Today I am joined by Sam D’Amico, Founder and CEO of Impulse Labs. Sam and his team have created the Impulse Cooktop- the most powerful and precise stove ever made due to its integrated battery, which doubles as a grid scale energy solution. Sam and I discuss his long history working on hardware related technology like AR and VR headsets with Google Glass and Oculus, how he got the idea to revolutionize the stove, hardware company business models and supply chains, and his thoughts on how AI will shape the hardware world in the next 5 years. If you’re interested in cutting edge technology, hardware, or even cooking, you won’t want to miss this episode! You can learn more about Impulse Labs and Sam here: Website | Impulse Labs LinkedIn | Sam D’Amico
Note: This post may contain transcription errors
Noor: Well, very cool. Today we have Sam D'Amico, the founder and CEO of Impulse Labs. You have a really badass hardware background. You built a lot of different products, so why stoves?
Sam: Yeah, so for folks who don't know anything about me, spent most of my career working on air and VR headsets. Started actually with Google Glass, if you're familiar with that program.
From over, I can't believe it's been over 10 years. Um, it was early, early. It was, it was maybe a little too early, but kind of showed the way a little bit and then spent a bunch of time at Oculus working on a number of different programs from like the first like commercialized headset that they launched once they became part of Facebook to various prototypes to leading a program up until 2021 over there.
I've always been into cooking and like in San Francisco, host barbecues and stuff in my backyard, you know, from time to time. I got kind of obsessed about cooking pizza really fast and I was like, it would be really cool if you get the same performance as a brick oven in a tabletop device. And this led me down this entire path of like, well, that's something that's fundamentally power limited, but what if there was another way?
And it made me realize that there was this opportunity that was much larger than, say, cooking pizza, which was if you could embed a high power battery in every appliance. You could actually reinvent the entire appliance industry where you'd be able to make fundamentally higher performance devices that could exploit the fact that they're intermittently used and the battery was there to boost the power well above whatever the electrical system in the home could provide.
That kind of got me down this path. I realized very quickly that the gas stove culture war, even ahead of all this controversy about gas stoves and all other stuff, that was gonna be the thing that was gonna be the toughest to crack because gas stoves are just like the best stoves as everyone knows.
And so realizing that this was actually a tool to crush that debate once and for all. Convince even the most recalcitrant people to swap their stoves out for something far better, you know? And, and that would be the entry point for a whole new class of home appliance. Now that's kind of the first part of it.
And then the second part was, well, I've just installed a battery in your house. And so once you've actually built this network of batteries across potentially every home in America, you can largely solve. The distributed energy needs for the whole grid and to, to kind of put a definition on that, the, uh, the, the grid's changing fundamentally right now, new sources of energy are intermittent, so like solar, wind, et cetera.
Our distribution system is out of date and increasingly aging and new demand is rising massively. So basically, uh, everyone's getting an electric car. They're putting electric car charger in their garage. Their appliances are going electric. They're getting AC because it's hotter out. All of these things are happening and basically gonna lead to a doubling of electricity demand.
But our distribution system is not meant for it, and our supply is gonna be intermittent. So we need to actually deploy batteries to solve this problem, ideally as close as possible to the homes or inside the homes if possible. And, uh, we can actually do this solely through appliances, having the batteries built into them, which is kind of a crazy statement to say, but that.
Is the foundation of what we're doing at Impulse.
Noor: Yeah. Yeah. That's super interesting. But I get, I dunno, I guess just backing up, I mean, you could have built anything, you know what I mean? Like, it's not, I dunno, I feel like it's pretty rare, you know, to have so, such deep experience building, you know, so many different types of headsets.
Was there like a real search of like, okay, do I want to do, I don't even know, there's like so many different physical appliances that, you know, you could build with that skillset. Was it just like this obsession with pizza and the power grid? Or like, was there like a, a search space of like oth other stuff that you were considering?
Sam: So I think there's kind of this interesting attitude of like, okay, so you need to pick on, I mean, if you look at Elon Musk in senses, like you need to pick on an unsexy industry that like people have neglected to give a dang about for a century, basically. If that makes any sense. Yeah. Um, maybe not a century, but like half a century and appliance industry is great like that.
We've already picked on cars, we've picked on kind of the legacy defense contractors, especially the ones that are building rockets. I think there's actually kind of two or three things that we realized. One was, there was an opportunity to innovate on the user experience from like a, Hey, let's solve the, like there's too many buttons and honestly you just need like two or three and then, and make it very clean and obvious and solve the problem.
The search path was basically like appliances at large. Yeah. Being a vector to deploy batteries. The key realization we made was if you do one that's installed, like it's like. Typically runs on 240 volts. Yeah, you can make the battery bi-directional with the grid and you can't do that with like a normal one 20 volt plug.
You have to wire it in two 40, like a stove or oven or water heater or something like that. And so because of that, we real, and also it's bigger so you have a bigger battery that can do more service for the grid. And so that was like the search space that kind of converged on. Why we're doing a stove first.
Um, but taking a step back, like why I went down the path of like, Hey, I wanna do home appliances and all this other stuff was pretty dead set on using the skills I developed in, in the hardware engineering space to do something in hardware engineering. I went and kind of looked at like, I. What was everyone else doing and what trees were already being, you know, barked up, if that makes any sense.
Everyone's like, Hey, I wanna do a venture space company, or something of that nature. Yeah. The issue is like, existing companies cover a huge amount of the potential. Let's call it opportunity space. If that, if that makes any sense. Like you're seeing right now incredible difficulties in the like small launch vehicle market now that star shop's coming online, like it's going to be incredibly challenging and you're needing gonna need a muster, billions of dollars to be able to actually enter the space launch business, for instance, in the next.
You know, if you wanted to enter in the next five years. And also like blue origin's also, right. Gonna also be shipping as well. So, you know, these are things you can't enter as a seed stage startup anymore. Likewise, I think the defense industry, there's been a, a Cambrian explosion of new defense startups, especially centered on LA or the gundo as they say, or whatever.
Um, it's going to be something where Unreal has actually been able to successfully deploy this portfolio strategy. That means that in some sense they are the late stage bet. And we'll make anyone who doesn't have like true secret sauce and truly has a new angle to this, um, it's gonna make it challenging for them to enter.
And then consumer electronics is another one that I think is, is really a big deal where it's like there are a number of very innovative companies trying to do the laptop of the future, like framework, for instance. Nothing has done a fantastic job selling smartphones. People are doing straight shots to kind of existing categories with a new spin.
As someone who's seen that space from the consumer electronic side with with VR headsets, it is something that you have to be mono maniacally better than everyone else to be able to pry open those markets.
Noor: Why didn't put in ge, which is like a massive ass company. Why couldn't they hire like one good hardware engineer from Apple?
Or like, you know what I mean? Like they sell so many appliances right across the way, different categories. Like, I don't know, do you have any, any insight into why?
Sam: I've got a big story on this.
Noor: Okay,
Sam: so there's, there's a big, big, big story on all this. So there's a couple things. One is, um, the appliance industry today, and actually this is something about like all kind of like, not legacy, I don't wanna describe legacy industries, but basically like all kind of like.
Product companies of the past era, like this is like. The big three Automo automakers, this is like white goods companies. They have actually over many subsequent efforts at greater corporate efficiency, have essentially pulled out a lot of the key disciplines from the core business. Like they don't have like electrical engineers able to do new power electronics designs for their systems.
They actually go to a sub supplier. That has solved these problems and has gotten kind of all of the safety and compliance certifications all sorted, they go and obtain that technology from them. Likewise, the same thing happening with major automakers where like Ford's infotainment systems are not delivered by Ford.
They're delivered by a subsupplier. And so when Tesla shows up with a 15 inch screen in the model three. Ford has trouble actually matching that user experience, that integrated user experience, because they have not vertically integrated any of this into their like full stack engineering team because the abstraction layers of like developing a product are actually closer to like.
If you've built like a desktop PC where you go buy the graphics card from some Nvidia distributor, you buy the motherboard from like a Taiwanese electronics manufacturer. You buy the case from someone else and you like all put those parts together. That's like what the appliance industry looks like today.
Versus
Noor: things like that would end up with a better result. You have all these specialists who are making like, hopefully like the lowest cost. They're like super optimizing, like whatever, each sub component. So why is it not the case that you have as someone who's trying to aggregate all of those components?
You have like infinite choice and like you get to swap it and like, I don't know. I guess could you explain to like the novice, like why does that result in shittier products?
Sam: It. It doesn't result in shitier products per se. It results in less differentiated products especially. The problem is when new technologies enter the scene, those end up having to change the entire device.
Like you have to kind of redefine the entire device around the fact that you've got a new set of fundamental trade-offs. And if you can only alter a couple bits around the edges where you have to wait for the entire supply chain to kind of catch up, you end up in a pro a situation where basically.
You're stuck and can't actually innovate. You can only like deliver one of the 10 things you potentially could do with this new technology. And so that's what we've noticed in the appliance space is like, hey, you go and add a battery to the product. We've now got an extra 15,000 watts available in stove, but the induction electronics manufacturer that is the main supplier for most of these brands.
They haven't updated all of their stuff to work with this additional power or at the right battery voltage that we're using or all of that sort of stuff. And likewise, we've also innovated in a couple other areas because we've been able to actually do those electronics designs in-house, including being able to hold a consistent temperature, kind of like an oven on a stove for the first time.
Which requires, in some sense, very deep level of like cross-functional integration for the system that you don't get if you can, if you're basically shopping off the shelf for the parts.
Noor: I see. Okay. So basically the reason why the products are less differentiated at the end it sounds like is because if you have one newer component that maybe you can buy off the shelf, the other components that you need that they're not integrated with each other because they're not updating to like the newest of X category.
But that surprising me. I thought that if like their, if their only jobs to get, integrate with all the things, I guess, I don't know. It's just surprising to me that they wouldn't be able to do that.
Sam: Well, everyone's doing development based in demand. So like, ultimately I can describe this in more of a product facing sense, where it's like, Hey, I'm the guy specking out the next gen stove.
I then have to convince all of these sub suppliers for different corporations that are maybe not the most profitable businesses on the planet. They, they, they have, they have a lot of revenue, but they're maybe not super profitable to be like, Hey, you need now need to spend r and d. For a product that you'll maybe roll out in five years?
Yeah, yeah, yeah. To do that sort of thing.
Noor: It's, it's like the innovator's dilemma where like they have a big book of business for like the legacy stuff. So like, do they really wanna, they'll wait basically to, to do. Yes.
Sam: Yes. It, and also the volumes that they're seeing in some of these space, like, you know, 55 million appliances are installed every year in the United States, like it's a big market.
There is a lot of inertia to that as is, right? And uh, if you're starting to kind of flip the table and be like, Hey, the cost structure's gonna change because you're adding a battery to this sort of thing, but the capability matrix is, is also gonna change. It's not something that will cause the supply chain to chase the new technology in the same way that like this is happening in like the AI space where like, okay, everyone assumes that the LLM market is gonna demand this much inference compute at this date.
And thus, essentially people are able to raise venture capital. To essentially fund, like data center buildouts or new chip designs, et cetera.
Noor: Yeah, yeah, yeah. Cool. Well, awesome. Do you wanna just like get into the guts of like, like how you made this stove so awesome.
Sam: So basically we, um, we figured out how to basically triple stove performance.
So, and when I say performance, it's like the amount of power it can put into a pot. So gas stove heats by basically forming a flame in the flame heating the pot, so you're actually intermediately lighting a fire, and the fire conducts to the air, which conducts to the pot. An induction stove is a different technology.
Effectively what it does is there's a electromagnet underneath glass that you don't usually see. That forms a switching magnetic field, which essentially induces a current in the pan, which causes it to heat. The pan itself serves as the heating element of the system, so you're actually heating the pan directly through electromagnetic field effectively, so, so a typical gas stove is like the equivalent of 2,500 watts of induction?
Yeah, a normal induction stove is like 3000. We're able to put 10,000 watts into the pan. Which means that if you have a liter of cold water on a gas stove, it ends up being like six or seven minutes to heat up to, to boiling. Ours is under 40 seconds. So we, we, we developed that within like six months of the company existing.
We basically built our own custom power electronics. We built our own battery pack technology, we built all this stuff. We then quickly realized. It's kind of a bad user experience if you have more power than you should on a stove, um, and don't have any way to control it. Like with great power comes great responsibility as Uncle Ben told us.
Tell us Spiderman. Yeah. And so we realized we had to come up with a, a way to keep that in control and we ended up developing a new type of temperature sensing technology for stove. So instead of like zero through 10 or low, medium, high. You can say 300 degrees and it will hold 300 degrees in the pan.
Noor: Wow.
Sam: Which is not something that exists in any like, full-size cooktop or range product today.
Noor: Um,
Sam: at this level. And once we developed that technology, which actually took much longer, that was something that was another six months to like, rather compelling performance. Um,
Noor: and what were some of the like, hardest parts of that to, to craft?
So
Sam: it's a real, so a big, uh, so if you look at the picture of the stove, our stove, there's actually a like. There's actually like a, a raised piece in the middle of each burner that touches the pan. Um, developing that in a way where it was not so that has to touch the pan. Um, we went and basically investigated every way to do it, whether it was like not using contact with infrared, purely like it's like a thermometer touching the pan.
We ended up settling on this active technology. Which was not used, has not been used in the culinary space at all. I don't even know if like our exact sensor has been used in any context really in this incarnation. Like I think it's a purely unique design, but what it lets us do is actually directly measure the inside of the pan's temperature, which is something that basically you can't do without something active effectively.
Mm-hmm. Um, and so that developing that intellectual property was actually. Really, really, really tricky. And then we had to make it manufacturable. So we actually forced it through the supply chain that we had developed, um, for manufacturing the stove and with a lot of pain and suffering. Got that out the other end in a way where it's like able to survive repeated abuse, like cycling temperature from like room temp to very high temp over and over and over and over again.
Um, and still working at good performance.
Noor: That's awesome. And so what's the feature that you love the most as a, um, as a chef? Like what you like, really like from like the cooking experience standpoint, and then what's the feature you're proud of more from like, the, uh, engineering standpoint of like, yo this's just crazy.
Sam: Yeah, yeah, yeah. And I think I didn't, I didn't, I I didn't go talk about like, okay, so we basically built this stove around the fact that we've got the battery and got very high performance and it's got very like, awesome control. Um. So I, the temperature sensing is clearly the best feature of the product and like it's something that essentially makes this different from an induction stove to something that's fundamentally new and better.
And like an example from what does that actually mean and what does that actually let you do as a consumer or as a home cook. And it's like, I can make the same egg every single time with like no sweat. I, you can set a pan to three 30 degrees. You can melt the butter. It will lightly brown the butter because that's what it happens at three 30 degrees and you crack an egg and you'll get the same fried egg like sunny side up every single time and it comes out identical.
And I could, if you had the same stove, I could tell you exactly what I did. It would come up the same for you as well.
Noor: Mm-hmm. Even
Sam: with like, you have a slightly different pan. Doesn't matter. Um, that feature is awesome. So there's two, there's, there's two things I, I love to do to demonstrate to people, and actually there I can do it in one, one meal, which is I can cook a steak super consistently.
Yeah. And I do it in this very weird way, which is you basically just get a pan with a lid and you put the steak in the pan with the lid. Yeah. And you set it to like 140 degrees or something cold. Like not very warm. It will conductively cook through the pan just like a SoFi machine would. Yeah. But without the water, without the plastics, all this other stuff.
Noor: No microplastics sick.
Sam: No microplastics. But you get the SoFi, it, it comes out identical to SoFi.
Noor: Yeah.
Sam: So you do that, you then let the steak chill for a second because you wanna make sure that the surface is a little cool before you fry it. And then I go get another pan. I set it to 400, let's say 420 degrees for meme value.
Um, I can actually use the, um. I can use like olive oil and it won't smoke at that temperature, and I can fry the steak in like a little bit of olive oil without it smoking up my kitchen or anything like that. Like I don't have to run a hood on high to get a well seared steak that looks pretty much perfect.
Um, like at home and don't turn the smoke alarm off. Doesn't like, you know, cause a lot of like air quality issues in my kitchen, et cetera. And, uh, that's an awesome experience. And likewise, like if I'm doing a stir fry or something like that, I can set a temperature that's like, Hey, it browns the veggies, but it cooks the meat full through.
But like, I don't have to worry about it, like, you know, the oil smoking or like me burning stuff or anything like that. Um, so like everything from like very low temperature, like, hey, I want to cook really precisely all the way through to like, let me fry stuff without making a mess is like very compelling.
Noor: So what about from the engineering perspective? What are you like, yo, it was crazy. We pulled this off.
Sam: So there's some, there's a bunch of stuff we did. So the, the level of vertical integration this thing is, is pretty crazy. I think there's like two things I can talk about here that are, that are impressive.
So one is we developed this new type of magnetic knob, which you can remove the knob clean underneath it without any sort of problem. But then the knob itself is like this very premium feeling, high quality knob. You can turn and click. So it lets us navigate the ui, which we have a much larger screen than a lot of other stoves in like a clear, coherent, awesome way.
So it's like key to the user experience and it's like, feels haptically really great and really premium. And that's been like exciting a lot of, uh, you know, like a lot of folks around the product basically. So that's, that's, that's one of the things
Noor: that's critical. That's like, I feel like the, the number one thing that I like I notice immediately on our auto product is like, how does it actually feel to turn it on and to use it?
Yeah. So, yeah. That's awesome. What, what, what is like the material for the knob?
Sam: The knob, the knobs, like it's metal plus like some plastics internally to make it, make it work, but it, it's spill resistant, designed to be very durable. We cook on these things all the time, so like we, we would know if they're, if they're wearing out and stuff like that.
Yeah. And obviously we can send you a new one if it breaks or you lose it. Um, yeah, so that's like one of the pieces. And then the other thing that I'm really proud about, and this is like very much in the weeds and very much not like a stove feature, is the fact that we figured out how to make. This thing a actual grid scale battery solution, despite the fact that it is a stove.
And so we have figured out how to interface a appliance bidirectionally with the grid in a way that essentially lets us stove by stove, build a giant distributed power plant out of stoves. And all that's modular, so we can put this in other appliances or you know, things like that afterwards. But that was like, that was a level of work that required everything from hardware to policy to line up and make that happen.
Um, and uh, I think it's something that I'm very proud of the team for being able to actually achieve.
Noor: Yeah. Um, so how much is the stove.
Sam: So it's, so we're targeting the premium end of the market, but like the premium end that you can still get at Best Buy, if that makes any sense. So it's 6,000 bucks MSRP?
Yeah. But then because we're eligible for the same tax credits as like home battery storage products, so we're. A power wall with a stove attached to it. Basically, if you wanna think about it from a tax perspective, um, it's uh, you get about 1800 bucks off as well off that. So it's like 4,200. That's
Noor: crazy.
That's a huge discount. That's amazing. Yeah,
Sam: it's much more aggressive than the automaker tax credits and stuff like that. Do
Noor: you think it actually, like from an engineering effort, it was worth it? Like did it take you, like how much cost does it add to the product versus how much does it take off?
Sam: To add the battery to the product, like it clearly makes the whole thing worth it because one, the big issue is like you're not gonna be able to compete with these legacy.
Like essentially the, the competition is effectively, hey, I'm gonna go buy a Miele or one of these other great brands products, and then maybe if I need to get an electrician in my house to like, help install the stove, you're kind of competing with like a Miele plus an electrician.
Noor: That much is like the cheapest stove you can buy.
Just so like, yeah.
Sam: So I mean, you can get like a induction cooktop for 600 bucks to a thousand bucks. It's not gonna be anything to write home about. The melay stuff is like three to five, three to six grand, depending on what you get. Um, for something kind of similar to what we're selling. Um, a big differentiating factor with our product, by the way, which I didn't even mention, is when you install this, it installs in whatever electricity supply you've got in your kitchen already.
So it'll run on 240 volts, or it'll run on one 20. So it, we don't actually need, you don't need to like rewire your house to be able to install our product, unlike. Any typical induction stove, it usually needs a 240 volt, plus 60 amp circuit to be able to run, um, run in your house, which is the same as like installing a Tesla charger in your garage.
So if you have a gas stove, it's often really painful to switch to induction because you'd otherwise have to like, find space on your electrical panel and run new wires to actually make that happen.
Noor: So what's like the, um. Like the distribution strategy that you think is like more useful to lean into? Is it like getting the celebrity chefs to show off the product so that like the, you know, consumers know about it?
Is it like talking to builders and being like, Hey, this is like, you know, the most energy efficient and like, you know, from an engineering perspective, the most epic stove to like install into new houses or like, I. Talk to us about like how the market works and like how do you break, break into it?
Sam: So it ends up being an everything bagel strategy, unfortunately.
Um, yeah. But, uh, but we're, we're very excited about how much traction we've gotten on just pure play, like social media traction. Like I've, I've gotten millions of views on videos I've posted on x. Yeah, I've gotten, um, like people have reposted these videos on Instagram and gotten millions of views or tens of millions of views.
Yeah. So we're very excited about how naturally viral the product is and how able we're what to, to kind of drive a D two C business. Um, and we're clearly like set up to do D two C as well. Yeah. Um, we have, additionally, you have to realize how appliances are sold today and meet kind of the industry where it is.
So it's like, it's again, the builders, the like. Architects. Utilities are also really interesting as well because they're interested in electrifying buildings. So there's a number of folks on the distribution pipeline you have to engage. Um, and retailers obviously, because people wanna see this stuff in person before they install it in their house.
Like there's definitely a, like, I need to touch this before I buy it.
Noor: How, how hard does it to get into a retailer?
Sam: I, I would say it's, it's, it's not, um, instant is probably the best way to put it. Um, and I. The exact, our exact strategy there are a little close to my chest, but like, it's something that there are best practices to follow.
Is,
Noor: is this similar
to like, you know, just kinda like at a high level, like, um, you know, with CPG products where you're like, okay, you pay for shelf space and then you get like a certain amount. Uh, they kind of watch it and they see, okay, how many of these products move or whatever. Or is it like a totally different model or just like what's like the high level how it works?
Sam: Yeah. It's, it's, it's a little different than CPG because of how unique our product is. If that, if that makes any sense. Yeah. The, the major retailer, some of them bought appliance distributors, so it's not like you, it some sense have to go through a different kind of window than like the, the CPG sort of window.
Yeah. But there are agencies that you basically work with that help you through the whole process and know every key person that. In that space. Um, there are also different distribution situations here where like you're making a physically large pro product, so you have to, like, the warehousing is not exactly the same as like how it works for Yeah.
For CPG, if that makes any sense. The other thing in the appliance space that's unique is there's actually a, I've talked to very senior executives in the space and like a lot of them are very jealous of Tesla, um, because they're in legacy distribution. Environments where essentially they distribute to someone who distributes to the dealers.
And so it's like, and each of those phases takes a cut and they're like, yeah, we're all desiring to get out of this sort of arrangement. Yeah. Um, and go direct and like, you should go direct because you're kind of like Tesla. And so you should do, like, I actually got that advice from someone, but he may go direct online and have your own showrooms, was kind of his feeling versus like, it would be like, look like, like for instance, like Bosch has a showroom.
And you could go to it, but you end up buying it through one of the distributors in your neighborhood. It's actually very interesting how this works. Oh yeah.
Noor: Um,
Sam: so it's like, even if you go to the Bosch showroom, you're not buying it from Bosch, you're buying it from like two steps away from Bos
Noor: anyways.
Yeah. It's just like you wanna see the thing. Yeah. That's super interesting. Um, yeah, you were mentioning just like how viral the product is, so like how do you feel like you guys made dos sexy? Right. You know, it's like not something that people would think that, you know, would go viral or.
Sam: I think because it's just so out there that's like, oh yeah, I have this stove that can do all this crazy stuff.
And just earnestly advocating for the product on social media has like, I think I've, I've seen some people annoyed at me about this, but it's like most people are like, I. I've actually had people like, get mad at me for some other take I've had on social media, but they're like, but the stove is really cool and I want one.
And so I think part of, I think part of this is like just being like
Noor: weird enough forward and like, yeah, and weird and a little
Sam: wacky, but like, I think like, like Lulu uh, from RA being like, you need to go direct. Like, I think it's like that. Like I, I've like internalized that lesson and being like, I have to just basically be the number one advocate for this thing.
And, you know, if someone asks me some, even like the, even a silly question, I'll just go and answer them. Um, and that, and it seems to drive engagement and people like it.
Noor: Yeah, yeah, yeah.
Sam: And then they also make fun of me for like searing a stake poorly, and then like three software updates later it comes out great.
Because our thing is really software defined. But, you know, and then I, then I come back and I'll be like, Hey, did I, you know. Look, look, look, I improved my, uh, my steak game. Um, yeah, you should appreciate me now. I'm saying, you know, and being willing to kind of make fun of yourself as well has been important, I think as well.
Noor: Yeah. What do you think is the worst, uh, cargo call startup advice where everyone's like, you should definitely do this thing, but it's actually a shitty advice.
Sam: I think the classic one is just like, you should do the thing that everyone's talking about right now, but I, I have a lot of stuff in hardware, um, where I think like every.
Bunch of years, people are like, Hey, there'd be amazing if we had like a hacker space for hardware that likes you. It would be so accelerating for hardware, businesses, et cetera. It's like what you actually want is you almost like wanna ignore the fact that you have a lab and figure out how to make a supply chain that actually makes the thing at scale and try to force as much of your r and d through that supply chain as possible.
Because it means that everything you do is kind of like playing for keeps versus like an experiment that you then have to productionize. Um, I, I think that the, related to that, actually there's, there's something very, very fun, which is so SpaceX and Tesla and like all of Elon's, like whole universe, there's definitely this meme of like, you should vertically integrate your business.
And I think that is a fantastic thing for Tesla and SpaceX and all those other folks to do. But for a new startup, it's actually this very weird situation where you don't have leverage. 'cause you're not gonna be able to rack, it's kind of like, it's kind of like giving someone running a typical bi a a regular startup like, Hey, you should hire a CFO as your first hire versus using one of these fractional CO services.
And so if you go in vertically integrate and like, Hey, I'm gonna get CNC machines, I'm gonna set up like, I'm gonna set up a production line for my hundreds of units in, in, in this warehouse. I'm gonna do all this stuff. And then most of the capital equipment you just bought, you're using like 5% of the time.
Or you bought all this prototyping lab space, but like, and you have all the equipment, but like suddenly everything you prototype there, you then have to like teach a manufacturer to make and then they make it on their equipment, not yours. And so there's actually this like counter narrative to vertical integration, which is almost like you should vertically integrate if you can use a hundred percent of the stuff you're vertically integrating.
Like, like if you can, if you can justify like. Like a hundred percent utilization of the stuff you're bringing in house. Um, and if you can't, you need to be really scrappy and figure out how to like, and figure out how to, how to get that fractionally somehow, whether through a vendor or, or through some other form of aggregation.
Noor: Yeah. What's something that you feel like you've learned about the business's been kind of hard won over the last year, where maybe you had, um, you really, really thought aggressively one way and now you're like, you totally flipped on it after a year of learning.
Sam: I think, I mean I think that, uh, a big thing that you run a hardware business, I would describe it as like hardware is hard in the fact that you have to actually do a number of things beyond make a killer product that are also really hard.
You have to make a insanely solid e-commerce. Experience. You have to understand how growth marketing works. You have to know how to do strategic finance for essentially a business that's like quasi multinational. You have to know like all of this government regulatory stuff and all of the local codes and compliance stuff if you're doing any electrification space.
So it's like there's, there's all this regulatory that's just huge.
Noor: It's so funny, ironic, I think with Stove, they definitely don't think of regulatory compliance, but you, but I'm putting a
Sam: battery in, I'm putting a battery in your house and so,
Noor: yeah, yeah,
yeah. Fair.
Sam: But, and that actually is a big story, is like to get a stove with new, unique, new technology in built, you have to pass all of these like UL standards.
And so, and UL is Underwriters Laboratory. It's a private company, but they have safety standards for consumer products and like, you know, things that touch the electrical grid and various other things like that. And because, um, they're not like a legal entity, but mm-hmm. What's funny is all of the local building codes are like if you install a stove in your house, it needs to comply with UL 8 5 8, for instance.
Noor: Mm-hmm.
Sam: And if it doesn't, you're actually out of compliance with your home insurance.
Noor: Mm-hmm.
Sam: And there's no like stove cop that will come and like, bust you for this or whatever. But like it turns out that like there is this diffuse, essentially quasi enforced non-governmental, uh, standards situation that you do actually have to abide by and make and make happen.
And CPS e also cares about this too. So there is actually an enforcement as well. So you end up, you end up in this situation where, yes, I wanna build this thing, but you actually have to get it through. A number of different permission slip processes, uh, to be able to actually scale it.
Noor: Do you have any, um, like inside baseball knowledge or just like a weird scam that, uh, you know, these companies, like any sort of appliance company tries to pull on consumers that you now know about that, uh, you didn't before building the business?
Sam: I wouldn't call it a scam. Um, but remarkably, many of the western appliance manufacturers use the same electronics internally.
Noor: Mm-hmm.
Sam: You could be like, Hey, I want to get this one brand versus other brand. They all may have this one company's, uh, like the induction experience will be the same between two of these different brands.
And yeah, maybe the finishes will be different. The industrials entirely different, but like under the hood it's like, it's kind of like intel inside basically. If that, if that makes any sense. Yeah, yeah, yeah.
Noor: No, that's, that's, and it's kind of similar to like,
Sam: Hey, I'm gonna go buy a Dell or an HP desktop in like 2005 or something like that.
So that's not a scam, but that's actually downstream of that compliance and regulatory thing I said, which is like, because that's a heavy lift. A lot of them end up going to the same vendors, to, to, to, to power their systems. They're
Noor: like, check the boxes. Yeah. Yeah. Yeah. That makes sense. Yeah. Do you think that there's been a vibe shift since, uh, Trump got elected and like, do you think he's gonna have.
Or his administration is gonna have any impact on, uh, startup land or hardware or anything that you do, or it's just gonna be kinda like a, a whole non burger. Um,
Sam: I don't think, I don't, I wouldn't describe Trump as a nothing burger under any circumstances. Um, um, what is it? Uh, what's the, is was George V.
Bush the miss underestimate guy, but basically like, do not miss under Yeah. You know it to, to pull out an old presidential meme. Do you not miss, underestimate him?
Noor: I feel like the first time you got elected there was like so much catastrophizing that was like, the sky is falling, like nothing will ever be the same.
Like it's end of democracy or whatever. Like, I don't know, I'm just, I'm just kind of saying like, where, where do you stand in terms of like how much of an effect it's actually gonna have over the next four years or even Oh, I mean, I,
Sam: I, I, I think there's, there's a bunch of situations around like, how will incoming Republican administration regulate X, Y, ZI, I suspect the tax policy stuff and like how the tax cut and Jobs act extension gonna happen is gonna be.
A potential giant, giant story, especially with like a fairly narrowly divided house of representatives. Um, and how that intersects with the tariff situation. So if you run a startup, the problem basically is like you gotta manufacture stuff in, like often have to manufacture stuff in Asia, period, just because that's where like integrated supply chains that will work with are willing to work with you exist.
And so there's going, there's this big push toward domestic manufacturing, but that's gonna run into this other buzz saw, which is like, Hey, we could raise tariffs on China to 60%. So everyone will just move to Vietnam and then it's, we could raise, we could have a global tariff to like offset potential tax cuts or whatever.
Um, but then there's also this talk of this mass deportation attempt. So like there could be this weird triangulation where like the sweet spot for manufacturing still ends up being Southeast Asia because, um, um, because oh, like 20% tariffs are pretty similar to what we're currently paying on China today, or 10% tariffs, whatever.
There's not a labor force to do these, the new manufacturing and whatever is there. It gets like, you know, bid up, you know, to the point where it's not gonna be necessarily affordable for new businesses, um, or non heavily automated businesses, if that makes any sense. And so I think that tension's gonna be very interesting.
The second tension is, did you know that every Apple product is mysteriously exempt from the Trump tariffs?
Noor: Wait, how? That's crazy.
Sam: Tim Cook showed up at Tim Cook, showed up at the White House, and then they were mysteriously exempt from the Trump tariffs.
Noor: That's Epic dude. How did he pull that off?
Sam: I think it was partially the like agreeing to make the trash can Mac in Texas or something, which they did at Low, which they did at low volumes.
But essentially there would
Noor: company can get like a carve out like that. That's insane. That's why you gotta be too big like that. You gotta be too big. And the
Sam: threat, the threat was, oh, we will just include the Trump tariffs on your checkout at the Apple store is my understanding. Wow. And then, because that looks really bad from a, like, hey, you just caused inflation in apple prices or whatever.
Um, but my understanding was there's basically, oh, that's what Tim
Noor: Cook threatened a Trump. I, that's my,
Sam: that, that's my guess. I don't actually know, know exactly what happened, but. It essentially Apple got a blast. That's so funny because just,
Noor: just like when you buy airline tickets where it's like, oh, this is like your TSA line or whatever.
Right. And then Oh, I see. Interesting.
Sam: And so the other thing that happened was a lot of people moved stuff to Vietnam and
Noor: mm-hmm.
Sam: Weirdly enough, this didn't actually change supply chains that much. Um, a lot of Chinese manufacturers just opened Vietnam branches.
Noor: Wait, wait, wait. Any other intel on Tim, this Tim Cook deal or.
Sam: I don't know. I don't know how, I mean, I, I just would say hats off to Apple's public policy team, but, uh,
Noor: yeah,
Sam: but, but basically my mind,
Noor: but it's been publicly reported. It's not even a rumor. Like they're just, it's known. No, no. It's just,
Sam: if you go look at the HTS schedules, it like no Apple products are, you know, or very few Apple, like laptops are exempt, phones are exempt.
You know, that's the ball iPad and no, and
Noor: no other vendor has the same thing. It's just that No, no, no. It's,
Sam: it's, it's blanket for the categories of products, not blanket for Apple specifically.
Noor: Oh, okay. So all laptops, all
Sam: laptops, all phones are exempt under that sort of situation.
Noor: I see, I see. Got it. Okay, cool.
But, but
Sam: again, given you, like given where the, where, who produces phones and laptops in China, you know?
Noor: Yeah, yeah, yeah. Sure, sure. So, so
Sam: that was, that was an interesting thing. But then that makes me like the big problem with kind of tariffs as a primary revenue stream or large tariffs is. Essentially it then means that there is a negotiation you can have with the feds and, and like, like I could basically say like, Hey, do you want to, you know, you know, our specific category of product could be made exempt under certain tariff policies, you know, if I go and lobby in a certain way.
Right. And so I think that there's, the arbitrariness of it is actually, I, I would describe that as like. A concern I have because people with more access versus small startup guys like me may be able to get, you know, systemic advantages that disadvantage younger hardware companies, um, versus say incumbents.
Noor: Yeah, yeah, yeah. What about ai? Are you a dor? Are you a eac? Like, do you think AI's gonna have any impact on stoves or the hardware industry at large? Or what? What's your take?
Sam: So I'll give you my big AI take. I am incredibly skeptical of the macro idea that I've heard many times from many people. I'm gonna build a God and if you're in my specific God's embryonic phase with me and you know, and our friends.
That, uh, we will be the high priest class for, you know, the, the, the cyborg apocalypse or whatever. And it's no one else will be, isn't it? Yeah. Well, I mean, I would describe it as like we had a Protestant reformation and now there's just a bunch of different subsect leaders effectively this point, and we're in a current bake off between them, if that makes, if that, yeah.
And I find it very entertaining. I also think AI is totally real and can be very transformative, if that makes any sense. Yeah. But I'm also like, I don't. I'm not sure if I described this in a religious sense. I, I, uh, I say this as like a, uh. There's this sort of like, I believe that AI will be fundamentally transformative and awesome and enable a lot of very big things, but like in that sort of religious debate, I'm like an atheist or something.
Like a militant atheist.
Noor: Yeah. Yeah. That's funny. So what do you, what's the stuff that you think is sort of, uh, like real, I'm gonna have an impact over the next like year or two, and then just like society, like, I don't know. Is there any product that you would like. To buy, you know, for, for impulse to help do your work faster.
Is there any like request for startup where like, dude, somebody need to build this thing. I don't know why everyone's like doing whatever, you know?
Sam: Yeah. So I, I, I think that the, the, the thing with, the thing with a lot of AI stuff is, and this is my second like skeptic take, not like, and I'm not like an AI dor, I'm not an eac, I'm not like getting this where people, but my skeptic take is like, is like.
It actually takes time to impact the built world. Like a fun example of like why our product exists is like, we can't upgrade the wires that are like getting electricity from the power plants to people's houses. Like, do you think that we're gonna have AI doom if like, we literally can't, like re like upgrade the copper wires in the streets?
Like it's sort of, it's sort of this crazy thing to think about of like, we've essentially made editing the built world illegal through various, you know, regulatory, you know, policies like that. That's maybe like one like silver lining Trump administration is there's going to be an attempt to shake up the regulatory, the regulatory state and potentially NEPA and stuff like Elon's beef with the California Coastal Commission.
Like all of those sort of things are directionally, I think correct for allowing us to decarbonize. So, but my point though is like the whole like AI going sentient and then impact in the built world at, at big scale, it's like we're still at the early innings even for like these humanoid robots, despite there being tons of companies like the actuators and like.
Thermal situation on the exploration stuff, like that's all gonna have to improve possibly under supervision of like new AI based motion algorithms, for instance. So we're like a couple cycles away. I would describe from like AI starting to impact real-time hardware outside of like the self-driving car space and stuff like that.
Um, the most near term stuff I'm really excited about is like essentially all of these coding assistants are relevant for hardware companies like. You can make firmware way, way faster. You can, um, you can potentially test and simulate real systems. You can, there's a number of companies working on printed circuit board, um, authoring tools where historically that was always a, I would frankly say like a dumpster fire of like any sort of like automated tools to help me design circuit boards.
Um, but it's actually gonna work now. And so I, I think that generally there's going to be a big step up in kind of tooling. Over the next two years or so, just like, just like it happened with software engineering over the past two years. I think that we're probably not immediately at the like single person hardware company era.
Like we may be close to with like software companies and app creation and stuff like that. Like I think we're fairly close to that right now. Um, but like from a product standpoint, it's like. Literally, the products I would advocate my team get are the same that I think you would advocate for a software team.
It's like, Hey, like everyone should be using Cursor or GoodHub Copilot or, or what have you. Everyone should be trying some of these new like coding agent tools. I, I think that all of that's kind of the first, the first thing, the other product we've found is actually just like PDF upload into chat GPT or cla and just like understanding these complex regulatory and compliance documents, that's been super useful as well.
So it's like. But it's, it's not anything that I would describe as like outside the box from like, what you might use, if that makes any sense.
Noor: Yeah. Uh, do you have any, uh, predictions for humanoid robots? Like, are you excited about them? Is it overhyped under hyped? Do you think you're gonna buy one? If so, when, what are they gonna be able to do?
What are they gonna fail at for a while?
Sam: So, I had a tweet about this yesterday. This is actually maybe me spilling some alpha, and I'm gonna dig this up right now so I can, I'll read it off for you. But basically what I have heard from talking to folks is there is, um, so we're getting these awesome demos and like, I, I actually think that like, I.
The robots are gonna go talk to you, sound like a human, just like, you know, your, you know, the AI apps, your phone are go talking, sound like a human. All that stuff's gonna happen. Um, the motion planning stuff, it's gonna work. They can spatially, they can understand the room they're in. They can be like, Hey, pick the box up from the wall and, and or from the rack on the wall and put it over there.
Like that sort of stuff is going to start happening at a very, you know, high fidelity level. But what I've heard is the robots, as they exist today, there's actually like a. Actuator problem, where effectively every humanoid robot is like sitting stiff or standing stiff.
Noor: Mm-hmm.
Sam: Um, and like, they're not like loose like a human is where like you kind of lock your knees, lock your hips, and you're kind of like, you know, you're, you're, you're only giving small adjustments to stay steady.
No, it's like everything is kind of just like super stiff and like holding position with a lot of power being burned. And this has a thermal and longevity for the actuator's problem that. Needs to be addressed both through the hardware, but also through the control systems and the, like ML that's backing those up.
Because a lot of the, like the, like this isn't like a hard program gate or like hand motion or anything like that. Like this is like machine learning driven. Um, um, motion control now. There's going to need to be basically like a couple more generation generational steps. Like when I say that, it's not like this is five years away, 10 years away.
It's like, no, this is like iteration of the hardware with software, with the hardware again, which could be quicker than that. Um, for sure. But it's gonna require a skillset that's not just like, Hey, you need to be really good at ai. You need to be really good at like. Developing a humanoid robot body, you're gonna need to be good at like motors and like power electronics.
Yeah. And like at a very deep level. And have those people talk to your AI engineers. Yeah. And so that's my, like, that's gonna be the thing to watch. And that's gonna be who, like, who's gonna really shine here is gonna be really good in those spaces. But my point being is I don't think we're gonna get, like, I think that you're gonna be able to get like a humanoid, like you can get those like robot dogs, um, and, and play around with it.
I don't know if they're gonna be personally useful for me in the next two to three years. I suspect, and I'm most bullish on like Tesla because they've got the factory to actually like, put them to work. Um, like they can be their own, they can be their own customer. On, on, on day zero even, and be patient with the thing not being great in at least initially.
Yeah. Um, so like that, that would be the reason of like, I'm, I'm probably most confident in Tesla's, Tesla's situation here, but then outside of Tesla, most people doing like novel actuator approaches. There's one company clone that's doing, I'm not, it looks like a Westworld hand or something like that.
It's pretty cool. Yeah. Um, but, uh, so, so that's the first thing. Second thing is
Noor: what do you think about applied intuition and like, I feel like they've been kind of on the scene. Um, oh, I mean, I
Sam: think, I think if you're doing shovels for this stuff, it's like genius. Um, yeah, just like, like I, I think anything like that is gonna, because it's like, oh, everyone's gonna be training like bottles with, or like everyone's be developing slam like techniques and like these sort of spatial understanding models.
You're gonna need platforms that know how to like help, can help you build those. Um, the, uh. Other thing is because there's so many people doing these like robot companies, there's so many people doing these like defense drone companies. Yeah. And then there's like, basically what's gonna happen though is there's gonna be this like explosion in motors and power electronics development.
Mm-hmm. And the performance of all of that's gonna get stepped up a lot of the next couple years. So there's actually gonna be another class of companies that will basically be developed off of the fact that the humanoid. Companies and the drone companies have made motors and actuators like cheap as all get out.
Whoa.
Noor: And so I'm really
Sam: excited about that because it's gonna be like some 22-year-old, like a, like the next Palmer Lucky type person is gonna be like, Hey, I just like found that you can make a VR headset from like modern phone parts and it doesn't suck. Someone's gonna figure that out with like the stuff that's getting developed now, basically.
Noor: Wow. That's awesome. Well, thanks much for the time, Sam. I mean, you've built something, uh, really new and really special and, um, it, it, it, it's not only a cool engineering feat, but it also gives off, uh, really tasty items for, for everyone to eat. So thanks so much for, for building that. It was a really fun, uh, conversation.
Sam: Thanks for thanks for hosting and, and, uh, I'll, I'll probably see you in person soon.