SPEAKER_00: I think what you're going to see from light matter over the next year and a half is extreme volume on groundbreaking optical products. And, you know, what I'm really aiming to do with light matter ultimately is to build the photonics company. So if you think of the word photonics, I want you to think of light matter ultimately. And the way that's going to happen is through the whole set of technologies that what we've developed can kind of be realized.

SPEAKER_01: This Week in Startups is brought to you by Carta now lets you launch and administer SPVs for your syndicate. Share your knowledge, capital and network to launch your syndicate SPVs through Carta. Get 10% off your first SPV at carta.com slash twist with promo code twist. Eight Sleep. Good sleep is the ultimate game changer. Now you can add the pod cover to any mattress. Go to eightsleep.com slash twist to check out the pod cover and get $150 off a checkout. And LinkedIn Jobs. A business is only as strong as its people and every hire matters. Post your first job for free at linkedin.com slash unicorn.

SPEAKER_04: All right. Listen, everybody. AI is changing everything.

SPEAKER_05: A big part of that is what's been made possible by GPUs, TPUs. All of this is powered by electric chips, basically, right? That's a standard for decades.

SPEAKER_07: We took sand, we made it into chips, and we changed the world. But electric signals, we all know that produces a bunch of heat, requires cooling systems. But there's something called photonic computing. And our guest today is at the forefront of that. And his name is Nick Harris, and the company's name is Light Matter. Nick, welcome to the program.

SPEAKER_10: Hey, good to meet you, Jason, and thanks for having us on the show.

SPEAKER_07: Yeah. So tell me, what is photonic computing? Because I am a neophyte in this.

SPEAKER_05: I'm aware of it, but I don't really know the state of it and if it's actually competitive and when this will actually be reaching people's desktops or servers.

SPEAKER_12: Yeah. So maybe I'll start with a little bit of a story to give context of it.

SPEAKER_00: So for the past 60 plus years, computers have done all calculations and communications generally using electrical signals. There were a bunch of incredible physics that were enabling those transistors, the fundamental component doing a lot of this computation and communication that we all use today, enabling it to get better, faster, more energy efficient, smaller, cheaper, all these good things. And that was kind of described by this idea of Denard scaling, which is that energy usage per device would shrink with time and also Moore's law, which was that we'd get more transistors per unit area as time progressed. So those were the two things driving all that progress, uh, in around 2005, that trend broke and a lot of people in the physics community, uh, you know, at universities, but also at the big companies doing the work, developing these fab processes, realize that there was a fundamental challenge. Um, I was working at Micron at the time when that wall was hit and you could feel it. And I love computers. I don't know why I've just always been obsessed with them. And, uh, for me, it was, wow, this thing that I love is kind of coming to an end or at this wall, how can we make it better? There are a lot of different ways that people have looked at trying to extend the computing roadmap. Um, the way that I've looked at and the way that, that others have as well is using light. Now there's two ways that you could use it. One would be to do the computation and that's the oddball really new thing. Um, the other is to do communication where you're sending data between points. Um, you know, a lot of the undersea cabling in the ocean, all of it actually is optical, uh, in data centers today. If you go into the data center, there's optical cables everywhere. They're using light to communicate signals. The, the new thing that's happening in the communication space, which we're working on at light matter is that people are sending, uh, data between chips that are actually within an inch of each other. We're using light and between racks and all of these scales, but what's fundamentally happened is that these optical engines are getting closer and closer to the processors that we use.

SPEAKER_18: Um, we're finding that there's a benefit actually to bringing them all the way into the chip itself.

SPEAKER_20: Wow. So if you were going to put a bunch of servers across the ocean from each other or across

SPEAKER_05: the data center, of course you'd use fiber optics. You can move a ton of, and you, and you see that, right? Some computers have ethernet plugs on the back. Other ones have literal, you literally, you can plug into the back of, uh, a server and optical cable where yes, all the data coming out comes out by light. And there are other constraints like hard drives, which of course have become SSD.

SPEAKER_23: So those are faster, but what you're saying here is the next step is, Hey, between the chips, we're sending the data over optical computing.

SPEAKER_16: That's right.

SPEAKER_00: So the distance that people are using optics over is getting smaller and smaller and smaller, and it's starting to make its way inside the processor. So we've actually developed two kinds of technology at light matter. One is actually doing the computation behind deep learning using tiny little optical components that measure microns by microns, and it does the additions and the multiplications that describe neural network, mathematics, deep learning, think about large language models. It does that math using light. That's a huge, you know, breakthrough, new field, um, you know, something that we've pioneered here. The other thing we're doing is we realized that these optical compute engines we built are very fast. We need to find a way to keep them busy. One of the big pieces of feedback that I've gotten from, uh, diligence groups, uh, famous people in the field when they've looked at us as like, ah, it's cool. You can build that. How are you going to keep it busy? Um, so we took that to heart and we invented our product line passage. And what it does is it allows you to feed the beast. It allows you to send bits at data rates high enough, both between the cores and memory, but also to other servers so that you can actually keep these things busy and not just sitting there waiting for work.

SPEAKER_05: Ah, and that is one of the keys right now. As part of this, uh, massive compute intensive language model building is that we're actually

SPEAKER_07: under utilizing a lot of the hardware, correct?

SPEAKER_00: Oh yeah. It's embarrassingly underutilized. I've heard numbers as low as 5%. So you'll have a huge number of theoretical maximum operations per second, and only 5% of them are being used. And the reason is that if you look at a profile of these jobs, let's say a chat GPT type workload, most of it is spent shipping data between memory and the processor core or communicating between cores in the same box or distant boxes. It's not doing computation.

SPEAKER_07: Yeah. So just moving the data to the point at which the computation can occur is the bottleneck. Uh, and, and you're trying to solve for that. So have, have your products hit the streets yet as it were, or is it still theoretical?

David Friedberg: Where are you at in terms of these things, um, being in the field? Yeah.

SPEAKER_00: So a little context on the company. We have about 130 employees, uh, offices in Boston and mountain view. We've raised about $300 million from, uh, investors, including Sequoia, spark capital, fidelity matrix, Google ventures. So we've really got a bench of all the top VCs in the world. Um, we have built many generations of chips, about, uh, three or four at this point. Uh, we have six customers on our compute product line, and we're going to be launching our interconnect product line at high volume next year. Got it. So pretty far along.

SPEAKER_07: And who are the customers for your products? Is it, is it the open AIs of the world, the, the, the cloud computing platforms, or is it,

SPEAKER_20: you know, deeper in the stack people who are building the servers and who are building the platforms?

SPEAKER_00: It's cloud infra. Got it. So we're building like heavy duty hardware for running these huge AI models. Um, so it's these big clouds.

SPEAKER_04: And so are you then.

SPEAKER_05: Being put up against say Nvidia in these cloud offerings and, and, and what are you, how do you compete against them today?

SPEAKER_00: So on the compute side? Yes, we would be compared against Nvidia on the interconnect side. Not really. Um, so on the interconnect side, we actually enable all companies. So you can think about companies like AMD, Intel, Nvidia, and even the internal product teams for TPU, Amazon builds their own chips and so on. They can all use our interconnect technology to help extend what they've got as far as it could possibly go. And those are generally chips that would be deployed in a data center. The way that we work with them is they have a chip, an accelerator or a network chip or a FPGA, and they want to scale it out using our optical interconnect. We figure out basically what the floor plan of their chip is. Where are the pins? And then we build a passage, which is a very up to a full wafer format. Silicon photonics chip with wave guides that interlink everything on top. Think about it like a chess board with light connecting all the chess pieces where the chess pieces are chips. Wow. So, so we basically lay their chips out in arrays. They can be processors, memories, any type of chip and allow them to scale the solution. So in that way, we're kind of helping everybody.

SPEAKER_07: And what impact if this is fully deployed to the cloud, what impact is this going to have on what we've already seen in terms of language models being built and, and, you know, the, the number of, uh, or the amount of data being processed by them?

David Friedberg: Yeah.

SPEAKER_00: Um, there's an interesting story here. So we're kind of bound by power usage and cooling at this point, not light matter, but the world, um, in terms of data center deployments. When I started the company chips were about 300 Watts. If you looked at a GPU, maybe 250, 300 Watts today, they're over a kilowatt. The next gen chips are about one and a half kilowatts. So things are accelerating really fast. You could even say they're heating up.

David Friedberg: Things are heating up. Yeah. Yeah. I mean, that is, that is the key, right? Is that this creates a bit of a heat issue.

SPEAKER_00: No, absolutely. And if you look at that heat issue, part of it is that the interconnect bandwidths, how many bits per second are leaving these chips is growing very, very fast. It used to be maybe a hundred gigabit per second, and the latest chips will push out seven terabits per second. So we're seeing a big acceleration in the number of bits that are shipped. The problem is when you ship those bits over electrical wires, it ends up being the case that a server, if you do a power profile, more than half of all the power used is just moving the bits between the chips, not even on the chip, just between the chips. And so if you look at what we're doing with Passage, it's about 10 times better energy efficiency. So 18 picajoules per bit for the typical type of inside the box communication that you would see with an Nvidia server or an AMD server, take that down to two picajoules per bit. And I know this sounds quite technical picajoules per bit, but the point is 18 divided by two. It's a big difference. What is a picajoules? 10 to the minus 12 joules. Joules is a unit of energy.

SPEAKER_05: Right. So you can massively reduce the amount of energy, which then gets rid of the heat and the cooling. And then you can put more compute in the rack.

SPEAKER_00: Right. And all that real estate that you build out is now more useful. If you didn't want to add more compute, the power bill went down. So that's nice. And by the way, the power bill is about 50% of the cost for running the data center.

SPEAKER_51: Listen, if you're in the tech industry, you know about Carta. Carta is the leading venture capital and equity management platform.

Chamath Palihapitiya: And they have huge news to share here on This Week in Startups. Carta now lets you syndicate an SPV. You know what an SPV is, a special purpose vehicle. So you create an SPV on Carta. Why would you do that? Hey, listen, you're an angel investor and you're putting 25K in a company like I did with Calm.com, but you got about 20 friends who also want to put in five or 10K. Now you put them all into an SPV. You tell the team over at Calm.com or whatever company you're investing in, it's going to be one line item. I'll sign for all 25 of those angels. And they say, oh, great. Can I put 10 other angels in your SPV? And then, hey, if you want to take carry on it because you syndicated the deal, great. Now you got a business model going, huh? They are used by more than 4,500 funds representing over $120 billion in assets under administration. They're going to support you at every stage of your fundraising journey, from doing your first syndicate to building a global venture capital firm. You can raise and deploy from anywhere in the world because Carta offers US and international SPVs. Also, Carta provides an automated back-off solution for you so you can focus on what matters, finding great startups, building relationships, and supporting the heck out of those founders. Here's your call to action.

SPEAKER_56: Go to Carta.com slash twist and use the code twist to get 10% off your first SPV. What a deal. Carta, C-A-R-T-A dot com slash T-W-I-S-T. Make sure you use the promo code twist for 10% off.

SPEAKER_58: Do you think we can catch up to the need?

SPEAKER_07: Because part of the issue here is we went from, let's say, millions of enterprises having no AI strategy and saying,

SPEAKER_23: yeah, you know, we're going to get to that at some point, to in one year, half of those, you know, million saying, this is the most important thing in the building. And the other half are kind of realizing, and I'm sure by the end of the year, they're going to say, hey, this is the most important thing we need to catch up on.

SPEAKER_05: So we're in catch up mode right now, clearly. Do you think we catch up? Or do we, do we think that all of this, um, energy by corporations, not actual energy,

SPEAKER_07: but all of this intensity that they're putting towards this is, um, we're going to be able to, between Nvidia Intel yourselves and everybody else working on this, we're going to be able to catch up or do you think that this is just going to spiral where once organizations get a taste of what AI can do, they're, they're just going to be doubling and tripling down.

SPEAKER_24: So are we going to catch up and what's the window of catch up? Is it a year or is it 10?

SPEAKER_00: It's going to grow like crazy. We're, we're not going to catch up. You know, I think it's going to be a case of, um, trying to build out enough ships, you know, companies like TSMC adding more capacity, Intel catching up on process leadership. And I think they will, um, it's going to take, yeah, Intel is going to catch up. Yeah. I think they'll do quite well. I actually think there's an end game for a process technology for semiconductors. There is a final node. I'll put it that way. And someone will reach it first, but then everyone will get there. That's, that's sort of my view on things. Having seen that the physics behind it and then what it takes. Um, yeah, so I think that the demand is insane. Uh, if you look at companies like Nvidia, they have more demand than they can possibly build out, uh, which gives them great pricing power and these sorts of things. And the whole world doesn't like it when someone has great pricing power. And so there's an opportunity for other companies to come in and bring a solution.

SPEAKER_20: Um, that does seem to be the truism here, which is it's very rare that everybody wants the same thing in the world, but Taylor Swift and H 100 seem to be the two 12 tickets.

SPEAKER_23: And H 100 seem to be the two items that are that much in demand. And, um, you know, there's only so much Taylor Swift to go around, but

SPEAKER_05: Nvidia has competitors, right? And they're, they're, they're going to catch up. They're going to have competing products and that will bring these prices back down.

SPEAKER_00: Yeah. Yeah. I think that that that's possible that it will happen. There's another thing to consider, which is that you can use the AI tools to build your AI chips. One of the big powerful things behind Moore's law was that you use the chips that you just made to make the chips that you're about to make. And that's something that you're seeing Nvidia do. And hopefully the whole industry do. Google's done good work on this. Our VP of engineering led TPU at Google, and they had a paper in nature a couple of years ago where they were using an AI to actually design the floor plan of the chip and lay out the circuits. And you can get big advancements in this. So, um, it's going to be a big non-linear feedback loop. You're going to see these tools, tools being used to design themselves. Yeah.

SPEAKER_04: And at the speed, these things are going, my Lord, um, that could be quite impressive.

SPEAKER_00: Yeah. Yeah. I mean, I have to tell you every year, one of these results like chat GPT comes out and it's mind blowing. It makes you question like what the future is going to be like. And if that happens every year, I mean, it's hard to imagine where, where this ultimately goes, but I can tell you, we're going to need to ship a lot more bits per second. The AI models are going to need to get bigger. You're going to need more compute hardware, uh, you know, and you're going to see exotic things. You're going to see water cooling is standard for computer chips today. So we run water over the top of the chip to pull heat out. Uh, the next thing that people are doing and in China, this is actually quite common is immersion cooling where they'll take the chips and actually put them in a, in a liquid that boils. Uh, when liquids boil, their heat capacity increases. So we're able to pull out more heat, uh, at that phase transition. So it's, it's a very interesting technology. I don't know what happens after that. Maybe data centers in space.

SPEAKER_23: Well, uh, it's pretty cold up there and yeah, that would, and it wouldn't, it doesn't cost you any energy except to get up there and doesn't question energy to cool it up there.

David Friedberg: Um, but you need to radiate the heat because there's no air blowing.

SPEAKER_00: There are no molecules bouncing into the, uh, the chips to cool them down or the heat sink. So you have to build a special kind of radiator. I think that's an interesting physics problem. Uh, but it's something to disperse the heat off the chip because you can't blow

David Friedberg: air onto it. You need to shine the heat away. You need to shine photons away.

SPEAKER_60: It's fascinating to think about data centers in space, the pros and the cons, obviously all the energy to get it up there and to maintain it up there, but then you do gain a lot by having it

SPEAKER_16: out there, I guess, in terms of maybe you'd have to look at serviceability, like the lifetime

SPEAKER_00: of the chips. There are all sorts of alien things that happen to silicon. I hear they grow dendrites when they're in outer space, like silicon chips will grow these dendrites and ultimately the packages will fail.

SPEAKER_12: Um, it's, it's crazy. It's crazy. But I guess the thing to look at, uh, and this should scare you a little bit.

SPEAKER_00: Um, we're at sort of the heat death with computers, um, with Denard scaling being broken. Uh, Denard invented DRAM at IBM.

SPEAKER_05: And he also predicted that theory, if I'm correct is, Hey, no matter what we do to these chips in terms of, you know, Moore's law or doubling the capacity of it, whatever. It's just going to be the same amount of energy is being used. Is that like the basic.

David Friedberg: Well, the point is density goes up.

SPEAKER_00: If you happen to find a way to fit more transistors in the same unit area, the, the amount more that you added, the heat's going to go up by the same amount. And this, this is the real issue today. Computer chips have the same energy density as a nuclear reactor. Hopefully that like sets the right picture here. Yeah. So ultimately what happens, there's a death spiral to the heat problem. Silicon, the resistivity, the thermal resistivity, and we're going into physics land here. It's nonlinear with temperature, the hotter it gets, the resistance changes and you generate more heat and it goes into runaway and sort of lights itself on fire.

SPEAKER_12: Um, we're already close to this with chips coming out that are one and a half kilowatts. We're pretty close. There's a limit to just going to go up in flames and there's no way to cool it, I guess. So, so how do you fix that?

SPEAKER_00: Um, the way that you fix that is okay. Stop scaling the chips. Like don't change the performance per chip. Just build a huge supercomputer. It's going to be very expensive. All of those like maker movements and things that were created by electronics becoming cheap. They die. Now you build gigantic supercomputers. That will continue the gauntlet of performance. That's it. So you only have a few companies that can build them and own and maintain them. And the progress is fascinating.

SPEAKER_05: Cause we went, you're, you're referring to like Arduino chips and like, oh yeah, we're going to get chips down to a dollar, $2, and we'll just make a ton of them. But that doesn't work for these jobs that we want to do. These jobs require much more compute. Therefore. Yeah, we are going to have, we're going to live in a world of supercomputers again, and we're going to be time rent. We're going to be renting time on a supercomputer. We're literally back to the age of vaxes and mainframes. You got back to parole computers.

SPEAKER_07: Where you're going to be renting time on some giant supercomputer. And the, these computers are going to cost how much to, to build and maintain. Do you think in the next? Oh, there are billions of dollars.

SPEAKER_00: If you look at, uh, you know, anthropic and all these other, uh, AI startups that are getting funded like crazy to build LLMS, Claude, uh, you know, from them, uh, they're, they're raising billions and it's all going straight to NVIDIA, which is amazing for, for Jensen and NVIDIA, but, but it's also kind of scary. I mean, uh, if you think about that future. So that's why at light matter, we've looked at the compute problem too, because fundamentally you can't just let the chips go on this singularity spiral of heat death. You have to find new ways for computers to make progress. And so for us, it's both the compute piece and the interconnect. You have to innovate on both to keep it going.

SPEAKER_42: I also suppose, um, the software needs to be optimized massively.

SPEAKER_60: It seems like we are so far behind on the optimization of these LLMS. And then a lot of these jobs are just getting, they're just throwing hardware at the problem.

SPEAKER_00: Yeah, a hundred percent. There's a big, there's a big, uh, improvement to be had through software innovations, but just remember, uh, in 1960, you could have done software innovations.

SPEAKER_12: So you do need to do that, but the hardware also has to progress. Um, that's, that's the tricky thing.

Chamath Palihapitiya: If you want to get ahead in your career, then you gotta be ready to fight every day. You gotta grind it out day after day. This is obvious. You're listening to this week at startups, you're grinding right now, but if you're a workhorse, you need to get a good night's sleep, period, full stop. We all know that getting great sleep equals better performance the next day. Sleep is a secret weapon and your secret weapon to getting great sleep is eight sleep. You know, I love my eight sleep mattress. I use it every night. I got it in both houses because I want to be able to get control my temperature. And I want to make sure I optimize my sleep. And you can do this with their new product. The pod cover by eight sleep fits on any bed, just like a fitted sheet. It's going to keep you cool all night, all the way down to 55 degrees Fahrenheit, if you like, and it will improve your sleep by automatically adjusting the temperature on each side of the bed. And you get personalized sleep reports. If you're one of those quantified self people, you're going to love this feature. And when you get great sleep, you get great performance. And listen, everybody's been talking about it publicly on twitter.com. I'm sorry, x.com. But no coastal Paul Graham's up myself. Everybody's been talking about how great eight sleep is. Invest in the rest that you deserve by going to eight sleep.com slash twist. They're going to give you 150 bucks off the pod cover. Because you're friends with this podcast. That's E I G H T S L E E P.com slash twist for $150 off the pod cover and their shipping, not just in the US and Canada, but they got the UK and select countries in the EU. And Australia.

SPEAKER_07: Knowing what, you know, it's always fun to speculate about, you know, general AI and feels like one of the interesting tricks that chat GPT has taught us is just how simple our brains are. You know, we next word prediction.

David Friedberg: Yeah, it's kind of like, you know what I'm about to say.

SPEAKER_00: Oh, I just had this conversation with a friend, you know, when you're talking to someone, you're close with, you kind of know what they're about to say. They don't even really have to. It gives you this hint that maybe there's something to that, that this next word prediction is actually fundamental.

SPEAKER_05: It is in terms of us speaking and us consuming the data. And so we may have just figured out something very basic about how the brain works. And then it does feel like we're getting to the point where, yeah, we forget about what test we run to

SPEAKER_07: figure out if it's general AI, but it doesn't feel like there's much a human can do that we're not going to accomplish with, you know, this, this latest volley. Where do you sit on when we might accomplish a general AI?

SPEAKER_00: Uh, that's a, that's a tricky, um, that's a tricky one. I think the one that is most tractable, the way to approach this that's most tractable to me is just parameter counts. If you're looking at trillions of parameters, we're almost there already. So just from that kind of standard, let's say two or three years, um, practically there are going to need to be algorithmic improvements and theoretical improvements to get this done. You can't schedule those. Unfortunately, you can lock theorists in a room with a schedule and they'll laugh. Uh, but let's see, you know, you know, maybe hardware wise, we could be ready for it in the next, uh, three or four years, but, uh, I'll put my bet on 2030.

SPEAKER_07: Feels like a good number to a good round number. And you, you, you will have a tipping point.

SPEAKER_60: It feels like when we look at vertical AIs, something like self-driving, we've certainly watched it in, in games of skill and video games. Um, feel, and we see it with words and just being able to write and communicate that we're kind of there, right? It feels like watching what Elon and some other folks are doing with self-driving, man, they seem like they're really close.

SPEAKER_103: It feels like this last AI push is going to solve it. Yeah.

SPEAKER_16: It does feel like you're really close, but remember you're really close and you're up against the wall of physics of this heat problem.

SPEAKER_00: So you're going to see some new technologies come to bear and we'll be one of them that will allow that to happen. But we're very at a, we're at a scary line. We've pushed the limits of everything. So whether they get to that without new technologies is, is the big question.

David Friedberg: Yeah. And quantum computing in relation to all of this. Yeah.

SPEAKER_00: I did my doctorate at MIT on quantum computing and, um, I love the field. I think that it's like sort of, um, an incredible Olympics for your brain. Great exercise. You're being in incredible shape, uh, but maybe not super practical yet. Um, in another 10 years, I think there'll be enough progress on the theoretical and experimental side to kind of scale quantum computers. But right now it's going to take companies like Google and IBM, just pouring cash into the program and believing in the vision of what it ultimately could be to get there. Um, it will be a useful tool, but it's not going to be super broadly applicable. And that's the thing that I think people don't understand. It's not like our savior, you need GPUs and things like GPUs to keep marching along. The quantum computer is not going to run your deep neural nets faster. And, and by the way, and this is very controversial and I think I'll probably get some, get in some trouble with my friends, uh, from grad school, but, uh, you know, it's not clear to me that with deep learning, you need quantum computing. There are a lot of problems. You look at alpha fold and some of these molecular challenges that people are looking at. It seems that, uh, deep learning is able to traverse the problem space in a way that reduces the amount of compute you need by huge orders. And so the question is, do you even need quantum computers at all? Maybe a sufficiently complex deep neural network could actually just do that for you.

SPEAKER_60: That would make sense, I think.

SPEAKER_07: And then let alone some optimization on the software driven by AI. Sure. So with AI writing the software, there's going to be massive gains there as well. I mean, that's the part that I think people are underestimating is exactly the role of AI

SPEAKER_05: in solving some of the problems around the ceilings in AI. Certainly you're using AI to design some of these optical solutions.

SPEAKER_00: Yeah, definitely. When we design optical components that is using AI algorithms, we'll use things that look like back back propagation to decide the shape of these optical components that are built out of silicon that the light travels through. So it's, it's a big, big tool in the toolbox. Um, I think there's a lot of work in ship design, computer aided design for, for ASICs that can improve energy efficiency, uh, the size of the circuits and things like this. And we're looking into it. You know, um, one thing that sort of came to mind when we were talking about where AI is going to go, there's a heat limit, but there's another one, which is a sort of operating cost expense limit for companies. Look at how much they're spending on these supercomputers. It's billions and billions of dollars. If you double that a few more times, it's going to really make a big dent in Google's bottom line. Right.

SPEAKER_05: There is. Yeah. I mean, billions of dollars are speeding tickets.

SPEAKER_07: Tens of billions are considered purchases and that's kind of as far as you can take it. Hey, speaking of like scientific breakthroughs, every nerd in the world is over the last four days watching this LK 99 room temperature, super computer.

SPEAKER_67: Yeah. I'm sure by the smile on your face, you've been on Twitter or I'm sorry, X.com. X.com. X.com. On X.com, watching people try to recreate this paper. Yes. I'm, I went straight to the prediction markets where people are placing bets because gamblers

SPEAKER_23: are so good at handicapping stuff. And they had it at 25% replicable in the next two years. What do you think's going on here? And how awesome is it that it seems like a dozen teams that are in some way credible decided it it's August, whatever.

SPEAKER_67: It's the summer.

SPEAKER_00: Let's try to replicate this for the of it at night. I, I love that. I, I feel like it's a new, a new kind of, um, scientific process. Someone posts a paper papers have gotten to the level where people actually read them. Non scientists, non like deep academics are reading them and they're trying to reproduce the results. That's actually an exceptional model. There aren't many things that you'd be able to go after this with, but this is very cool. Um, the odds that it work works. Um, I think there look like there are some holes in the data. I worked on superconductors at MIT for detecting single photons for quantum computing. Niobium nitride is what I was working on. Um, the, the critical temperature there was around 10 Kelvin here. They're saying like room temperature. Um, it is a massive leap. I think that if, if it works, it will have a big impact, but maybe not quite as big. As people think it will definitely help reactors for fusion. It'll make the magnets cheaper. We'll have more interesting experience experiments for LHC. It's not going to make your chips, uh, dramatically better. Um, there are a lot of batteries, batteries, batteries, it can help batteries. Um, you can think about EVs, things like this, but I always think it's important to kind of look and temper the expectations. Like quantum computers are not a general purpose computer. Superconductors are not going to just make everything free energy and energy transports free everywhere. Remember there, there's inductance and capacitance, not just resistance. When you calculate the power dissipated in a wire, you have L R and C R is a really important one, but it's not the only one. Okay. Capacitance. If you look at a computer chip, you know, maybe a 30, 40% of the energy dissipated capacitors. Will not go away with superconductors.

SPEAKER_05: What is your gut tells you is going on here? Like, do you think that they stumbled onto something? Do you think it's a fraud or a mistake? What does your gut tell you?

SPEAKER_12: I think they believe in it because, uh, you know, they did reduce the author count to three, uh, positioning themselves for sort of a Nobel prize.

SPEAKER_00: That takes, you know, they're very excited. So I would assume goodwill here. I think that they really do believe they found something. There might be another physical mechanism that's going on. Um, and I think we're going to find out soon. You can bet researchers at MIT and Caltech and Stanford and everywhere else are looking at it right now and trying to see if it's real, because it would be a landmark result. It would be a huge deal.

SPEAKER_07: I just love the fact that like you're saying, people know where to find the papers. The papers are kind of roadmaps.

SPEAKER_05: Malcolm Gladwell, whatever, 15, 20 years ago, started turning these papers, you know, obviously social science ones, whatever into pop culture. Yep.

SPEAKER_07: Darryl Morey started reading them, you know, from, you know, the, the basketball, uh, GM,

SPEAKER_05: everybody started getting into, you know, like, Hey, let's, let's read some fundamental science here. Uh, Michael Lewis, et cetera. And it became part of our culture. And I don't think it's, I don't, I don't think it's, uh, either we're in a simulation

SPEAKER_23: or this Oppenheimer film and Christopher Nolan releasing the Oppenheimer film at the same time

SPEAKER_67: as his paper means we're in a simulation because it's all happening at the same time.

SPEAKER_23: But I do think it's sparked in a lot of scientists and people with the, you know, capability

SPEAKER_58: that, well, it would be quite fun to do science in public.

Chamath Palihapitiya: Yes. And what an amazing moment for society that you could capture people's attention and everybody wants to try to replicate it. I think it's absolutely fantastic to inspire a lot of young people to pursue hard science.

SPEAKER_126: Yeah. A hundred percent.

SPEAKER_127: All right. Listen, you're listening to the next unicorns, right? That's table stakes. Well, if you want to start on the path to becoming a unicorn yourself, you're going to need to find and hire great candidates. And how do you do that? If you want to attract LinkedIn, of course, you're on LinkedIn. I'm on LinkedIn over 900 million users. Now the March to a billion continues for the amazing team over at LinkedIn. Then you can attract both the passive and the active job seekers.

Chamath Palihapitiya: And all you have to do is put that purple hiring ring on your profile. Then you post some interesting content. When they see that purple hiring ring, it says, whoa, wait a second. I know that founder. I know that CEO. I know that VC. Well, they're hiring. Let me click and check it out. And then all of a sudden you start getting this great inbound, right? We love LinkedIn. I mean, I've gotten some of the most amazing people. In fact, I just got a new personal assistant because things are going so well. And I've got the new accelerator that we have coming in San Mateo. And I need somebody to help me run that and set it all up. And I said, you know what? I need to, I need to have an executive assistant again. So we found somebody amazing on LinkedIn, of course. And so here's your call to action. LinkedIn jobs helps you find qualified candidates. You want to talk to fast. Post your job for free at linkedin.com slash unicorn. That's right. LinkedIn.com slash unicorn to post your first job for free. Terms and conditions apply because LinkedIn is so generous.

SPEAKER_16: You know, I remember when I was in undergrad and I was sort of asking, why are scientists not rock stars that are paid like crazy?

SPEAKER_00: That was around 2005. And, you know, Google was, was coming up on getting ready to make a lot of rock star, uh, you know, scientists and engineers who are building these things. Um, but I heard that the first time that really happened was around the Manhattan project. Hmm. You had Feynman and Einstein and Oppenheimer and all these people, and they really were celebrities and people were inspired and they were excited about science and engineering. I think we're at another moment like that right now. I think it has a lot more to do with AI than anything else. Yes. Yeah. Uh, but yeah, you know, I think that the science rock star moment is here. I'm a scientist and you know, I'm not a rock star. Uh, maybe someday. Yeah. Uh, but you know, that's a potential potential potential.

SPEAKER_05: Well, I'll tell you what the diversion was. I think the diversion was there were so many applications to hit consumers that building consumer apps in this and consumer products in the style of Steve jobs and bill gates and then eventually Zuckerberg and Elon that that was a quicker path to, you know, delighting consumers. Right. Um, but now deep tech is kind of coming back and it's like, well, wait a second. If we're going to this next more interesting. Level of applications. What's the millionth app going to do for society, right?

SPEAKER_07: Like what's the next Uber going to do for society or next Robin Hood or Facebook or Twitter, all due respect to incredible products and incredible founders. But we kind of need a next leap. We need to kind of make some fundamental, uh, leap here. And I think companies like yours are, are part of that. This LK 99 thing is part of that.

SPEAKER_60: Certainly what's, you know, the, the team at open AI has done is been eyeopening and inspiring. I think to a lot of people to maybe let's get into some fundamental.

SPEAKER_23: Deep tech and, and see if we can up and make some actual progress. Right.

SPEAKER_16: And yeah, there was a lot of skating. There was, I felt like there was a period of 15 or 20 years where everybody's just riding the internet platform, building apps.

SPEAKER_00: And that was like the whole thing, but now it's like, okay, well, this is boring. What's what's next. You know, how can we actually make progress?

SPEAKER_07: Yeah. And, and you know what we, we're, we're going to need to the next generation of problems. I don't think an app's going to solve, you know, like all due respect to Fitbits

SPEAKER_05: and apple watches and quantify itself all very important, but we really need the AI to analyze what's happening in our bodies in order to make the next generation of change. Right. It's great that I can wear an apple watch and know how many steps I did or, you know, Peloton. All this stuff is great, but man, when AI looks at what's going on inside our bodies and tells us like, Hey, here's a, here's what you should do next. Here's how to extend life. Here's how to cure cancer. You know, that's going to be a different level, drug discovery, et cetera. All right. Listen, Nick, congratulations. Um, you raised a bunch of money. That means you got to hire a bunch of really hardcore folks to go to Boston and help you build this. You got an in office culture.

SPEAKER_07: I assume you can't do this stuff remote, right?

SPEAKER_12: Yeah.

SPEAKER_00: So, um, we do three days a week and we find that's a pretty good sort of even thing, but I have to say in person, there's a huge difference when you're trying to innovate and you're trying to do something really hard. You need to be next to each other to talk and to kind of go through things. That's not novel, but a lot of people tried to get away with the remote thing and it really hurts culture. Ultimately, uh, it was a very damaging period that COVID period, uh, to come. Yeah, I agree. And, uh, it's time to kind of undo a little bit of that and remind people that, you know, we're alive to be with each other and to, to hang out and enjoy the company and invent things together. And that's, what's fun. That's what you do.

SPEAKER_05: You know, and the right people, I'm sure since you put your foot down and said,

SPEAKER_23: Hey, three days a week, quite reasonable. You're getting paid 60% office time is not a big ass. Um, I bet you they're stoked.

SPEAKER_12: People are stoked to hang, right? People are excited. They're excited to see each other. And, uh, you know, with 154 million, actually 160 million that we raised, um, we're expanding.

SPEAKER_00: Uh, so we're hiring a bunch more engineers and product and sales and, and sort of, uh, going into these motions. I think what you're going to see from light matter over the next year and a half is extreme volume on groundbreaking optical products. And, uh, you know, what I'm really aiming to do with light matter ultimately is to build the photonics company. So if you think of the word photonics, I want you to think of light matter ultimately. And the way that's going to happen is through the whole set of technologies that what we've developed, uh, can kind of be realized. Super easy.

SPEAKER_05: If this sounds like the job for you, you go to light matter, you.co.co slash people.

SPEAKER_07: You see all the great people working there and you go ahead and look at the open positions and go change the world. Listen, uh, congrats on the all-star roster of investors and, uh, really excited that you're working on this, Nick, and we'll see you all next time on this week.

SPEAKER_01: Next up, we have a great talk from the Founder University podcast. LinkSquare CEO and founder Vrshal Sanak breaks down how he grew from 1 million to 10 million in ARR in just two years. Stick with us. And if you like the content subscribe to Founder University available on all podcasting platforms and YouTube.

SPEAKER_151: I'm really excited to share this awesome playbook. How to grow your startup, uh, one to 10 million ARR in two years. What does it mean to be the best in SAS? Well, SAS as an economy is like three quarters of a trillion. And so when you, when you look at data around who's made it to 1 million in ARR, it's about 4%. And who's made it to 10 million is 0.4%. So we're really going to focus on the strategies to first put you in the 4%, then put you in the 0.4% today. Uh, be here in the moment. The slides are all available and I know this is recorded. So go back and take a look at it and, and all the content is really available to you. As an introduction, I'm Vishal. Um, I've recorded this in, uh, Boston where our headquarters is for the companies I founded LinkSquares. I have a couple engineering degrees and spent the last decade working in SAS.

SPEAKER_159: Little side about LinkSquares founded in 2015.

SPEAKER_151: Uh, we're the leading, uh, multi-solution enterprise legal management platform. Uh, we are an AI company. That's a fancy way of saying we sell, uh, software to in-house legal teams. I've raised 161 million bucks. I have over 400 employees. So we talked about the 0.4% and there's so much pattern matching that venture capitalists have to describe the journey of the best SAS companies of all time, specifically first to a million ARR, which really doesn't matter how long it takes. I think UiPath did it in like 14 years. So take your time there. When you reach a million ARR, then you're really on the clock on the pattern matching of all the other great SAS companies who really did the one to 10 journey in under three years. I'm here providing these tips because we did it in two years and three months. And I'm going to tell you how it comes down to four P's people, product development, predictability, and go-to-market philosophy on venture capital. Let's talk about people. As a CEO, my job is to hire the best executive team I possibly can hire experts. And I've been trained in my decade of being a individual contributor and working at SAS companies with amazing experts that really taught me everything. So I created that here, this expert and apprentice model works, works really, really well. And this enables everything to happen. So definitely hire top down with the best world-class executive team that you possibly can. And I'd caution you on handing out executive titles too early. It's really hard to have a conversation with someone that you thought might be running engineering or sales from the early days to when the company gets much bigger, like say after 50 people to layer them or give them a boss. That's a really hard thing to do. So don't set yourself up for failure by handing out too many executive titles too early. Definitely have to balance the ego on that. But a company of six people don't need five senior vice presidents. Do things smartly in terms of executive titling and try to avoid it if you can, unless you know the person you hire is the best that there could possibly be ever. As CEO, I tell my executive team what is important, but really never how. Shame on me if I hire a marketing expert that I'm trying to tell how to do marketing too. Uh, and that really enables them to run the company in like the quarter we're in or the year that we're in. They're the one going out and running the, the operating plan that we put together maybe a year ago. That means I'm solely responsible for the future, like all of it.

SPEAKER_164: The next time we raise capital, where are we going with the product? What do we want to do strategically? What's our next move? What's our three year move?

SPEAKER_151: I often say the best days for me as CEO is I have like figuratively nothing to do because all the experts have really hired more experts underneath them. We've hired other great folks and they are the ones actually running the company. I'm more like managing the executive team. Think about it that way. Don't tell an expert how to do their job. Product development. So super important as a software company, I will give you a guaranteed way to fail on this journey, which is not doing your research and not understanding your ideal customer. Customer development and that understanding of your ideal customer is the number one priority before you build one piece of software feature or functionality. And I often say you have to know your buyer really inside and out. You have to know everything about them. And for me, uh, selling software to the general counsel, me not being a general counsel, I really didn't know anything about them. We had an idea thought, you know, understanding what was in contracts might be a great idea, but we have to learn so much about the general counsel. What motivates them? Where do they go to school? How do they think about the world? What do they read online? Where do they not read online? What software do they use today? Are they interested in buying new software? Where are they on the power line of the executive team? All of it. And until you know these things, you're not going to be able to grow as fast as you want with confidence. I always say the goal is before you build the software, you have talked to 100 of your ideal customer. I cannot stress how important this is. Please do not try to shortcut this because you will build the wrong products for the wrong problem for a buyer that does not care about it. And yes, solving the right problem has never been more important in the early days of the journey. And there's so many misconceptions about how you become a startup founder or a software startup founder. A lot of people think, hey, we have a business problem. Immediately, I'm going to build a solution for it. Then I'm going to put it next to a customer.

SPEAKER_169: And I got to tell you, that's just going to end up in a big sad face.

SPEAKER_151: You have to validate what you're doing. You have a business problem, validate that that problem exists with your ideal customer, your ideal buyer, then build the solution. Another way of saying it is ask yourself if you took a multivitamin today, did you take a multivitamin today? That's right. I didn't either because people don't buy vitamin type of products that will make your life better, but not actually fix big problems that are going on. But if you break your arm, you're going to need a real strong painkiller. That's the type of software that you're going to have to try to create painkillers. People don't take their vitamins. We use a train car philosophy. We've thought about it this way over the last seven years, running the company around what should our engineering team do? How should products think about giving them work to do? There's three cars of the train figuratively. The blue car is infrastructure, bugs, technical debt, kind of behind the scenes stuff, performance related stuff, super vital to keeping the app healthy and alive, usable, fast. The purple is the customer, everything. So bugs that they have, but also features that they desire, features that you can make money with. And then the green is like the roadmap and the vision. So I would challenge you every month or every quarter to really know what is the mix in these three categories of the stuff your engineering team has been doing. Because if you never get to the green part of the train car, you're actually never going to get the company to actually where you want it to be in the future. And remember, if you're the CEO, you're in charge of the future alone. That's your responsibility to make sure the company gets to where it has to go. If you do too much of the blue and the purple, it can cause derailment. Know what you're doing, track it and assess it at that high level quarterly. I always say our customers have their hands on the steering wheel and the company is in charge of the gas pedal and pushing it down. We build software so that our customer is more delighted. And when our customer is more delighted, either new customers that we're trying to acquire or existing customers, good thing happens in the form of ARR. So the faster you build it, the faster the ARR is made, but try to let your hands come off the steering wheel, let your customer's hands get on steering the product roadmap. If you ask them, they will surely tell you what they would love to have improvements in your product. These are the five most dangerous words inside your company. It would be cool if blank. And I feel like in 2023, we're at a fever pitch level of amazing, shiny technology that everyone can chase and embed and really get excited about internally. But you have to validate it with real customer feedback. Embedding some sort of technology will not always do the things that you wish it would. And if you're finding yourself saying, it'd be really cool if we did this. Well, yeah, things that are really cool don't really matter. I always say that unless our customers think it's really cool. Avoid the things that you think are cool. Do the things that buyers want. Predictability and go-to-market is so important, especially as you start getting your revenue flywheel running. In the early stage, track data on every opportunity you create. The source, like where it came from and the tactic that was used. So maybe it's an outbound method through email. That's how we got the meeting set with the buyer. Then two dates, the initial qualification call. And then the date that the demo happened. That's basically what you need to look at trends over time. Oh, what sources are really doing really well? And which tactics are not working so well? Which ones should we experiment with? Because ultimately you need predictability. So you're experimenting until a million and then you're really using what you learned to accelerate to 10 million at ARR. You will likely hire an expert in sales. I would recommend that you do probably at a million at ARR. Founder led sales cannot go on forever and it shouldn't really as the founders and other early team members are going to go on other missions inside the company. But these metrics are the ones that are the standard set for everything revenue, everything ARR on the new business side, right? Finding new customers. Product productivity, conversion rate, weighted pipeline, average selling price, sales cycle. You will track all of these and you should and monitor them and view them over time, how they move up and down. One thing that I hear a lot from founders is when they hire a sales rep, they might expect that they are productive maybe in the first month that they hire them. And so I would just tell you that when you're forecasting productivity for new sales reps that are focusing on new business, you build in an appropriate amount of ramp time. It will not be 100% productivity the month that they start. Figure out what that ramp time is for your reps and then build that into your financial modeling. There's a huge help for you to actually get closer to what reality is and you really need to when you model out the future. Data is the key to success in sales. I remember partnering with my sales team in the early days and saying, y'all, I know that tracking this stuff is a headache.

SPEAKER_176: And yes, there are so many required dropdowns that you have to fill out before you hit save on

SPEAKER_151: an opportunity, but the more data that we have in the executive team level, the easier it's going to be for us to figure out how you're going to make more money faster. And that partnership is so important. You're going to create that partnership forever started in the early days and really build a culture around, we need this data to help the company go faster. When it comes to hiring folks in the sales role, the first 30 days are really like the next 3000. You'll know whether they can understand the script, the pitch, the ejection handling, they understand the buyer, the software kind of through the training. Unfortunately, in 30 days, you probably get a good understanding that if it's not going to work, it's just not going to work. You're going to have to make an adjustment to essentially keep on your hiring plan. One thing that we didn't do that I wish we did was making more reusable sales training content. We did everything back on a whiteboard. We started the company when a hundred percent office was kind of the standard. So if you can make Google slides or PowerPoints of training materials so that as you keep on hiring more sales reps, you're going to have the ability to train them much faster. Believe me, you don't want to do it. One offs forever philosophy on venture capital. Yeah, we know what's going on, right? There's a correction in the public markets resulted in the private markets being corrected, being heavily impacted. But the good news is precedence seed is as open as it has been in the past. It's really no change there. In fact, there's more seed capital available than ever before. So get out there with a decent pitch deck, a great idea, a big market, good things should happen. Series A I always thought was the toughest round to raise, kind of that awkward teenager phase. You're not a full blown awesome company that knows everything about how you make money, but you're not like a seed stage company either. You're somewhere in between. And so it is the toughest round to raise under any market condition. In good times, they said it was like 80% attrition rate seed companies that are not going to make it. And probably in this world, it may even be a little harder. So focus on the retention, the go to market, the product efficiency. And remember, don't be disheartened. You only need one investor to keep the dream going. If you're a late stage company, you know that this is the hardest hit part of the macro economy and the private market. So you got to focus on efficiency. You got to focus on doing it durably and maybe even to profitability. I think about venture capital, like bottles of water. There is no differentiation between these bottles of water. The only differentiation is the marketing that was marketed to you and the taste of the water. I personally don't like Evian. So that's why it's not on this list. And so if you need capital to grow, which a lot of companies do, we're an AI company.

SPEAKER_181: So it's capital intensive to build an AI company back in the day, the way that we did it.

SPEAKER_151: Water, just like capital, is an essential element for your body and also for your company. Don't put too much overemphasis on it. At the end of the day, I always value choosing better people over pre-money valuation. The people that you work with for investors are basically more difficult to get rid of than, say, 10 marriages that you've had or you're in 10 different marriages at the same time. You're not going to be able to get rid of people that you don't like easily. So don't make decisions based on just price. And if you can, don't do 30 day fiance with venture capitalists. Try to learn and get to know them over an extended period of time. That way, when you show up at the time, it's time for a capital raise, you know who the good people are, people you can work with. Always choose better people. Fundraising is not real life. The things you read on TechCrunch are not real life. Running a business is real life. Having customers and a product in market and happy customers, NPS, writing you GT reviews, that is real life. Don't get too wrapped up into fundraising and reading TechCrunch. Because honestly, kind of vanity anyways. Unit economics. These are all the things that I track. I'm not going to go through them, but this is the standard playbook of what you're going to track basically forever. And as the company gets bigger, you're going to have to try to optimize and control all of these. Do not make a mistake on the formulas. It's a huge no-no. Make sure you get the formulas right. These are the formulas that we use. I'm going to leave you with this amazing framework, Georgian Partners. They're not even an investor in link squares, but this G7 framework of the seven SaaS metrics that they track and think highly about. They have a whole guide to them. They have a free XLS that you can use. I think it's Google Sheet. You can put all your information in there. And it's been massively helpful and valuable to me as a free resource. Closing thoughts. Yeah. 10 million ARR, super hard. Everyone says it, but it's not impossible. And your company really is the SaaS metrics. And in the future, you're going to have to try to control them, optimize them, know them, and forecast them. I always say we are what the spreadsheets say we are. Remember the four P's on your journey? Keep a level head. Good things will happen and terrible, bad things will happen also. But you got to try to keep a level head and a fixer attitude. Celebrate everything as much as you can. When we started this company, we took a whole bunch of pictures. I wish I took more. You will make it to 10 million ARR. You can do it. And it'll be nice to have all those as a memory. Celebrate everything that you can. Big thanks to Founder University for having me today. Thank you.