The IPO Comeback: Why Tech Giants Are Finally Going Public | All-In Liquidity IPO Panel

Summary

Brad Gerstner moderates a Liquidity Summit panel with Cerebras founder/CEO Andrew Feldman (three weeks post-IPO) and Planet Labs founder/CEO Will Marshall (four years post-SPAC) on the lived reality of going public. Feldman delivers the line of the panel — that you “get the timing right by getting it wrong for a decade” — and describes the dissonance of a massive listing where “you’ve sold no more stuff, your engineering projects have made no progress” the next morning. Marshall reframes the IPO as a legitimizing event, particularly for selling to agricultural giants, civil governments, and the 60% of Planet’s revenue that comes from defense and intelligence customers who “want to know you’re going to be around.”

The technology arc of the conversation centers on two infrastructure rebuilds. Marshall lays out the case for space-based data centers: launch costs need to fall from ~$1,000/kg to $200-300/kg, and at Starship’s trajectory that’s two to three years away. The intuition is solar — a dawn-dusk orbit gives you 5x more energy per panel with no batteries — and Planet is already launching Google TPUs and Nvidia GPUs into space as early experiments. Feldman explains Cerebras’s bet on a dinner-plate-sized wafer with memory next to compute, which makes them 15-18x faster than a GPU on OpenAI workloads, because “you will not wait for AI” the way you waited for dial-up.

The closing third is Gerstner’s thesis on the IPO pendulum swinging back. He argues that Anthropic, OpenAI, and SpaceX going public at multi-trillion valuations is the wrong frame — Planet Labs is the right one, where 90% of the value was created in the public market after going public at $2B via SPAC. Andreessen’s “stay private forever” decade pushed all that upside to private investors; Gerstner sees portfolio companies now wanting to go public at $1-5B and play in the big leagues. Feldman pitches Cerebras’s “dribble lockup” innovation — shares released over six months on performance hurdles — as the structural template SpaceX will likely use.

Highlights

”You Get the Timing Right by Getting It Wrong for a Decade”

“I think the answer is by getting it wrong for a decade. I mean that’s really the right way to get timing right.” — Andrew Feldman, 6:24

yt-dlp --download-sections "*6:17-7:25" "https://www.youtube.com/watch?v=jLICvWE7w2Q" --force-keyframes-at-cuts --merge-output-format mp4 -o "timing-wrong-decade.mp4"

”Indexing the Earth and Making It Searchable”

“Just like Google figured out how to index the internet and make it searchable, we are indexing the Earth and making it searchable.” — Will Marshall, 0:25

yt-dlp --download-sections "*0:21-1:00" "https://www.youtube.com/watch?v=jLICvWE7w2Q" --force-keyframes-at-cuts --merge-output-format mp4 -o "indexing-the-earth.mp4"

Data Centers Move to Space at $200/kg Launch Cost

“When launch costs come down to about $200 to $300 a kilogram, it would be cheaper, just simply cheaper to put the data centers in space. Now we’re about $1,000 a kilogram… I would expect it, the launch cost come down there in two to three years.” — Will Marshall, 15:00

yt-dlp --download-sections "*15:00-17:00" "https://www.youtube.com/watch?v=jLICvWE7w2Q" --force-keyframes-at-cuts --merge-output-format mp4 -o "data-centers-200-per-kg.mp4"

”How Big Is the Market for Slow Search?”

“When OpenAI uses us, we’re 15-18 times faster than a GPU… How big is the market for slow search today? It’s zero. How big is the market for dial-up?… You will not wait for AI.” — Andrew Feldman, 24:00

yt-dlp --download-sections "*23:54-24:50" "https://www.youtube.com/watch?v=jLICvWE7w2Q" --force-keyframes-at-cuts --merge-output-format mp4 -o "market-for-slow-search.mp4"

”Engineers Own Ties — I Didn’t Actually Know That”

“I learned that engineers own ties, I didn’t actually know that. And they didn’t die when they wore them. And second, I was surprised at how big a deal it was for them and their families. They were really proud in a way that sort of their parents might have heard of it or that somehow this was like a Bar Mitzvah or Quinceañera.” — Andrew Feldman, 6:24

yt-dlp --download-sections "*6:24-7:30" "https://www.youtube.com/watch?v=jLICvWE7w2Q" --force-keyframes-at-cuts --merge-output-format mp4 -o "engineers-own-ties.mp4"

Planet Labs as the Counter-Example to “Stay Private Forever”

“I hear a lot of people thinking that Anthropic, OpenAI, and SpaceX are the new normal. I actually think the public markets will be shifting back in this direction… We had this period of a decade where Andreessen was really pushing stay private forever and I see the pendulum swinging back.” — Brad Gerstner, 29:34

yt-dlp --download-sections "*29:34-30:30" "https://www.youtube.com/watch?v=jLICvWE7w2Q" --force-keyframes-at-cuts --merge-output-format mp4 -o "pendulum-swinging-back.mp4"

Key Points

• Cerebras IPO mechanics (4:40) — Gerstner: priced at $185, range was taken up 2x; opened at $320; today ~$230, $50-60B market cap
• CFIUS overhang from UAE investors (4:40) — Made going public hard under the Biden administration
• Three weeks in, nothing has changed (3:13) — Feldman: “you’ve sold no more stuff, your engineering projects have made no progress”
• The garbage of going public (3:13) — 130-person Zooms, commas moving in documents, no value added
• Going public as legitimizing event (8:30) — Marshall: agricultural customers and defense customers “want to know you’re going to be around”
• Planet Labs is 60% military revenue (11:33) — Bigger than they would have guessed at IPO
• 200 satellites imaging Earth daily (10:34) — Time-series of everything happening on the planet
• Miniaturization is the bigger story than launch cost (12:10) — Satellites went from $1B / 20 tons to a few kg; mainframe-to-desktop revolution for space
• Space-based data center math (15:00) — Sun-synchronous dawn-dusk orbit gives 5x energy per panel, no batteries needed
• Already launching TPUs and GPUs (17:00) — Planet has Nvidia GPUs in space, partnering with Google to launch TPUs
• Compute in space within 10 years (17:50) — Marshall: “most compute will be put in space… trillions”
• Counter: cluster comms in space is unsolved (18:08) — Feldman: “we’re not good at doing it on the ground, we’re really not good at doing it in space”
• Last 10% problem (18:37) — Feldman compares to self-driving’s decade-long last 10%
• AI opened compute to images and language (19:46) — Computers were historically bad at both; could only store
• The Cerebras bet: don’t look like a GPU (22:50) — “If you want to be 20 times better than somebody, your architecture can’t look like them”
• Memory-next-to-compute (23:32) — Dinner-plate chip vs postage-stamp; the fundamental problem in AI is moving data from memory to compute
• 15-18x faster than GPU on OpenAI workloads (24:00) — Feldman: “you will not wait for AI”
• Planet at $2B SPAC IPO; 90% of value created in public markets (25:17) — The counter-example to private-market value capture
• VCs who held mostly captured the 10x (25:46) — Google “hasn’t sold a share”, largest single investor
• The MongoDB lesson (26:31) — Altimeter distributed at $3-4B, went to $50B over 24 months; LPs pressured the distribution
• Cerebras dribble lockup innovation (26:31) — Shares dribbled over 6 months on performance hurdles; SpaceX likely to copy
• More money is made after IPO than before (27:40) — Feldman: “every single study shows” this in percentage and absolute terms
• SpaceX would need quadrillion-dollar liftoff (28:18) — Marshall: most big tech went public at “a few billion, not a few trillion”
• The pendulum swinging back to early IPOs (29:34) — Portfolio companies now thinking about going public at $1-5B
• “Iron sharpens iron” (30:30) — Gerstner: public-market scrutiny sharpens focus and accelerates innovation
• Planetary intelligence as the next AI frontier (31:13) — Marshall: LLMs are “blind” to the real world; large earth models needed

Mentions

Companies

• Cerebras (0:04) — Wafer-scale AI silicon; recently public, $50-60B market cap
• Planet Labs (0:17) — 200-satellite Earth imaging fleet; SPAC IPO 2021
• Altimeter Capital (0:33) — Gerstner’s firm; Cerebras pre-IPO investor
• SpaceX (10:07) — Going public; reference for “dribble lockup” template
• Starlink (14:26) — Transporting far more data around the planet
• OneWeb (14:26) — Satellite communications
• Google (16:00) — Planet’s largest single investor; data-center-in-space partnership; TPUs to be launched
• Nvidia (13:17) — GPUs already launched into space by Planet
• AMD (22:05) — Lost share when cell phone compute emerged
• Intel (22:05) — Lost share to ARM in cell phone
• ARM (22:05) — Won the cell phone era
• OpenAI (23:54) — Cerebras customer; 15-18x faster than GPU
• Anthropic (29:34) — Multi-trillion private-market reference
• NASA (12:10) — Planet customer
• Anduril (10:07) — Reference for military-tech framing
• Groq (13:17) — Different approach to AI silicon design
• Draper Fisher Jurvetson (25:06) — Earliest Planet investor
• Capricorn (25:06) — Early Planet investor
• Founders Fund (Peter Thiel) (25:06) — Planet investor
• DST (Yuri Milner) (25:06) — Planet investor
• MongoDB (26:31) — Altimeter case study: distributed at $3-4B, went to $50B
• Cisco / Juniper / Arista (22:27) — Companies born from Nortel’s collapse
• Nortel (22:15) — Couldn’t build chips when data networking emerged

Products & Technologies

• Wafer-scale chip (Cerebras) (23:42) — Dinner-plate-sized, memory next to compute
• TPUs (Google) (17:00) — Being launched into space
• Starship (15:00) — Trajectory to bring launch costs to $200-300/kg
• Sun-synchronous dawn-dusk orbit (15:00) — 24/7 sun exposure, 5x energy per panel
• Earth imaging time-series (10:34) — Planet’s daily-imaging product
• Dribble lockup (26:31) — Cerebras innovation; 6-month share release on performance hurdles
• SPAC (25:00) — Planet’s 2021 path to public markets at $2B

People

• Brad Gerstner (Altimeter) (0:33) — Moderator; Cerebras pre-IPO investor
• Andrew Feldman (Cerebras CEO) (0:12) — Three weeks post-IPO
• Will Marshall (Planet Labs CEO) (0:25) — Four years post-SPAC
• Elon Musk (15:00) — Starship and SpaceX reference
• Marc Andreessen (29:34) — “Stay private forever” architect of the last decade
• Donald Trump (POTUS) (1:49) — Jason’s Davos moderating story
• Peter Thiel (25:06) — Founders Fund as early Planet investor
• Yuri Milner (25:06) — DST as Planet investor
• Tucker Carlson (1:37) — The Japan ski trip Jason cancelled

Surprising Quotes

“What’s clear so far is I need to upgrade my name drop game. I mean that — that was a tour de force.” — Andrew Feldman on Jason’s POTUS-Davos story, 3:04

“You go there and you have this enormous event, and the next morning you’ve sold no more stuff, your engineering projects have made no progress since the day you weren’t public, and you go back to work.” — Andrew Feldman on the IPO, 3:13

“One of our leaders, her father, a Chinese immigrant, said ‘I thought it would have happened faster’.” — Andrew Feldman on immigrant parents at the IPO party, 7:05

“You took a big pizza-shaped die and said, fuck it, yolo, this is it, and you were right.” — Brad Gerstner on Cerebras’s chip bet, 13:17

“If you want to be 20 times better than somebody, right, your architecture can’t look like them.” — Andrew Feldman, 22:56

“How big is the market for slow search today? It’s zero. How big is the market for dial-up? It’s zero. You will not wait for AI.” — Andrew Feldman, 24:24

“All the cool stuff that we’re doing with LLMs now is really based on just the text of the internet being absorbed into these models, which is incredibly powerful already. But they don’t know shit about the real world.” — Will Marshall, 31:13

“Now for the equivalent liftoff, SpaceX would have to be aiming at quadrillion dollar valuations. Now I know Elon has those sort of ambitions, but you really have to believe in that.” — Will Marshall on SpaceX as IPO frame, 28:18

Transcript

Jason Calacanis: 0:00 Hey, 2026 could be an all-time record for IPOs.

Chamath Palihapitiya: 0:04 The AI IPO of the year so far, that company is Cerebras.

Jason Calacanis: 0:09 Cerebras Systems founder and CEO Andrew Feldman.

Andrew Feldman: 0:12 We are participating in something extraordinary. On everything we do, we are the fastest, bar none.

Jason Calacanis: 0:17 Will Marshall is the co-founder and CEO of Planet Labs.

Chamath Palihapitiya: 0:21 Space and AI are really a match made in heaven. They’re getting married, in fact.

Will Marshall: 0:25 Just like Google figured out how to index the internet and make it searchable, we are indexing the Earth and making it searchable.

Jason Calacanis: 0:33 He’s got his glasses, the famous red glasses, Brad Gerstner’s here, founder and CEO of Altimeter Capital, a leading tech investment firm.

Brad Gerstner: 0:41 I believe that the wave is the biggest wave in the history of technology. It will be incredibly beneficial for America. I’m rooting for all of them because I’m rooting for America.

David Friedberg: 0:51 Ladies and gentlemen, please welcome Brad Gerstner, Will Marshall, and Andrew Feldman.

Jason Calacanis: 1:01 Good to see you on the couch. Let’s switch it up on the couch.

Brad Gerstner: 1:05 Hey, good to see you man, how are you? Happy birthday.

Jason Calacanis: 1:13 Thank you buddy, nice to see you. Last time I saw you, we were in Davos.

Brad Gerstner: 1:15 Yes.

Jason Calacanis: 1:16 We were in Davos causing trouble.

Chamath Palihapitiya: 1:17 Another name drop! Another J-Cal Davos name drop! Do you hear that little Davos yawn-o?

Jason Calacanis: 1:21 I was just, you know, pre-IPO, we were chopping it up, Davos.

David Sacks: 1:25 We were in Davos hanging out at Davos.

Jason Calacanis: 1:30 Well, no, listen, I was supposed to — everybody knows the story. I’m supposed to go on my yearly Japan ski trip. Sacks calls me.

David Sacks: 1:37 With Tucker.

Jason Calacanis: 1:38 Yeah, anyway, we don’t drop that name, but I’ll pick it up for you, put it over here. So anyway, I cancel on Tucker, I cancel the ski trip because Sacks calls me, says ‘Listen, POTUS needs you, the world’s greatest moderator, in Davos.’ I said ‘no problem.’ I said ‘Sacks, POTUS, and Davos.’

Brad Gerstner: 1:45 So you said Sacks, POTUS, and Davos.

Jason Calacanis: 1:49 So I said when? He says ‘in three days.’ I say ‘you got it.’ I go and they give me a badge, and it’s like the special green badge and they buzz you through the security. And I look at the monitor and it says ‘Jason McCabe Calacanis with Donald J. Trump.’

Brad Gerstner: 2:06 Oh wow. How did you feel?

Jason Calacanis: 2:08 I thought it was hilarious. So then I went and we did a great interview there, and we did like six or seven of these great all-in interviews, and it was fun. Let’s start this because the two of you guys run two of the most interesting and consequential newly public companies in the stock market. Andrew Feldman is the founder and CEO of Cerebras, Will Marshall is the founder and CEO of Planet Labs. But you are also the insight and gateway for all of us to understand these two big trends. One is in AI silicon, the other one is in space data centers. I think it would be a really interesting thing to —

Will Marshall: 2:44 And emerging.

Jason Calacanis: 2:45 And emerging, yeah. But let’s just take one step back. You just heard the last conversation about being public, going public early. Let’s just talk about that because I’m just very curious, how’s it been? It’s been three weeks or so for you, it’s been about a year and a half or two years for you. How’s it been? It’s more fresh for me, you know. Was it every — was it everything that you thought it would be?

Andrew Feldman: 3:04 What’s clear so far is I need to upgrade my name drop game. I mean that — that was a tour de force.

Jason Calacanis: 3:10 Yeah! But by the way, you were — you were in Davos with JCal.

Andrew Feldman: 3:13 I was. I was there, but that — um — tour de force. Look, I think you do all this work and I think it’s really difficult to overestimate the amount of garbage that’s involved in going public. The number of meetings where you look on the Zoom and there are 130 attendees, and the amount of times you review these documents and the commas move and just no value’s added. You go there and you have this enormous event, and the next morning you’ve sold no more stuff, your engineering projects have made no progress since the day you weren’t public, and you go back to work. And um, you have some new constituents that you have to address and communicate with, but the core parts of your business — you have more money in the bank, but not a damn thing changes in the important parts of your business. If you still — if you need new supply, or if your relationships with your vendors are bad, they’re still bad. If they’re good, they’re still good. And so I think what we’ve seen is your employees have a party, everybody’s really excited, and you put your head back down, you high five and you go back to work.

Brad Gerstner: 4:40 Can I — can I just give a little context and then I want to hear from Will. You know, if I can, Andrew. You know, we were investors in Cerebras, I was on the board a year earlier where we were trying to go public. And, you know, aside from just being a warrior who weathered a decade worth of storms that would have taken out any normal human being, the path to going public for Cerebras was a particularly challenging one. One of their investors was the UAE, so there were questions about CFIUS in the prior under the Biden administration challenging to get public. My observation outside looking in is everything was really hard until it got really easy. Like nine and a half years of really hard, and then 12 months, you know, of really easy where everybody wanted to get in. They priced the IPO at 185, which was up — the range was taken up two times. Okay. The stock opened at $320 a share, I think. Today it’s at 230 bucks a share, 50-60 billion dollars in market cap for a business like, you know, and Andrew’s just one of these people let’s let’s get back to work and build shit. But my just add-on question to that is from an employee morale perspective…

Jason Calacanis: 6:00 Like distraction perspective etc., has the last three weeks, you got a lot more capital, you got a lot more profile, presumably it’s easier to sell to enterprise customers today. Net-net, if you were advising me, if I was in a similar position, would you say go public?

Andrew Feldman: 6:17 I think the first thing is a lot of people asked us about how we got the timing right.

Jason Calacanis: 6:23 Right.

Andrew Feldman: 6:24 And I think the answer is by getting it wrong for a decade. I mean that’s really the right way to get timing right. Um and I think first I, we’ve been at this for more than a decade and we brought everybody who’d been with the company more than nine years to share and we brought their families. And first I learned that engineers own ties, I didn’t actually know that. And they didn’t die when they wore them. And second, I was surprised at how big a deal it was for them and their families. They were really proud in a way that sort of their parents might have heard of it or that somehow this was like a Bar Mitzvah or Quinceañera or something.

Jason Calacanis: 7:04 Yeah. Yeah.

Andrew Feldman: 7:05 And then you had these sort of the children of immigrants. One of our leaders, her father, a Chinese immigrant, said ‘I thought it would have happened faster’.

Jason Calacanis: 7:25 Yeah, right.

Andrew Feldman: 7:26 And but I think we are sort of by nature sort of in the trenches people. And so we love solving hard problems and so when we had this excitement everybody went and they were so excited and we had a party and I think it gave external validation and then everybody turned around and they said ‘now what are we… now back to work’.

Jason Calacanis: 7:58 And so you started off kind of bang right out of the gates. Will, you had a little bit different experience in terms of, you know, the entry to the public markets, but over the last 12 months your stock has gone from five bucks a share to 50 bucks a share, some 10x move in the public markets. So talk us through the other side of this where you come public, nobody really notices until they notice.

Will Marshall: 8:30 Well we were one of the first space stocks and and I think people just had no idea what on earth is going on in space, how it was changing everything and they were just like what the heck is that and but you know I have similar opinions. I mean in the end you’ve just got to get on with executing the business. Going public gives you access to liquidity for early shareholders whether that’s the early employees or early investors and that’s great. It gives you cash for the company, that’s great. And I do think it helps your business as well because the maturing event gives you more credibility… ability to various customers. And for us we work with biggest agricultural customers, big governments, civil governments, defense and intelligence, all of those sort of actors, they want to know you’re going to be around.

Jason Calacanis: 9:10 Exactly.

Will Marshall: 9:11 And not going to disappear. I mean we have countries fully dependent on us giving them information. They don’t want to just disappear. So they really care that we’re going to be around and being a public company gives you the kind of force in the world that people go, okay, you’re here to stay, and you have access to capital if you need it and so on, right?

Jason Calacanis: 9:28 It’s a legitimizing thing.

Will Marshall: 9:29 And you know, you, you know where the stock is at any one day, you know, we’re not focused on that day-to-day. We’re focused on how we build long-term value for our shareholders, right? And, you know, the market is I think starting to really understand where space is going, why it is changing the world. You know, people forget how space is part of your everyday life, every time you use a phone, you’re using communications, using satellites, or GPS using satellites, or satellite data in some way or another, it’s sort of integrated in your lives, you may not realize it. But it’s just booming now.

Jason Calacanis: 10:07 And the story’s changed as well, obviously with SpaceX going public, but has the framing of Planet gone from like a data source for people who need data from space and maps to, hey, this is a tool to accomplish tasks and military, like, post-Anduril success, like, you probably would have been bucketed into Anduril as a military tech company. So is that framing what’s driving a lot of the…

Will Marshall: 10:34 Well, I think it’s a bit more nuanced than that. I mean, firstly for the audience’s benefit, what Planet does, we have satellites doing Earth imaging, we have the largest Earth imaging fleet, about 200 satellites, they image the entire Earth every day. So think of it like the Google, satellite layer on Google Maps that you can look at, except it’s today’s date rather than three years old. And we have every day going back. So it’s a time-series analysis of everything going on on the Earth. That’s useful for farmers, it’s useful for energy companies, it’s useful for civil governments for flooding and fires, it’s useful for security applications like you’re getting at. And it’s a wide variety of use cases. I think that where we’re seeing this is that AI is now enabling the, it’s basically reducing the barrier to entry so that more people can get access to this, right? And, you know, there’s a lot more to say on that, but AI is only as good as the data it’s trained upon.

Jason Calacanis: 11:27 What, what percentage is military, I’m curious. What percentage of revenue customer base is military?

Will Marshall: 11:33 It’s about 60% of our revenue today. Security is part of the initial thing that we said we would do, out of the gate, but it’s true it’s a bigger fraction today than perhaps we would have guessed. But the needs of the geopolitical situation right now demand what we’re doing. You know, just as an example, what this does is enable them to see threats around the corner, and then, you know, give them weeks or months advance warning of things, and then that allows you to provide… enables them to more likely do things that stop conflict. So, we believe this is, you know, really better for the world.

Brad Gerstner: 12:06 Are you reticent to be perceived as a military company?

Will Marshall: 12:10 Not really. I — but I wouldn’t say we’re limited to being perceived like that, right? We are helping farmers, we are helping, you know, energy companies, civil governments. We work with NASA, we work with what have you. And so, it’s a — it’s a bigger play than that. But back to the space piece of it, what has changed, obviously rocket costs have come down about four, five X over the last 10 years, which has helped tremendously. But a thing that people don’t know that is actually perhaps more important is that we’ve had a miniaturization of satellites. So that the same satellite that used to cost a billion dollars and weigh 20 tons now costs a few kilograms or few tens or hundreds of kilograms and can do just as much stuff, if not more. It’s — it’s the same as the sort of mainframe computer to desktop — it’s computer revolution for space and it’s unlocking — just like mainframes to desktops unlocked loads of applications, this is unlocking loads of applications and it’s — so both go in combo, the launch costs coming down and this.

Brad Gerstner: 13:17 Let’s build — let’s build on this. So, I think — I’d like first you maybe take a few minutes and then I want to talk to Andrew the same question. Both of you guys are at the foot of what are probably huge secular trends in technology. How I would frame this is we are rebuilding the data processing infrastructure that has existed on the earth in the sky. And first you do the satellites, but I would love for you to explain space-based data centers because I think everybody’s hearing about that, are they really viable, what are they, how will they work, etcetera. And then Andrew, this is the rebirth of silicon. We’re gonna find the next version of Moore’s Law, which I think is more time-bounded, not transistor density bounded. We now hear a lot about domain-specific architectures, we hear — I mean, your chip was just a complete transformation in terms of the design principles that, you know, like at Groq we took a very different approach, Nvidia’s taken a very different approach. You took a big pizza-shaped die and said, fuck it, yolo, this is it, and you were right. Just explain where we’re going in silicon. So maybe, Will, you start and then Andrew, you start.

Will Marshall: 14:26 I mean, what we’re seeing firstly in space is — is all these new applications based on data and AI. So, you know, we’re — we’re collecting vastly more data about the planet and with SpaceX and Starlink and OneWeb, they’re — they’re transporting far more data around the planet. As you say, we’re sort of changing the — the nature of data using satellites and that’s basically doing what was once the province of governments only and giving everyone else access to satellite capabilities. And that’s going to — I mean, I — I estimate there’s a 75 to 100 billion dollar market just on earth observation. The vision that’s kind of data we collect and AI on top of that unleashing all that application. So that’s the near-term thing. Applying large language models to Earth imagery data, unlocking agriculture, you know, energy, civil government applications, permitting, you name it. This is going to make everything more efficient.

Will Marshall: 15:00 And then where we’re going is indeed, space is… we did a study with our partners at Google about eight or nine years ago looking at what are the costs of data centers on the ground, what are the costs that it would take to put them in space, and when might it make sense to do it non-terrestrially? And we figured out that when launch costs come down to about $200 to $300 a kilogram, it would be cheaper, just simply cheaper to put the data centers in space. Now we’re about $1,000 a kilogram, just over that, right today. But that’s come down about 10x in the last 10 years. On the current trajectory with Starship in particular, I would expect it, the launch cost come down there in two to three years. Elon might say it’s next week, but at least realistically a couple of years. So we’re not far away from it literally just being cheaper. Then in addition, and the intuition there that helps people understand that is you would naturally use solar panels for doing power. Data centers are a power problem, it’s a power game. And you would normally use solar panels, that’s the cheapest way to get a watt today by far. But you don’t want intermittent power. So then you have to have batteries or then you have to have gas or then you have to have nuclear, and then it gets really expensive. In space, you can put a solar panel in a sun-synchronous dawn-dusk orbit where you’re 24/7 looking at the sun. So you can have a solar panel that collects and gathers five times more energy per solar panel than on the ground. And you don’t have to have batteries or anything else. So the infrastructure for compute in space is literally just solar panels and the chips and then the RF signals up and down. So it’s actually really quite simple. It was just a question of when it’s going to be cheaper to launch all those solar panels and chips into space than putting on the ground. And it turns out that’s going to be in a few years. So we’re partnering with Google to launch some of their TPUs into space. We’ve already launched some of Nvidia’s GPUs into space. We’re launching Google’s TPUs into space on an early test. There’s lots of technology to figure out.

Andrew Feldman: 17:41 We may want to launch some Cerebras.

Will Marshall: 17:43 Let’s have a conversation. But it’s early days, but I think no question within 10 years most compute will be put in space, which to give you a sense is a lot of money, like trillions, and will be bigger than any of the other space businesses today: comms, Earth imaging… This is why we’re getting into this game early.

Jason Calacanis: 18:00 Andrew, tell me about Cerebras. Do you believe… Sending data centers to space makes more sense, or is it just the regular…

David Friedberg: 18:05 Wait, can you have him explain the business first and then…

Jason Calacanis: 18:07 Oh, of course, yeah.

Will Marshall: 18:08 I think there’s, you know, with all due respect, one or two hard problems still left beyond putting, putting GPUs in space right now. I think, um, we, we’re not super good yet at building the clusters in space necessary for the communication between…

Jason Calacanis: 18:30 Between, exactly.

Will Marshall: 18:31 Exactly, between. We’re not good at doing it on the ground…

Jason Calacanis: 18:34 We’re not good at doing it on the ground, we’re really not good at doing it in space.

Will Marshall: 18:37 I think this is an extraordinarily important and interesting problem and one we should be spending money and attacking. I’ve got it at a slightly different time frame, but one that certainly will occur. And the hard part is, is it one of those problems where the last 10% is 80% of the time? Now self-driving was a problem like that, right? Where the last 10% proved to be a decade’s worth of work and just now we’re over the hump. And we don’t know yet. But I think the interesting work they’re doing at Planet is really important and I think the fundamental driver to experiment to even get insight into whether I’m right or not is to get down the cost of launch vehicles. Then you can start doing experiments and getting it wrong and fixing it and figuring it out. And until then it was mostly on paper.

Jason Calacanis: 19:27 Yeah, so for the, for the foreseeable future, you’re going to be terrestrial. Explain your business and how you made these critical decisions that kind of took you on a different path and, you know, you versus Nvidia versus AMD and what you think the future of AI silicon looks like.

Andrew Feldman: 19:46 Well, I think there were two parts. Your first question was around sort of the rise of silicon in general. And I think what AI did, and it’s, it’s rarely sort of framed this way, but it allowed computers to address a class of problems that before AI computers were bad at. We were bad at images for almost the entire history of compute. We could store them and that’s about it. We were bad at language. We could store it but that’s about it. We could transform numbers. We were magical with numbers. And what AI did, starting in about 2015, ‘16, is it opened the door, the aperture, to say maybe we could use computers on images. All right, maybe we could find insight in images. Maybe not only could we store language but we could generate it. All right, maybe we could understand it rather than storing it and regurgitating it. And what this did is it, it opened up sort of to compute huge areas that were previously foreclosed. And at the same time, we were adding to those areas. We were taking vastly more images. All right, terrestrially, we’re… In satellites, and what this did is it simultaneously opened up this entire area and allowed compute to attack it. And this is what’s underpinning both Nvidia’s growth and sort of all the growth you’re hearing about in AI compute is, as a processor builder, as a hardware builder, suddenly our tools could attack more and different parts of knowledge. And that was sort of the first part to answer your question.

Andrew Feldman: 21:39 Now how you do that, there’s lots of different strategies, there’s tons of different ways to skin cats. What we saw in 2015 were several things. First, we saw that AI would be an enormous consumer of compute. And historically for computer architects, new workloads were the opportunity for share to change. Share changed when the rise of graphics emerged and you got the dedicated GPU that’s how Nvidia was born. Share changed when cell phone compute emerged, and Intel and AMD who had fabs and the best architects got zero share and it all moved to ARM. Share changed in the late nineties when Nortel and all these companies we’ve forgotten about couldn’t build chips and couldn’t do, data networking. And what you got was Cisco and Juniper and Arista and this collection of new companies. So we knew that this new problem would present an opportunity for massive change.

Andrew Feldman: 22:41 So we saw that. We made two bets. The first was dedicated silicon would be the answer. And the second was it couldn’t look like a GPU. And our view as computer architects is if you want to be 20 times better than somebody, right, your architecture can’t look like them. Right? It can’t. They have enjoyed and eaten all the low hanging fruit. So if you build a GPU, the odds that you’re better than Nvidia in our view are approximately zero. That led us to a fundamentally different architecture. All right. The hard part here, the hard part is moving data from memory to compute.

Brad Gerstner: 23:26 Mm.

Andrew Feldman: 23:27 This is the fundamental problem in AI. And we solved it with a way that very few others had even attempted, which was to build a very big chip and to put memory right next to compute. By building a big chip, a chip the size of a dinner plate, whereas most chips are the size of a postage stamp, we could use a different type of memory. And by using a different type of memory, a memory that was vastly faster, we opened up all sorts of opportunity. So when OpenAI uses us, we’re 15-18 times faster than a GPU. That means your answers are delivered more quickly. It means your engagement with the AI is more enjoyable. It means you can use the AI to solve harder problems and not wait. And the way to think about this is sort of to ask yourself the counterfactual question: How big is the market for slow search today?

Jason Calacanis: 24:24 Right.

Andrew Feldman: 24:26 Right? It’s zero. How big is the market for dial-up?

Jason Calacanis: 24:29 It’s zero.

Andrew Feldman: 24:30 How long do you wait for a website to resolve before you click away? Three seconds? Five seconds? You will not wait for AI. We have to deliver it to you in real-time. And that’s what we saw. That’s what we built.

Brad Gerstner: 24:47 So the panel’s on going public, a lot of LPs in the room, they need to get liquid. I’m curious about the journey for your investors. Okay, so Will, you guys went public what year?

Will Marshall: 24:58 Uh, 2021.

Brad Gerstner: 25:00 2021 by way of a SPAC.

Will Marshall: 25:02 Correct.

Brad Gerstner: 25:03 Okay. And your VCs were who?

Will Marshall: 25:06 Uh, Draper Fisher Jurvetson was one of the earliest, Capricorn, Peter Thiel’s Founders Fund, then we got, uh, Yuri Milner’s DST, all sorts of guys.

Brad Gerstner: 25:17 Great. Okay, so your investors come in, you go public at 2 billion via a SPAC. Now we’re four years later, really it wasn’t until year three or four that 90% of the value was created. Okay, so did those early investors capture this 90% move? Did they stay in it?

Will Marshall: 25:46 Most of them did. Yeah, most of them did, which was really smart on their part, obviously. I think they should hold on even more. If I didn’t think that, you’d think I was a bit self-interested.

Jason Calacanis: 25:59 Wait, what’s interesting about this…

Will Marshall: 26:01 No, but really they did. And I mean, Google hasn’t sold a share. They are our largest single investor. Capricorn didn’t until very recently. So basically most of them stayed really well in and they got all of that upside, and good for them.

Brad Gerstner: 26:22 And the reason I think this is so important is that there are a lot of LPs in this room who they’re like, ‘When a company goes public, give us the shares.’

Jason Calacanis: 26:31 No, no, no, yeah. That’s the beginning.

Brad Gerstner: 26:33 Give us… give us the shares. This is a counter-example, right? This happened to us in MongoDB ten years ago. We invested pre-IPO at a billion dollars. We distributed the shares, I think, at three or four billion. And then it went to 50 billion over the course of the next 24 months, and we had people who called us who said, ‘Well, why didn’t you hold on to the shares?’ And we’re like, ‘Because you’re pounding on us to distribute the shares.’ So you’re an example. Now, in your case, Andrew, you have an innovation, right? You’re just now public. So all of your investors are still under lockup like Altimeter, but you guys have innovated with the banks on what I call a dribble lockup. So over six months, the shares can be dribbled out according to a bunch of performance hurdles, which SpaceX is going to have a very similar…

Jason Calacanis: 27:05 When did we start this process of the, the dribble?

Chamath Palihapitiya: 27:10 The dribble? Constantly you started it years ago, but…

David Sacks: 27:13 We’re all of that age.

Brad Gerstner: 27:15 With respect to the lockup, I think this is the most innovative and I think SpaceX is going to have a very similar innovation. But Andrew for your investors, if you were talking to my LPs, right, in the room, should Altimeter be distributing the shares when they come out of lock? How do you think about, you know, your VCs holding on to the shares kind of post lock?

Andrew Feldman: 27:40 Well I think historically more money’s made after IPO than before. I think every single study shows that there is more money to be made both in percentage and in, in what we care about, which is absolute. And so, the amount of money that it’s possible to put to work in most venture companies is very modest. I mean there are two or three or five outliers, but for the most part you can only put a relatively little bit of money to work. By the time we get public, there’s a lot more money there, things are going well, and the opportunity to make vastly more is after IPO, not before.

Will Marshall: 28:18 And if I could just add on that, one interesting question is what’s going to happen with SpaceX on this because, you know, a lot of the value is, is in the future, right? But I mean most of the big tech companies went public at a few billion. Not a few trillion. Like there’s a lot of zeros in between those, right? And you’ve got all this upside afterwards. Now for the equivalent liftoff, SpaceX would have to be aiming at quadrillion dollar valuations. Now I know Elon has those sort of ambitions, but you really have to believe in that to get, you know…

Brad Gerstner: 28:53 This is, this is kind of the point I’m getting to, right? We have three mega IPOs, you know, that we keep talking about that are multi-trillion.

Jason Calacanis: 28:59 And software…

Brad Gerstner: 29:00 All of that value accrued to private market investors. Planet Labs is a great example of venture capital in the public markets where the 10x has occurred in the public markets. We’re all advocates of these companies coming public sooner. Had Andrew had his way, he would have been public 18 months ago probably at 10 billion dollars rather than 50 billion dollars and that 5x over the course of last two years would have gone to public market investors. So go ahead.

Andrew Feldman: 29:30 Way better to be lucky than good.

Brad Gerstner: 29:34 So, so I think that I hear a lot of people thinking that Anthropic, OpenAI, and SpaceX are the new normal. I actually think the public markets will be shifting back in this direction and a lot of the companies in our portfolios are now thinking about going public at a billion or three billion or five billion. We had this period of a decade where Andreessen was really pushing stay private forever and I see the pendulum swinging back. The companies are like man, I want to be like Planet Labs and get public, right? And have to play in the big leagues and do it in the public market. So here’s what I’ll say maybe just like to the two of you guys. Both of you guys have had enormous pressure because there’s visible competition that’s always sort of in your periphery. But I do think that getting public sooner, having the scrutiny of public markets, having the scrutiny of having to deliver, sharpens the focus. It steel sharpens steel.

Andrew Feldman: 30:30 For sure.

Brad Gerstner: 30:30 Iron sharpens iron and I think innovation tends to get better. And so the idea that you allow everybody to participate but you also put yourself in the spotlight, to me, is where great things happen.

Chamath Palihapitiya: 30:46 I agree.

Brad Gerstner: 30:48 And so anyway, I just wanted to say to both of you, just as we wrap, you guys are incredible testament to entrepreneurship, both of you. I mean we’ve been talking literally since day one, me and Andrew, because we went in different paths and then we kind of reconverged. And then Will, same with you.

Will Marshall: 30:58 I’m happy it worked out for you, Chamath.

Chamath Palihapitiya: 31:00 Well, it’s worked out for both of us so it’s fine.

Brad Gerstner: 31:04 You guys are incredible testament to entrepreneurship and I just want to say thank you for everything you guys are doing. The next few years are going to be really spicy.

Will Marshall: 31:13 Yeah, if I could just spend a 30 seconds on the next few years because I think it’s going to be so exciting with, as I mentioned, AI and space merging together. We’re going to see a takeoff of applications. I’d like to say that all the cool stuff that we’re doing with LLMs now is really based on just the text of the internet being absorbed into these models, which is incredibly powerful already. But they don’t know shit about the real world. I call them blind to, you know, they don’t know about that farm field, that flood, the security situation around the corner. If you give them real world data, then they can answer real world problems. And that’s going to open up gazillions of applications for these AI models. I call them instead of having large language models, large earth models or and instead of AI, planetary intelligence where you have planetary sensor systems in space, planetary compute systems in space and we can disagree or agree on the exact timeframe but I think it’s going to happen and and then that’s going to enable a huge economy. So it’s an exciting time in the next few years.

Brad Gerstner: 32:16 Will, Andrew, thank you guys very much. Well done.