was a major innovation to kind of feed feed the planet, right? What Monsanto did with their wild seeds, right? So those are all innovations that created abundance. Well, it turns out maybe too much abundance is bad. So I think it's interesting when we talk about the physical AI stuff, is how much of it's just gonna be like mainstream. We're never gonna question that like it, like, oh, that took someone's job away. Right. I mean, I I think there's there's just a lot of things we over, I don't know, we over-index into the future thinking that we have some like, oh my god, what are people gonna do? Yeah, and and people people will do different things, yeah. But along those lines, right? I had a conversation with a friend yesterday about Claude's new design capabilities, you know, and of course, you know, the Twitter sphere, right? Oh my god, Claude just announced XYZ. It's insane. All graphic designers are toast, right? Like that you you see all those narratives on Twitter. But what's funny is like, I mean, you know, some graphic designers. How many of them do graphic design for corporates? Because that's what they want to do. Every graphic designer I know is like in a band, has some other like they're doing graphic design, yeah, to make a living, but they generally hate it. They you know, there's nobody at Coca-Cola making posters and flyers as a graphic designer that loves that. Chances are they're sculptors, they're in a rock band, they're doing stand-up, they have other artistic yeah.

SPEAKER_04

The core substrate and a graphic designer is artist, just like core substrate and a copywriter is writer.

SPEAKER_01

Right.

SPEAKER_04

Like they probably want to be. If you're a copywriter, like you probably want to be writing some, you know, whether it's a book or novel or you know, journalism, you probably don't want to be writing advertisements, is is uh is a safe assumption. But yeah, yeah.

SPEAKER_01

So does it become a forcing function for people to take those risks? To you know, I was listening to Tetogrammaton yesterday, and it was like the founder of Substack, right? Oh yeah, yeah. Yeah, yeah. Maybe you work on being a Substack writer and you monetize your art that way, right? So so I think like the physical AI stuff, people are seem also very fixated on, oh, it's gonna take jobs away. But I I don't I don't know how that's going. I don't think we any of us know how it's gonna work out.

Construction Robots Spark A Backlash

SPEAKER_04

I feel like so. I I have I have some interesting commentary on this because I mean to your comments on on Twitter and posting, I've been posting robotics content on Twitter from some of our portfolio companies. So last week I posted a video of Okibo. And for anyone who doesn't know, Okibo is a shadow portfolio company that does interior finishing, fully autonomous. It's a so they do they focus primarily on wall finishes now. So dry drywalling and painting, they essentially you essentially put the robot in a in a space, it can reach up to 24 feet, which is obviously much higher than what humans can reach, much higher than 12 feet KP. I think you're you had a question about that.

SPEAKER_01

Yeah, it's much much higher than who's that company? Canvas.

SPEAKER_04

Yeah, I don't know.

SPEAKER_01

Yeah, yeah, JLG, they bought them.

SPEAKER_04

Yeah, they're they're they're great, they just can't go up to 24 feet.

SPEAKER_01

They're very short.

SPEAKER_04

So anyway, I posted this video of Okibo, just the machine working, because we were there with them last week. 300,000 views. And the reason I it got such a high view count is because it was controversial. People were, you know, half, you know, half the people were like, amazing, like, you know, we're gonna get to reduce cut construction costs, like the work to your point, the workers will be able to do more meaningful work and like less backbreaking work, like hanging drywall and taping is tough. Like, who actually wants to be taping drywall? Like, no one, like ask the drywallers, they don't want to do that. And so, but there's there's of course the other contingent that's like, oh my god, they're coming for the physical jobs too. Like, we've just we've already we're already having this AI conversation, and you thought you were safe going to work in the trades, going back to trade school, robot that's doing fully autonomous trade work, and it just created like this firestorm. And so, like, I saw all the commentary, I had a couple hundred comments, and I think there's something to the physical form that is different from AI, which is the digital form. And when you see a blue-collar work worker be displaced or not necessarily be used for the same sort of work, it creates a level of discomfort as a human, as a human being, because you're you're now wondering, like, okay, what is my value? Like physical labor at the end of the day, if we all went to zero with knowledge work and white-collar work, it's always physical labor. Yeah, well, go get a physical labor job. Well, in the future, like, that's not true. It's not going to be all physical labor. There's always going to be, you know, some, but it's not going to be for everyone. And that reality, I think, is tough for the human to accept. And I think there's also the demographic dynamic where like blue collars, blue-collar workers, people feel bad for. They're not making as much money as the white-collar financial analyst that's being displaced by AI. But yeah, so like people feel a different level of guilt around you know, displacing them and taking their jobs. Now, to your point, we've always tried to use the best tools that exist to push humanity forward and build a world of abundance. We did it with machines. So mechanic, you know, mechanicization mechanicization was a big part of the industrial revolution. And were there arguments about taking human jobs and reducing the amount of humans in a factory? Yeah, 100%. Yeah. Were we better off for it? Yes, 100%. We got more things, like we, you know, we advanced society, we advanced civilization, and we were doing less tedious manual work that, like, I don't know, most people probably didn't want to do. And so it's just like getting to robotic autonomy is just that next phase. And to immediately think that those humans are not going to be leveraged on a on a job site or leveraged to still do physical work in and around that activity is silly. Like, humans are still going to be in a factory. We're going to be like they're like, we'll be around, we'll be operating the machinery. Like there will be use cases, especially like, you know, in the immediate future, like over 20 years, like, I don't know, maybe humanoid is actually just on the job site and we're doing other stuff.

SPEAKER_01

Yeah. And and my my point, right, was like, okay, so you know, we're all we're all very fortunate people, right? So next time my cleaners come by the house, I'm gonna take away their mop, give them a rag, and tell them to like what hand wash all the dishes, and they're gonna look at me like I'm insane, right? They're like I'm like, no, but I'll pay you the extra hours to hand wash and hand dry all the dishes and to scrub the floors by hand. I'll pay you more. I don't think they're gonna care. I think they're gonna walk out if I did that, right? In other words, I'm gonna take all this these tools from you, do it the old-fashioned way. I'll pay, I'll pay the same, I'll pay by the hour anyway. I already pay you by the hour. Yeah, you get to work more hours, and doesn't that make you happy to work more hours and make more money?

SPEAKER_04

I think I and it's a it's a great, it's a great response to that argument because that's exactly what it is. Yeah, like it is using an inferior, you're opting to use an inferior tool to complete work if you're not leveraging the best system, which is going to be obviously robotics.

AI Anxiety And Losing The Craft

SPEAKER_01

Right. And so I think, you know, and so I was thinking about like the graphic designers, right? Like, is it does it become a forcing function for them to try to like find joy in what they actually have meaning, you know, what what they believe is the things that are interested in. Does it it's kind of like if you're gonna side hustle, you know, we used to say, like, if you're gonna side hustle your startup and never quit your job, I'm not interested in funding that, like, you're not all in. So does some of this stuff maybe push artists to go all in on what they really enjoy and do and try to monetize that, which is kind of interesting. But then I was talking to one of my founders, one of our founders, and he said he had one of his software engineers quit. And I was like, really? Like, what happened? I was happy. He said, We started doing more cloud coding, and he despises AI coding and quit. Yeah, this this is happening, right? Is leaving the tech industry.

SPEAKER_03

So existential crisis, yeah.

SPEAKER_01

What's that for? I mean, at least like my friend that's a graphic art every graphic artist, amazing, and she's in a band. So if in her world, if she got fired, she'd be like, Great, I'm gonna go do more gigs and try to make make a living playing music, right?

SPEAKER_03

Yeah, yeah.

SPEAKER_01

What's the software engineer guy gonna do?

SPEAKER_03

Do great question. Like at least I go be a baker.

SPEAKER_04

I don't like I don't know, right? Depends on what, yeah, what do you what's what's the core interest?

SPEAKER_01

But it was funny, is like I feel like to me, like the the what we have out there in terms of coding has kind of made coding fun again. And for him, it's made it like not fun.

SPEAKER_04

It's it changes so there's this if you're no if you're a software engineer pre-AI, there's a level of craft with it because you're actually writing the syntax, and you're thinking you're you're you're thinking and creatively coming up with the right syntax in the right order, using the right commands. You don't have to do that anymore. In fact, like you're at a disadvantage if you try to do that, and that's probably why that's probably his realization. There's an amazing article about this. We should link to it in the show notes. Like we have show notes. I don't know if we've ever done that. Send send it, send Angela a link repeating what I hear on other podcasts. Yeah, and the show notes that we use for this episode, it's called We Mourn Our Craft, and it's written by a software engineer. It was written earlier this year, I think in February. And it's just about the crisis that exists and like the decisioning on the accepting reality and that it's still painful, there's an identity crisis, you have to go through it, and you also have to learn how to embrace the new tool to move through it if you want to stay in the career. And so I and it frames it really well because it's written from a software engineering standpoint. The guy's a the guy's an engineer. But yeah, like if you don't want to, if you still want to do craft work, and I don't know, I would argue like getting you know 10x faster engineering through, you know, some of the agentic tools out there and Cloud Code and Codex, like there's still some craftsmanship to that, getting the right output, yeah, and moving quickly and you know, integrating systems and all of that. But if but if you're seeking like true like written syntax type crap crafts work, where you're you're a graphic designer that's like a graphic designer comp is like, are you using AI art? Are you using mid-journey? Are you creating from scratch every time? And there's like the graphic designers I know are facing the same existential thought.

SPEAKER_01

Yeah, they don't know. I was trying to think about it, you know, because I'm a big vinyl guy, right? I love the nostalgia of it. I love the experience of vinyl, right? I love the experience of vinyl. And people can argue about whether it sounds better. People like to make all these arguments. Half the time I fall asleep listening to music, so I'm not sure that I'm tuned into the sound quality, but love the experience of vinyl. Do we think there's gonna be a retro, like a retro aspect of like coding, right? It's like you know, Palmer Lucky has that company, what's it called?

SPEAKER_04

Retro something, Rod or Mod Retro.

SPEAKER_01

Yeah, mod retro. He brought back the Game Boy, right?

SPEAKER_04

Yeah, yeah.

SPEAKER_01

It's not a better video game, right? It's the same, it's Tetris, right? But but we people love it because of the nostalgia, right? So I'm just trying to think about that. Like, what is what's the nostalgia for coding? Like, people are gonna like, oh, I'm gonna go do some vi I'm gonna go do some visual basic. Like maybe, maybe, maybe like I don't know, right? Like it's it's like you walk into like I love going to these like Japanese stationery stores. If you were gonna do these, yeah, and they have like hand pressed paper, sure, they have quills, they have watercolor, it's amazing, right? It's like we don't need any of these things, right? I could go to Office Depot and get much better paper, yeah, yeah, much better pens, right? Terms of functionally, but it's it's kind of I was I was just went down that rabbit hole to be like, what is a the nostalgia of software programming gonna look like?

SPEAKER_04

Yeah, it's interesting to think about the I mean, gosh, even just the last 10 years, like like think about libraries. Like libraries have been have been completely dematerialized. You can access all of that information on the internet at a in a moment's notice.

SPEAKER_01

Like why do you have to go to Half Moon Bay? We have an amazing library.

SPEAKER_04

Yeah, which you showed me.

SPEAKER_01

It's amazing, yeah.

SPEAKER_04

But like you don't need to go to the library. No, you go there for you know the experience to your final comparable. And uh I yeah, I think my guess is that there's gonna be a surge of nostalgic activity, human activity over the next couple decades when yeah, we have the ability to access anything that we want at our fingertips, intelligence, information, physical forms, you know, you can build a house, like you can build a house in 10 days. So, what do you spend your time on and how do you find meaning? I don't know.

SPEAKER_01

I've I think that's such an individual thing, right? I'm for me, it's just always like pushing to the next thing. Like, what's the next the next thing I can do? Right? Like if I can build a house in 10 days, I'm gonna try to figure out how I can build one in five days is twice as big, right? Like, it's just always like yeah, more, yeah. My you know how I live, more is better, better is not better.

SPEAKER_04

Always on the chase, always on the chase.

Nostalgia In A Post-Driving Future

SPEAKER_01

Funny, the nostalgia thing. Yesterday, Devin was teaching our little one. If a car has three pedals, what are those pedals? And he's repeating a gas, a brake, and clutch, and then what kind of car is that called? And he's like, a manual, and what if it only has two pedals? It's a gas and a brake, and it's called an automatic. This is what they were talking about on a walk the other day, and I just thought it was so funny. Like, he, I mean, Ronan might in by the time he gets to drive, there may be no pedals.

SPEAKER_03

I don't think we're gonna be allowed to drive.

SPEAKER_01

Yeah, no, it'll be considered dangerous 100. Like, how dare you?

SPEAKER_04

Data already shows it. Yeah, we're I I think it's just a matter of when we're outlawed from driving, for sure. Yeah, it's crazy.

SPEAKER_00

We'll all be rogue rolling our cars out in the middle of the night going for a drive.

SPEAKER_04

My god, yeah, we'll get arrested for driving. That's a crazy thought. Like, oh my god, we have to arrest this human. He's out, yeah. He's out driving driving a car. Way too dangerous.

SPEAKER_00

Very dangerous.

Why Physical Materials Are Scarce

SPEAKER_04

So, one more thing on the physical AI front. So I think an interesting observation and a big part of our thesis from an investment standpoint moving forward is there's been a shift in terms of what's what's scarce and what's abundant right now compared to 10 years ago. What was scarce 10 years ago was intelligence. We had to go build the software systems that were, you know, had a had a baseline level of intelligence that could compute a lot faster than we could compute. And but we had like there was a there was a shortage of engineers to go build that's that that intelligence, right? And even a limit to how intelligent and cognitive that system could be, just given that you was, you know, largely like these are distributed databases doing computation, right? We're now in a in a world where intelligence is being commoditized, and that's abundant, and it's only gonna get more abundant from here. It's gonna far surpass human ability. What's now scarce is physical material things that are that can be used for the build out of that world and for the build out of post a post a post-AI world, a post-scarcity world where things are more abundant. Like we can we can produce more things because we have the intelligent systems to do it. And so when we think about investing, like if you believe that thesis, that the physical form is now scarce, like an example would be all the components that you would need to stand up a data center connecting to the grid, transformers, turbines. So you got, you know, you got infrastructure, you got building components, you got power, you got fuel, right? Electricity, all these things. Every single attribute on that value chain that's needed to stand up those those data centers that commoditizes intelligence is scarce. And that's the stuff we've been investing in for the past seven years at shadow. And when we look at the future, that's the stuff that will remain and continue to be more and more scarce. Like look at the transformer backlog. Yeah. Can't buy a can't buy a turbine today if you you know had if you had if you had your your checkbook out, can't find one.

SPEAKER_01

Yeah, there's there's a guy here in the Bay Area that started transformer companies called IR A Y R. He's building he's R D and creating next generation transformers, right? So I think about that and I'm like, okay, three years ago, if he would have pitched Shadow, yeah.

SPEAKER_04

What Transformers? Transform like what doesn't yeah, doesn't Siemens do that?

SPEAKER_01

Doesn't Schneider do that? Isn't that done? Haven't we already done that? You know, and and so it's fascinating, and like if of course he p I mean we're we can't, I think he's doing his series B now. I think we're out of the zone, right? But if he pitched us today, we'd have been we'd be all over him. And so I think like this idea that I mean, I I think to your point, like I don't I don't know that intelligence is gone, right? I think I think we still have creativity, I think we still have different ideas and the application of these different ideas, right? But I think what it's what it's doing is really raising the bar around entrepreneurship. Like you just can't be like, I'm gonna build a dating app where I'm gonna make the Uber. I mean, you know how many pictures we get? The U We're building the Uber of, right? I mean, we we never did any of those things, but some people did and did fine at it, right? But we're always like, nah, like I don't get that, right? So I think in many ways it's gonna push people to come up with better ideas around any of this stuff. But no, they I mean, look, the physical world is is where we live, and I think it turns out the the TAM of the physical world is increasing because we're gonna be on the moon. Yeah, I always joke around about like icon when they pitched me. It was like, this is the TAM. It didn't it did not include the moon. So strange.

SPEAKER_04

It's gonna be wild when that's the biggest segment of their business.

Rapid-Fire Wrap And Disclaimers

SPEAKER_01

Well, cool, man.

SPEAKER_04

Cool. That's probably enough for today.

SPEAKER_01

Yeah, peptides, everybody peptides.

SPEAKER_04

Yeah, reach out. Not medical advice.

SPEAKER_01

All right, man. We'll see you. All right, so