Mike Schroepfer: Lessons from Scaling Meta to Investing in the Energy Transition

Summary

---

Inevitable (an MCJ Podcast)

Episode: Mike Schroepfer: Lessons from Scaling Meta to Investing in the Energy Transition
Date: March 20, 2025
Host: Cody Simms
Guest: Mike Schroepfer (“Schrep”), Partner at Gigascale, former CTO of Meta

---

Overview

This episode features Mike Schroepfer, renowned technologist and former CTO of Meta (Facebook), now the founder and partner at Gigascale, a climate-tech venture firm. Host Cody Simms and Schrep trace the arc of his journey—from early engineering, through building and scaling Meta, to investing in the next generation of climate solutions. The discussion covers the challenges of company building, the shifting narrative around America’s ability to “build hard things,” the convergence of AI and energy, and lessons learned about talent, scale, and impact along the way.

---

Key Discussion Points & Insights

---

1. Early Career & Building Things (00:00–07:17)

• Schrep’s roots as a computer graphics nerd:

  • First job: Startup in Sausalito with special effects pioneer Scott Squires; wrote motion tracking software used in Star Wars: The Phantom Menace and Gladiator ([02:27]–[03:46]).
  • Memorable for seeing code he wrote directly impact big productions:

    “I get to make things in the world that other people use... that feeling has stuck with me.” – Schrep ([03:47])

• First startup & dot-com era perspective:

  • Founded a server management software company, funded by Sequoia at the height of the dot-com crash ([05:17]–[06:08]).
  • Recounts the intense challenge of perseverance:

    “It's really, really hard to bust through all of that... the game of being an entrepreneur is fundamentally irrational.” – Schrep ([07:17]–[08:16])

  • Lessons on timing and opportunity, referencing how Webvan’s concept was too early and only succeeded years later with Instacart.

• Empathy for entrepreneurs:

  • The importance of persistence in the face of repeated rejections.

    “You just need one yes. The ratios don’t matter. I could take a hundred no's, as long as I get the one yes.” – Schrep ([08:22])

---

2. Lessons from Scaling at Mozilla and Meta (09:27–19:13)

• From Sun Microsystems & lessons on market disruption:

  • Saw how Sun failed to adapt to modular, cheaper hardware, a “big lesson... in disruption, how it impacts big companies.” ([10:24])

• Mozilla: “Business school” for open-source scaling:

  • Learned to scale a product made by 2,000 global contributors for hundreds of millions of users ([11:51]–[12:17]).

• Joining Facebook (Meta) in 2008:

  • At a time when no one knew if social networks could make money.
  • Decided based on first principles:

    “Our market is unlimited. We just got to execute and then there’s a big risk on the business model. We got to go figure that out.” – Schrep ([13:17])

• Early work at Meta:

  • Tasked with scaling engineering, keeping the platform up (“the fail whale” reference), and expanding into data centers, system software, and new products ([15:17]).

• Reflections on management & humility:

  • Making a 35,000-person org work: Ask questions, admit what you don’t know.

    “I just trained myself to be like, hey, I guess I should know that. Don’t know. Can you explain it to me?... The best part of my job is like, I get to run around and get educated by all these amazing, brilliant people.” – Schrep ([16:06]–[16:10])

  • Importance of “unique help” and not trying to out-expert engineers, but rather unlock the team’s potential.

---

3. Debunking “America Can't Build Hard Things” (19:13–21:25)

• Challenging the narrative:

  • Pushback on the idea that the U.S. doesn’t build “hard things.”

    “I would say that whole framing... really irritates me. It misses SpaceX, drives me crazy... We have built hard things... what we have to do is actually invest our time, our money and our talent in hard problems that solve things that we care about.” – Schrep ([19:13])

• Industrial policy & infrastructure:

  • China’s role acknowledged, but asserts unique U.S. startup innovation and the need to recommit to industrial investment.

---

4. The AI-Energy Convergence (21:25–26:29)

• AI’s exponential needs:

  • Recounts Facebook’s early AI lab, watershed moments in deep learning (ImageNet), and trajectory toward “outsized” AI impact ([21:25]–[23:54]).
  • Infrastructure constraints now:

    “When you break down the cost [of AI], it basically boils down to two things. Chips, buildings, and power – chips I think there's a lot of innovation happening... Power is going to be the big one.” – Schrep ([23:54])

• Power as the new bottleneck:

  • Solar is already the “cheapest way to make electrons,” and innovations like offshore floating data centers could cut compute costs by 50% ([24:42]).

• AI as a driver of energy innovation:

  • Growth in datacenters, manufacturing onshoring, and electrification will lead to a 5x increase in U.S. utility demand by 2050 ([26:29]).

---

5. Distributed vs. Centralized Infrastructure & Lessons from Compute (27:48–32:01)

• Parallels between compute and energy:

  • The evolution from monolithic to modular, mass-produced systems in data centers ([28:24]–[30:55]).

    “Anything where you can get mass manufacturing... you get learning rates, you get cost downs, the whole flywheel goes in your favor.” – Schrep ([30:28])

• Commoditization and verticalization in AI/software:

  • Ongoing tension between platform providers (e.g., OpenAI) commoditizing layers above them and the need for domain-specific data flywheels ([32:01]–[34:37]):

    “If you, in your first couple minutes of pitch, haven't explained to me how you’ve got a proprietary data flywheel going, I’m out.” – Schrep ([34:37])

---

6. Power Procurement as the Next Business Competency (35:14–37:40)

• Power strategy now central to large firms:

  • Increasing relevance for hyperscalers/Fortune 1000:

    “The way we make power and distribute it is fundamentally changed. It's both cheaper but way more volatile.” – Schrep ([36:00])

• Innovation pressure in consumer and enterprise power sourcing, especially in unregulated markets (like Texas).

---

7. Personal Climate Journey (38:06–41:04)

• From technologist to climate investor:

  • COVID gave time to reflect—slow-moving crises require new responses ([38:06]).
  • Start with philanthropy, shifted to investing after realizing,

    “It’s a $10 trillion problem. You don’t solve $10 trillion problems with government or philanthropy. You need markets.” – Schrep ([39:15])

  • Key: “Commercial potential is a way to align goals. We can have a better future and more prosperity for everyone.” ([40:21])

• Meta advisory role continues in parallel with climate focus.

---

8. Talent Movement into Climate Tech (41:04–43:30)

• Big uptick in tech-to-climate interest during the pandemic, but some decline as enthusiasm chases trends (AI boom):

  “I think what we saw was a massive wave in 2020–2022... I think it's waned since then, just to be frank.” – Schrep ([42:27])

• Importance of making climate tech jobs “zero trade-offs”:

  • Great careers and financial stability can and should align with climate impact.

---

9. Gigascale: Model & Motivation (43:30–55:52)

• Gigascale’s thesis:

  • Align commercial and climate impact by backing businesses that are fundamentally better and more efficient, making them lower cost and lower carbon ([43:32]).
  • Team combines operator/technical backgrounds to guide founders in deep tech and company scaling.

• Focus:

  • Pre-seed to Series A is the sweet spot ([45:24]), but opportunistic in later rounds for the right companies.

• Operator advantage:

  • Help founders skip “obvious mistakes” and focus their innovation where it’s most needed ([43:32]).
  • Talent and storytelling as key differentiated services (see below).

• Strategic approach:

  • Sector deep dives to identify thematic opportunities, but also flexible when serendipity uncovers standout startups ([49:49]–[50:42]).

---

10. Talent & Storytelling as Scaling Engines (51:15–55:52)

• Storytelling as essential:

  “Humans convey knowledge through stories. This is how we learn things, as nerdy and data driven as we are... The way you’re going to get customers and investors and employees... is by having a very clear story.” – Schrep ([51:37])

  • Brought on The Regenerates (communications agency) as core Gigascale team members to operationalize this for portfolio companies.

• Hiring as strategic advantage:

  • Established in-house talent service to help founders with both tactical hires and broader people strategy.

    “What’s your talent strat? ...You need this other thing instead. That’s so much higher value than just ‘okay, cool, I'll run your process’...” – Schrep ([54:10])

  • Only refer talent to founders “if they and I disagree on a topic, take their advice.” ([54:10])

---

11. Industry S-Curves, Market Cycles, & Near Future (55:52–59:38)

• Reflecting on cycles:

  • Parallel between dot-com, Clean Tech 1.0, and current energy/climate wave—caution against over-accelerated expectations.

    “Everything’s going to happen tomorrow is not true. This is what happens... It took a little longer than we thought. Still faster than anything else.” ([56:49])

• Technology cost curves as powerful tailwinds:

  • Solar 10x cheaper per cell, lithium-ion batteries 90% cheaper; “My business is premised on batteries are going to get cheaper and so they’re going to be used for more stuff.” – Schrep ([57:43])

• EVs, design, and the surprise of exponential change:

  “Why Instacart works now versus Webvan didn’t. In 2000, everyone had a fricking smartphone and so the density of deliveries made sense. That was it. That’s the difference. Otherwise, it’s the same business.” – Schrep ([59:00])

---

Notable Quotes & Memorable Moments

• On company building & irrational optimism:

  “The game of being an entrepreneur is fundamentally irrational. If you just look at expected value, it’s worse. You should just go work for a big company... But the possibility for an outsized impact... you have to have unshakable belief in the opportunity you're going after.” – Schrep ([08:22])

• On America building hard things:

  “That whole framing of we don’t know how to build stuff really irritates me... What we have to do is actually invest... in hard problems that solve things that we care about.” – Schrep ([19:13])

• On the power of compounding technology:

  “Anything where you can get mass manufacturing... the whole flywheel goes in your favor.” – Schrep ([30:28])

• On AI and data flywheels:

  “If you, in your first couple minutes of pitch, haven’t explained to me how you’ve got a proprietary data flywheel going, I’m out.” – Schrep ([34:37])

• On the importance of storytelling:

  “Humans convey knowledge through stories. This is how we learn things, as nerdy and data driven as we are... The way you’re going to get customers and investors and employees... is by having a very clear story.” – Schrep ([51:37])

• On exponential change:

  “Solar is 10 times cheaper per cell and getting cheaper still. Lithium ion batteries are 90% cheaper... my business is premised on batteries are going to get cheaper and so they’re going to be used for more stuff. I think we’re going to be okay.” – Schrep ([57:43])

• On climate as opportunity:

  “If you’re listening to this and you’re like, wait, I’m like a biz dev person or I’m really good at financing, there is a role for you. We need you, badly. All the engineers too. We need all of you. We need all of it. And just stacking these companies with talent is one way to help us make sure that we can get these next set of trillion dollar companies.” – Schrep ([60:20])

---

Timestamps for Key Segments

• Schrep’s motion tracking and Star Wars story: [02:27]–[03:46]
• Dot-com crash, perseverance for Sequoia funding: [05:17]–[08:22]
• Joining Mozilla and business school for scaling open source: [10:24]–[12:17]
• Early Meta/Facebook: risks and first principles: [13:17]
• Scaling engineering orgs & leadership lessons: [15:17]–[17:45]
• America "not building hard things" myth: [19:13]
• AI and energy’s convergence, datacenters, floating compute: [21:25]–[25:35]
• Distributed vs. monolithic infrastructure: [27:48]–[30:55]
• AI commoditizes software layers; data flywheels: [32:01]–[34:37]
• Power procurement as business differentiator: [35:14]–[37:40]
• How and why Schrep dove into climate (Covid catalyst): [38:06]–[40:21]
• Gigascale’s “operator” and talent/storytelling model: [43:32]–[55:52]
• S-curves, market cycles, and cost tailwinds in climate: [55:52]–[59:38]
• Schrep's encouragement to listeners on joining/staying in climate: [60:20]

---

Conclusion

Mike Schroepfer’s journey embodies the fusion of high-scale technology and impact investing. Lessons from dot-com booms and busts, scaling global platforms, and rallying talent toward mission-driven outcomes now inform his approach to climate tech investing at Gigascale. With a strong belief in innovation, humility, and the power of compounding technologies, Schrep encourages anyone—regardless of skill or background—to play a part in building the next generation of trillion-dollar companies for the energy transition.

---

For more episodes and info: Visit mcj.vc

Transcript

[A]

Today on Inevitable, our guest is Mike Schrepfer, or Schrep as he's known. Mike is partner at Gigascale, a venture firm that he's created and self funded as a way to invest in and support startups building venture scale businesses with the potential for outsized climate impact. Mike was CTO at Meta, the parent company to Facebook from 2013 to 2022 and he originally joined Meta in 2008 as a VP of Engineering. Mike and I have a wide ranging conversation about his career journey and lessons learned. Our shared counter opinion on the oft repeated myth that America doesn't build hard things, his views on AI and energy's role, and on the work he and his team are doing today at Gigascale. Mike's been a supporter of our work here at MCJ for years and we're honored to count him among our investors in our MCJ venture funds. It's inspiring to me to see a technologist as accomplished as Mike not just lend his resources to the space, but also roll up his sleeves and help support startups as they tackle hard problems and build the future. From mcj, I'm Cody Sims and this is Inevitable. Climate change is inevitable. It's already here, but so are the solutions shaping our future. Join us every week to learn from experts and entrepreneurs about the transition of energy and industry. Mike, welcome to the show.

[B]

Hey Cody, happy to be here.

[A]

Boy, this feels like a long time in the making. We've collaborated on a number of things for many years and you've been a great supporter of our work here at mcj. I know so many people on your team. It's a lot of commonalities between some of what you're doing at Gigascale and some of what we've done at mcj. And it's fun to finally sit down and have a conversation all about your background and what got you here in the first place.

[B]

Yeah, it feels a little bit like a reunion, so it should be a lot of fun.

[A]

I was digging in, learning more about your background before this conversation, and I thought there was a fun little almost trivia nugget to this very start of your career, which is it looks like you started your career working on a digital video software that ended up becoming part of Star the Phantom Menace and Gladiator. Yes, that's pretty cool.

[B]

It was a funny formative moment. If we can do story time for a second, that's nothing to do with climate, but it's more about building stuff. I was a super nerd in college, loved computers. Got really into computer graphics because it was 3D programming and graphics and how you create things on a computer. This is the 90s when this stuff was sort of new and novel. Well before the rise of GPUs and all the rest of it, I worked at this tiny little startup in Sausalito. And it was started by someone who worked at Industrial Light Magic and legendary special effects going back to like Close Encounters of the Third Kind. Sort of person named Scott Squires. And he was one of these polymaths that basically didn't sleep. I think he had the gene where he had three hours of sleep a night because he would be on set all day and then he'd stay up all night coding. And he was self taught. So the code was a nightmare. And it was like one giant function. He, like, started this little company. Me and one other person joined as the first engineers. And we were like trying to turn this into a product that we would sell to other people. And part of it is they could prototype a bunch of effects and shots on set. So rather than taking three months post production, they could try some things out and see how it works. And the formative moment was we sold the software to ILM and they were using it on the Phantom Men. And they were using the feature that I built, this thing called motion tracking, where you could track something through a number of frames and apply an effect. So if you wanted to like light a little ball, someone was juggling on fire and they're like, you need to go over to Skywalker Ranch, help them with your thing. Is that cool?

[A]

Super cool.

[B]

Yeah, that's cool. Drive over there. Sign an NDA that I think had a clause in there. It's like, if you disclose anything, we have Jedis on call. Open the door, and there's just a table full of all the models of all the characters and a bunch of people stuck in the corners. And it's the scene from that movie where they're in the throne room with the queen and they were using code I wrote to like work on it. And they're like, oh, we're having this issue with the motion track. And so like, I sat there and I helped them for a bit, driving home, like, all right, this is awesome. I get to make things in the world that other people use. And that was 25 years ago. Yeah, that feeling kind of has stuck with me. Worked on consumer products and we build all sorts of crazy stuff now, but there's something special about that opportunity.

[A]

Such a cool story about when you work in tech or in software, particularly at a startup, and you see your product getting used in a way that is really driving things you can tell your parents about or something, it's a magical moment. And you got that right out of school, like your very first gig. That's pretty cool.

[B]

Yeah, pretty awesome.

[A]

We won't need to talk about how you were involved in the creation of Jar Jar Binks.

[B]

I suppose I take. No, I had no creative control. Effects are great.

[A]

I got to know you as this guy who is the CTO of Meta Facebook, leaning into climate. But then your next step in your career was you co founding a startup and you went out in your very first startup.

[B]

You.

[A]

You weren't yet the CTO of Facebook. This guy with this incredible, illustrious career. You were a guy who built some visual effects software. You raised money for your very first startup from Sequoia. That's no small feat either.

[B]

It was a crazy time to do it because it was the fall of 2000, and there's this thing that happened in April of 2000 called the dot com crash, where basically everyone went out of business. And so we were the only software deal that Sequoia Capital did in Q4 of 2000, by the way, for any.

[A]

Listeners too young didn't live through that era. The number one website in 2000 for any of us who lived in the San Francisco or the Bay Area was a site called Fucked Company, which literally tracked companies going out of business. That was the entire website. And every day you would wake up to find out where were the layoffs, what company was gone. Are your friends okay? Like, it was a complete extinction event.

[B]

Yeah, it is really hard to overstate. Like, traffic on Highway 101 was markedly different. All the offices were open and like, for a little while, I basically played repo man because we just raised $7 million, literally the only company with any money. And then immediately I was getting spreadsheets emailed to me every week. Here's all these assets from the store.

[A]

Yeah, you got some cheap air on chairs, I bet.

[B]

Mostly was like servers and Cisco switches. I like, drive my little Honda yet again over to an office and was awful because everyone's getting laid off. I'm like, I'm here to pick up the two sun servers in the Cisco Switch because we got it half price. And the joke I made. So Mike Moritz was our investor, legendary investor, and a ton of amazing companies. And we bought this Cisco switch and the first company that bought it, we bought it used because these things were expensive and we Want to save money was Webvan. Now, Webvan's probably not a company that anyone has heard of. They raised hundreds of billions of dollars and just completely imploded. And they were basically delivering groceries. It's Instacart. It's the same business. It's just bad timing. And so this is a real important lesson. And sometimes there's great ideas, but you gotta nail the timing because 15 years later, that's a $10 billion business.

[A]

We're gonna come back to all the timing from lessons learned from dot com 1.0, boom and bust to your time at Facebook when the Internet became this truly global thing. And compare that to Clean Tech 1.0 and where we are now. So put a pin in that. I want to come back to that for sure.

[B]

Can I tell one story about fundraising in that timeframe, though? Yeah, yeah, yeah, yeah. Great success at the end, scoring capital. Mike Morris couldn't ask for more. I can't tell you how many no's we got. It was like 30. I definitely remember the pitch where the partner fell asleep in the pitch. So we're doing the pitch and they were like full on snoring asleep while we're doing the pitch. Definitely a lot of your company sucks. You're never going to do it. Our portfolio, Company X is doing what you do 10 times better. You definitely go out there and like the 30th time you get a no, you're like, maybe they're right. It's really, really hard to bust through all of that. And so it is a lesson that I sort of take with a lot of empathy to entrepreneurs that I back today.

[A]

Your yes was tier one, top of the tier investor. What was that like? How did that happen?

[B]

We just fricking worked it. We just did it. Kept pitching people. And this is the hilarious thing, Cody, is the first 12 slides of our pitch was trying to convince you that there was going to be a lot more servers in the world. IDC reports. Dell forecasts all these things of like, hey guys, there's going to be a lot of servers. Because we were basically. My idea was like, as servers go up, we need better software to manage them. As it gets old by hand. That idea turned out to be right. But you just got no, no, no. But this is what I tell entrepreneurs is you just need one yes. The ratios don't matter. I could take a hundred no's, as long as I get the one yes. And that yes could be a customer, it could be an investor, it could be a recruit. You're Trying to hire like the game of being an entrepreneur is fundamentally irrational. If you just look at expected value, it's worse. You should just go work for a big company and you'll make good money. But the possibility for an outsized impact in the world and impact in you is there. And so you have to have unshakable belief in the opportunity you're going after. And that's after like the 25th or 30th. No, we just like kept going. Was like, I don't know. This problem seems like it needs to get solved and I think we have a good solution for it.

[A]

You're building into an inevitable tailwind. It was just the question of how far away is the updraft of that wind. Are you going to be able to survive a funding cycle given the market drawdown at the time?

[B]

I suppose just in the small world saying this because I just was talking with them the other day. One of our biggest competitors that wasn't a competitor when we started, but became one, was a company called opsware, started by someone you've heard of called Marc Andreessen. They were trying to do basically AWS at the time and they were trying to host everything. And I was like, people are going to want to host their own stuff. We should write software. So it's over. About a year after we got funded, Opso was like, no, no, no, we're going to sell software. So then all of a sudden, Marc Andreessen was my competitor. I've worked with Mark on a variety of things since then. But 25 years ago, we had to cross sell against Marc Andreessen, creator of the Internet and a bunch of other things, which was fun. Amazing.

[A]

You sold that company to Sun Microsystems. You were there for a little while. And then also kind of parallels to Marc Andreessen, given what he invented. You then went to Mozilla. You were then running a web browser.

[B]

Yeah, the time at sun was actually really interesting because. And it's something I brought actually to Facebook and Meta. Every part of the background had some experience. But sun in the 90s was the company you wanted to be at. I mean, this company made Java, it made the Spark chip, it made Solaris, which is an amazing operating system technological marvel that just created a ton of amazing technology. By the mid-2000s, the world had changed. They made their money selling these really expensive bespoke vertically integrated servers. They sold a 64 processor server when you couldn't get Linux to really run well on a two processor server. And then Linux x86 was just like a cheaper alternative and it just crushed them. And I was there when I watched it happen. It's like there's this change in the market and you've got this big company with a lot of prestige trying to adapt and fit. Just couldn't manage it. Well, it was a big lesson to me in disruption, how it impacts big companies. But yeah, after I sold that company I sort of, what's next? We explored starting another company. I'll tell you the story amongst friends that like I found my old co founder. We were going to go start another company. We didn't end up starting one. And part of the reason why is like our motivations were wrong. The first company was like, I think this problem needs to get solved and here's a really good idea to solve it. Second time around was like, I want to prove to everyone I'm like good, not lucky. And so I'm going to start another huge company. It's got to be a billion dollar company. So we looked through all sorts of ideas. No idea was sort of good enough. Kind of like our heart wasn't in it. We didn't like get fixated on a problem and a solution. We're in it for the mechanics. And so about six or nine months of that, a friend of mine called and said, hey, you heard of Firefox, this open source web browser? It's like, yeah, of course. It's like, oh well, the people behind it need some help scaling. Can you come help? And so he and I joined that company and I joked that was my business school because I was like, I don't know how 2,000 people on the Internet I'll never meet make a product my dad can use, come together to make one thing that is so simple the average consumer can use it. That's what we did. And we scaled Firefox, had hundreds of millions of users a couple years I was there. That was amazing.

[A]

How did you go from that to meta?

[B]

It was about three years into the journey at Mozilla we had shipped Firefox 1.5, 2.030 was like a three year long software project and there's a whole set of lessons around that. But someone introduced me to Mark and the Facebook folks originally wasn't that interested. But it was kind of one of these like, well, you know, you kind of take a meeting, see what they're like. Met Mark and said, wow, he's actually really smart. And I was like, well, why don't you meet some other folks?

[A]

2008 Facebook way pre IPO on the path to a hundred million in revenue. It was a turning into a growth stage company at that point.

[B]

Back to your F company thing. I think it was TechCrunch or someone had an article of like a cash pile on fire that's like Facebook burning cash, gonna get out of money. The rep actually wasn't super get there. It was kind of like I don't know. I actually had some friends try to convince me not to join because the reputation was look messaging services or social networks. So think email, think ICQ, think AOL, instant messenger at the time MySpace. These things are used by everyone but.

[A]

Didn'T make any money for sure.

[B]

MySpace had this very big deal with Google where that Google was going to provide ads and then there was some minimum. If the ads didn't work they had to pay the minimums cause the ads weren't working. It was like lore on the street that these businesses are going to crater because they get so used but can't be monetized. So it was a huge risk joining it. It was one of these again back to first principles, met all the people. It's like wow, the people here are really smart. They shared a bunch of data and usage of the products and the product was growing. And then you just sort of think about do the thought experiment. It's like what problem are they solving in the world? Well, trying to make it easier for people to stay in touch with friends and family. Well who needs that? Basically everyone. Eight billion people. And so our market is unlimited. We just got to execute and then there's a big risk on the business model. We got to go figure that out. That's the risk we're underwriting. I joined worst case, some good stories. I'm going to learn a lot. That was Mozilla, that was Meta trying to start that second company. I was trying to fit some extrinsic reward. I'm going to prove to people I'm awesome. That didn't work out. That's the other lesson I take forward is like internal motivation always trumps any sort of random external reward.

[A]

What was the first product you worked on at Meta when you first started?

[B]

I started in about two weeks. I had this really weird role and I was actually supposed to run product and design and like half of engineering. Like two weeks in I this pre planned trip. So I took a trip, came back after a week and then we reorgan the company and Chris Cox took over product, I took over engineering. So mostly what I did in the early years was really drive the technology behind all the products, and that was a lot of recruiting and building the teams. But at the time, most people didn't think of it as a hard technical problem. But scaling a social network was a really hard, novel thing. Friendster, very famously, which was an early social network, kept crashing. Everyone probably remembers the fail whale from the early Twitter days. These services were really hard to keep running and keep scaling. And so a lot of my energy was on that. How do we build the software and hardware systems? And we started building data centers by ourselves rather than leasing space. So we moved into the hardware space, built all sorts of crazy system software, and then it just sort of spiraled from there.

[A]

Before you joined Facebook, did you view yourself as an expert manager, for lack of a better term? I mean, I don't know how big of orgs you had managed previously, but I think by the end of your tenure you had what, 35, 40,000 people in your orbit? Honestly, I don't even know how one goes about doing that or thinking about that time type of day to day.

[B]

A lot of humility I think is the seriously is like, I think people get themselves in trouble in two ways when they're polar opposites. One is not having enough humility to just say, I don't know what that is, or I don't understand. They use an acronym. You're like, yeah, I guess I should know what that means. I don't know. I've watched people sit through a presentation and like not know what was happening and not ask any questions. And I just trained myself to be like, hey, hey, hey, I guess I should know that. Don't know. Can you explain it to me? Because otherwise you're missing things and you're.

[A]

Working with incredibly smart people all day long who are experts in their field.

[B]

The number of times I've approached a super expert in field, they're like, hey, can you explain this to me? 0% of those times did someone get mad at me? A hundred percent. They're like, cool. I get to like lecture you for half an hour. The best part of my job is like, I just get to run around and get educated by all these amazing, brilliant people. I think the other mistake people make is the over abstraction mistake. I either don't know this so I can't learn it and I don't need to get into enough detail. I've got like, people can handle it. There's times when I had to go real deep on something because we were having an issue and we had a major outage. I was on IRC to figure out what's going on. You can't just like sit at a thousand foot view and hope everything works. You've got to figure out when to dive in. And then when you're like, I've got great people, they've made good decisions. We're going to let them roll. But my first management job was at my startup in 2000, 2001, we started hiring all the engineers. I was like the super nerd. I wrote all the code and as we hired people, I was like, well, I kind of know the code the best, so I guess I'll just manage. I know it's handfuls of people, six, eight people. I know it's kind of easy mode because it was my startup and my code. And then it was really the switch to Mozilla, where I took over a small team. But then we grew that team to hundreds of people.

[A]

So it's a legit, like, engineering manager, large organization role.

[B]

I joined, I was not a browser expert. I was not an open source expert. They're like, okay. For some reason they thought I could learn how to scale these things. And so that's what I did. And then didn't try to come in there and outsmart people on the browser team. I was like, what are the things I can do to uniquely help this organization grow? So that was the beginning. And then I think Meta was another 12, 14 years of that at mega.

[A]

Scale, learning how to manage people who manage people who manage teams of hundreds of people.

[B]

Yeah, in the beginning, I was managing engineering team. When I joined Meta, it was maybe 100, 150 engineers. As you said. The team grew up to 35,000. When I started, it was one thing, it was facebook.com Then we had to go to mobile. We had to build an ads business. We built data centers, we built an AI research lab. We went into consumer hardware with Oculus Quest and the Ray ban stories, Instagram, WhatsApp, Messenger. It was like it just profligated. So it was a much more complicated business now. And at the end of my tenure as cto, I would often describe my job as a very involved board member. I'm running Instagram, I'm running the data center group. Kind of have your CEO equivalent of each of those things and a whole bunch of stuff underneath them. And I've got to be involved and helpful and figure out what to do, but I'm not in there running things day to day.

[A]

I want to shift gears a little bit, but use what you just shared as backdrop, which is we Are in this time in tech in the US where a lot is being talked about about America hasn't built hard things. We're moving back to this world of building hard, challenging things. Look at SpaceX. Look at Boom Supersonic, which just had their big moment a couple weeks ago. Look at Tesla. America starting to build hard things again. I would say that feels like incredibly selective memory. Have we forgotten that Apple, Facebook, Google, Amazon are American companies that were built over the last 20 years? And I would have to imagine you felt like you guys did some fairly hard things.

[B]

Yes. Not to make it in financial terms, but I would name the number of trillion dollar companies that aren't us. It's a very, very small number. And I would say that that whole framing of we don't know how to build stuff really irritates me. Yes. It misses space, it drives me crazy. A whole bunch of other things. You know, I know Intel's having issues. What's underpinning the AI boom is like Nvidia, distinctly American company. So I think we have built hard things and this is very relevant to climate tech, industrial policy, government policy really turns out to matter. China has had a very, very concerted effort over multiple decades to build an industrial base in a variety of different categories and has been highly successful. It is true that most batteries are made out of China. Most solar cells are made in China. Almost all the chips are made in Taiwan, which is next to China. For different reasons. There's been a huge investment there in things. And so I think this is why the response to me isn't like, it's not like we can't do it. We still have the most innovative by startups. Pick a metric no country can match the us. What we have to do is actually invest. Invest our time, our money and our talent in hard problems that solve things that we care about. Yeah.

[A]

And most of the hard things we've solved have been infrastructure to build software businesses. We as a country have built fewer hard infrastructure companies that you identify as an infrastructure company in some way, shape or form. But every company that we just talked about has an incredible amount of infrastructure behind it. It's just powering the ultimate software based business.

[B]

Got it.

[A]

And so on that note, I'm curious as you think about now, taking that notion of innovating and bringing it into the energy space, I'm starting to think of energy as almost the picks and shovels of AI. We need energy for a lot more things than AI. But when we talk about the AI boom, investing in this space is investing in the thing that makes AI happen. I'm curious how you've spent a lot of time thinking about AI and data centers and how they grow. I'm curious how you think about the convergence of energy, innovation and AI as it becomes more and more part of our daily lives.

[B]

The things I tend to think the most about are like, what are the rate limiting steps? What are the things that are slowing us down from achieving progress and things that matter if you look at AI. So I started the Facebook AI Research Lab in 2012, 2013, in that timeframe. This is right after we're going to nerd out on AI for a second. But there was this moment that was a big deal in the AI research community where a convolutional neural net won the ImageNet challenge, this image labeling challenge, image recognition challenge, eke out these small 1%, half a percent wins every year. And then like, boom, you had a convolution neural net win by like 10% of the next best. One of these rare watershed, like, whoa, wait a second. This technique works. And so part of the reason I started that lab then and Facebook started investing, and a lot of credit to Mark for really triggering this is you sort of looked at and teased it apart and said, like, okay, cool, what are the component parts of this that are going to make progress or slow down progress? And where are they trending? Say, okay, we need more data. Internet's getting bigger. That seems like a good thing to bet on. There will be more data, not less. We need more compute power. Still got a lot of Moore's law left and a lot of creativity going there. And then there's this other X factor of like, well, maybe we'll need to tinker around with the algorithms and have some other ideas. The time was convolutional neural nets. Now it's transformers. And so those changes are fewer and far between, but you kind of were hoping for those. But the tailwinds to your point, were there massive rate changes? You fast forward to now 12, 13 years later. I spent a lot of time in the early years trying to convince people that AI was going to be important. There's nothing like a product you can touch and feel. As soon as people try ChatGPT, they were like, oh, yeah, I guess this is going to be really useful. And I think we're in the beginning, not the end of it. This, the pace from here to the end. I don't think people are prepared for how much things are going to change. Even in the last couple days, there's Been some really interesting voice models, dropped the reasoning models. My concern with LLMs for a long time is the next token prediction thing was clearly not all there was. And then we've already bolted on reasoning on top, which is sort of. You like run a loop on top of it to run RL to optimize what happens and kind of treat the LLM as this lower level component which is like getting us closer. I don't think we're totally there, but there's a lot to happen there. Okay, so then you say, okay, well what's going to happen over the next five or 10 years? We're kind of running close to out of data, but we could get more, we can make more, we can do synthetic, we can do video, there's more compute to be had. But it's hard to keep scaling at the rate we're doing. The costs are starting to get somewhat insane in terms of how much people are spending to train these things. And then when you break down the cost, it basically boils down to two things. It's really like chips buildings and power chips. I think there's a lot of innovation happening in and I think there's a lot of ways to get pretty massive wins on both chips. And I'd say software efficiency buildings are pretty hard to innovate on. We've done a bunch of things, we've moved to liquid cooling. There's sort of limited headroom there. Power is going to be the big one. We can definitely make power way cheaper than it is today. It's actually getting more expensive in the short run because there's such a demand. But when you start looking at solar and you just say like, wait a second, solar is the cheapest way to make electrons. Again, back to first principles, like look at a solar cell. What is it made up? There's like so little material there, I just put it on the ground and I walk away. Once it's wired up, maybe get the dust off of it. Therein ends the maintenance of most solar panels for 20 years because you're like, huh, for a teeny bit of money I get this thing that just like makes electrons all the time. Solar is one. Things like fusion are like a total other X factor. That because it's the most energy dense reaction in the universe, you gotta be able to make some power out of that thing cheaper than burning trees. And then there's some really out there ideas. So we can talk in specific, but there's a company I'm really passionate about that's basically building an Offshore floating data center. And the interesting thing about it is when you look at how they're building it, it's an incredibly simple machine. It really kind of devolves down to the cost of steel to make it. And then it makes its own power and does its own cooling. Actually innovates on both the building and the power side at once. And it is probably cheaper by 50% than any other way to make compute out there. I think there's a lot to do here and sort of how to innovate all these things we can dive into. Anything you want to talk about.

[A]

All of these are supporting this massive growth in compute and energy consumption that we're forecasting. I'd say the picks and shovels of AI. And if folks are listening, don't know what that means. You think back to the Gold Rush era. The people that made money, the people who became millionaires from the Gold Rush mostly weren't gold explorers. It was people who sold picks and shovels. It was Levi Strauss who made blue jeans and it was Wells Fargo. Right. Who financed the whole thing. Levi Strauss and Wells Fargo are like the lasting companies that came out of that Gold Rush boom. I'm curious right now how energy evolves to build the next enterprises that grow large and support this AI boom. And I don't think it's going to be the traditional power companies. I don't think it's going to be oil and gas majors and the like. It is likely to be a next generation of power producers who can scale at the level that's needed and required. Which means you have to be relatively limitless, which, AKA renewable.

[B]

Yeah. And I think if you just look at the US AI is such a big thing and it's growing. So it's a really fun thing to talk about. But it is one of three megatrends happening in parallel that is driving, driving up electricity demand in the United States. The other thing is just like a massive investment in manufacturing. So there's a onshoring of manufacturing to the United States. And then the third is electrification of everything. EVs have been on a steady growth path to date. You know, we're replacing gas stoves and induction stoves. Like we're electrifying a lot of stuff. And so you add these three things together and you get a 5x utility demand for electricity by 2050ish is what most estimates say. That's a pretty crazy number. Again, I think it opens up a ton of opportunities to make power differently and to start laying down infrastructure that basically Makes unlimited clean power with very little work or input. I need to like go back and do these calculations, but one of our companies, Exmoor Energy, says that if they build a large scale gigawatt power plant, you could fuel it with a pickup truck for a year. One pickup truck's worth of fuel for delivery would fuel a gigawatt for a year. The equivalent for a coal power plant would be like, I don't know, multi mile long train of coal cars. Just intuitively you're just like wait a second. If I can unlock a power source like that. Literally every sci fi book you read. Interstellar travel powered by fusion on board and we know we can do it. Humans have made fusion happen on earth multiple times. So it's an engineering and commercial problem.

[A]

At this point it strikes me that you're referencing history before and when you're talking about how compute evolved. You went from these big 1950s mainframes to were talking about your first business was we're going to bet there's going to be lots of servers. You end up with this commodity server layer that powers our data world today. Today power is mostly large monolithic power plants with solar and wind we're moving to smaller, more distributed systems. Do you see that replication happening even in other alternative forms of clean energy in the future? Or do you think with fusion you got to go back to like a large scale, like big plant?

[B]

Yeah, Fusion is the one that breaks my rule on this because I think it's a special thing. But I do think the lesson if you walk through a multi hundred thousand square foot data center, it is packed with pizza boxes. The pizza box is 1U, it's 1.75 inches tall. Maybe you have a 2U or 4U box which is 2 to 4X that size, the units there. It isn't a rack size machine, it isn't a container truck, 40 foot wide long thing. It's a relatively small thing. And you have to ask yourself like why did that happen? And it's like well because you're going to sell tens of millions of those a year. And when you sell tens of millions of something, you can get really good making them incredibly cheaply. That's when like all of a sudden robotics starts to make sense. You like replicate processes and this is back to anything where the core component and I think the best example of this is chips. If you just think about the amount of engineering R and D that goes into the design of the chip itself and then all of the manufacturing, there's a very big machine set of machines to make a chip. But that net result is something I can hold in my hand. And we're going to make again millions, tens of millions, hundreds of millions of these things. And so we get very, very good at it. When that component looks like that, then you get to get on what I like, which is a cost down curve chips about a million X the compute power per dollar from when I was in school to now. Solar is on about a 10% a year cost down curve. Lithium ion batteries. Why is my car filled with AA battery size esque battery cells rather than big ass ones? Because you can make a lot of them really cheap, 10% a year. The way I like to talk about entrepreneurs is like you can go home, take a year off, come back, your inputs are 10% cheaper, you got more competitive without doing any work at all. That's a massive tailwind. We're seeing that in solar, we're seeing it in batteries for sure. We're seeing it in other technologies too. Anything where you can get mass manufacturing. So electrochemical cells is one that I'm really passionate about. A company that's building a ethylene plant and the core component is a meter.

[A]

Square dock cycle, right?

[B]

Yep. Doc cycle.

[A]

They've been on the show. So listeners, go listen to the archive.

[B]

Go listen to that one. Their commercial unit, they're going to make one, they're going to turn it on, be like look, this thing works. I mean like this pitch to like commercializing is like I just make more of this thing. There's no like giant scale up or other thing. I just like manufacture a lot of them. Anytime you can do that, you get learning rates, you get cost downs, the whole flywheel goes in your favor. And so I think there's a ton of technologies that fit into this category where the core innovation is actually something that can be replicated with frequency.

[A]

I love that lesson of how you've seen scale come down the cost curve and how technology tends to modularize over time and why. A great lesson for folks in our space who are building or thinking about the future. Going back to the AI question. As we think about the future, another thing that I've heard big thinkers in our world say recently is five plus years from now, the whole notion of enterprise SaaS and software business models are going to be eaten by AI agents. Anything that is right now kind of a vertically specialized piece of software will not require being its own coded thing that a generalist AI platform will be able to be smart enough to go solve that problem and as long as it can get connected to a customer and connected to data, you can command it to do the things you need it to do. A do you believe that? And I think Satya Nadella at Microsoft was the one going on that talking point a little while ago and I heard Avinode Khosla say the same thing last week at Upfront Summit here in la. Do you believe that's where we're headed from a software perspective? And if so, how does that impact how you think about software based opportunities?

[B]

I think there's elements to truth to it, but I think it's an oversimplified story. And here's what I mean by that. I think two things happen at once that fight against each other. One is in software you have a constant commoditization of the layer above you from the layer below. In the Windows days there used to be all these companies that would make applications. Norton Utilities, I don't know if that's something that anyone remembers, like a bunch of extra utilities for my Windows device. Microsoft just went like, why don't we just put that in the operating system? Why are you buying a separate this and that calculator and an antivirus, you're just like going to bundle it in. And so for anyone building on top of OpenAI or anthropic at some point they're going to go, wait, the guys making coding agents are making a ton of money. Maybe I should go into that business and go ahead and get some of that myself. And so you've got the platform providers constantly trying to move one up and sort of eat into the businesses above.

[A]

Look at your iPhone and the apps that didn't used to be native that were Apple apps and now how many native Apple apps they have. Who would have thought Apple was a weather company 20 years ago?

[B]

100% tough to differentiate because they're going to take any margin they can from you. On the other hand, there's always this challenge in these broad platform businesses of they're doing a lot of things and so it's hard for them to specialize. If I'm building a medical advisory AI system that needs to respond to HIPAA and needs to like be calibrated properly so it doesn't give you advice that's dangerous for a patient. There's a bunch of very specific things, data sets and other things that you'd have to do and integrate. OpenAI's probably not going to go do. I think there's actually still a really interesting opportunity there. And so I think there's vertical opportunities where like I found this market where customer access, data access and all the rest of it are really, really important. I'll give away my secret of how I test for interesting AI companies. It's like if you, in your first couple minutes of pitch, haven't explained to me how you've got a proprietary data flywheel going, I'm out. And data flywheel means that the more people use your product, the more the more that usage creates data that is unique to you that then builds a better product and then you're off to the races that you were required to do. Otherwise anyone can immediately just step in and take your business from you. Hey everyone, I'm Yin, a partner at mcj, here to take a quick minute to tell you about the MCJ Collective membership. Globally, startups are rewriting industries to be cleaner, more profitable, and more secure. And at MCJ we recognize that a rapidly changing business landscape requires a workforce that can adapt. MCG Collective is a vetted member network for tech and industry leaders who are building, working for or advising on solutions that can address the transition of energy and industry. MCJ Collective connects members with one another with MCJ's portfolio and our broader network. We do this through a powerful member hub, timely introductions, curated events, and a unique talent matchmaking system, and opportunities to learn from peers and podcast guests. We started in 2019 and have grown to thousands of members globally. If you want to learn more, head over to MCJ VC and click the membership tab at the top. Thanks and enjoy the rest of the show.

[A]

The other trend around AI and energy that I'm curious to hear your take on is 10 or so years ago. More than that, 10, 15 years ago, every company had to think about how to become cloud literate, how to use cloud in their business and mobile too. How are these things part of what I do every day? We're starting to see signs now at least big tech companies are having to be thinking about energy procurement in a major way. They are having to fund energy procurement teams. How they buy power, how they access power, how it's fundamental to their company. Do you see that extending out to the broader like Fortune 1000 or whatnot, where every major company has to become to some extent not an energy company, but needs to at least have deep energy expertise in house?

[B]

I think that's a really interesting question. I actually would like answer it, but with a twist which is I think we're already seeing this on the consumer side and unregulated markets. When you look at Texas in particular where you can like choose your power provider. I don't have to use whoever's there. And then it opens up a whole bunch of new business models. Am I on vacation for six months a year? Do I want a battery? Do I have an ev? What are all the things that are different about me than everyone else? This sort of one size fits all power system, everyone signs up, you get power 24, 7 and nobody has needs that are different than anyone else, is obviously inefficient and wrong and doesn't allow for innovation. So I think whether it be companies deciding to source and build their own power, whether it be consumers deciding that they're going to effectively create their own power by having a battery or having solar and sort of using that, I think it's going to put a lot of innovative pressure into the system which is what we really need to sort of align needs to cost and usage because that's when you then expose all the efficiencies to then go optimize. So I think it's going to be interesting. I don't know that every company is going to do it, but I think that in energy intensive businesses is really going to matter. And you're seeing hyperscalers pay for future plants, for fission, for geothermal others. I think big industrials are going to do the same. I think it is going to be a competitive advantage over time. People aren't paying attention to. It's not just like power is a thing, it's like the way we make power and distribute it is fundamentally changed. It's both cheaper but way more volatile. And the second part is the thing that I think is lost on most people and that just really changes the game in terms of you've got a different set of knobs to turn on being really cost competitive if you can be creative on using that volatility.

[A]

Shifting gears a little bit, Mike, for you personally, I'm sure you had to think a bit about power and how it powered your data centers and whatnot at Meta, what caused you to decide I want to focus all in on this area. Was it starting from an interest in power markets and how it's changing and clean energy coming on or did it start from a climate change concern and then it brought you into thinking about clean energy? What was the Mike Schrepfer path?

[B]

At my heart I'm a nerd. So I love technology because technology shows up to zero sum problems and removes the constraints because you're like, oh, I just made this thing twice as efficient. So many problems are Can I do this or can I do that? Well, if we make it half as expensive, you can do both. And there's just nothing else in the world like that. That is the core of everything I do. I had the first Nissan Leaf I could buy, which is an electric car, which had awful range, but it was fun to drive around. So I was kind of nerdy about this stuff, but honestly, it wasn't top of mind. I'd seen Inconvenient Truth and other things like that, and basically what happened was Covid. Covid happened, Lockdowns happened. My life got thrown on zoom all the time. Kids not going to any activities had a little bit extra free time. I was kind of watching humanity's response to a crisis, and it just reinforced a view that I have, which is like, acute crises where it's right in front of our face. I think we're pretty decent at figuring out how to rally around it. The, like, 10 or 20 years out thing we kind of suck at. And climate change is like the mother of all. It's happening. It's just like a slow burn. It's real hard to get people motivated. But there's a bunch of all the stuff we talked about. Are we going to make fusion work? We need to be working that for 10, 15 years. You can't show up and be like, can I have it tomorrow? And that's the technological trends. Like, can we invest in the things so that they're ready for us when we really need them? I thought it was originally Cody just going to be like a side project for me. Do this nights and weekends. Hired Bridgepan to help get me smart on climate, and said, like, I want to start doing climate philanthropy I can do on the weekends and Wednesday evenings. And then I started doing that. So I've done a bunch of climate philanthropy then and continued to do so. But as I learned more, I was just like, wait a Second, it's a $10 trillion problem. You don't solve $10 trillion problems with government or philanthropy. You need markets. The only reason markets care is that they can make money. And then coincidentally started bumping into people like you and entrepreneurs and others are like, hey, I have this really good idea. And it like, turns out this way of doing this is cheaper than the current way. Oh, and by the way, it's lower carbon. So the more I found that, I was just like, huh. So I started kind of again, nights and weekends, angel investing. But then it just sort of snowballed and I said, look, I think this is a place where investment in the right technologies with the right entrepreneurs that have commercial potential is a way to align goals. We can have a better future and more prosperity for everyone. So. And it sort of happened from there and I eventually decided to focus more of my energy on it.

[A]

Were you already planning to leave Meta before you decided this was your thing or did this thing pull you out of Meta?

[B]

Actually, still. I have a role at Meta called Senior Fellow, which is, I'm there as an advisor to Mark and others. I love what the company's doing and I love the teams there. And the work we're doing in AI is really critical, particularly open source. And so this was a like, can I have my cake and eat it too? Mark is amazing. I was like, look, I want to go spend a bunch more time on this stuff. I don't really want to say goodbye. Can we figure out how to do this? And luckily I had someone who'd been working with me for a long time or had mentored to take over as CTO and I moved into this other advisory role. And so over the last three years I've been continuing to advise Meta on AI and others while I've been doing the climate work. It's sort of a, I kind of wanted to do both, honestly, and it got very lucky that I'm able to do it.

[A]

Well, you and I got to know each other as you were right in the middle of that transition, I think, and we, we got to plan a fun event together in San Francisco that brought together a bunch of top technologists who were thinking about moving their careers into this space and going through the same journey you and I were in the middle of. I'm curious how you have seen talent interest continue to grow in this space. I mean, you managed 35,000 engineers. You have lots of relationships with lots of very smart people who, many of whom are starting to put their hands up and say, this is an area I want to work in. There was a huge wave of it. Like you said in the start of COVID How have you seen the interest ebb and flow and grow and shrink and where do you see people wanting to jump in today?

[B]

That was a fun event, by the way.

[A]

It was so fun.

[B]

Think about the speaker list we had John Dorr was Tom Steyer, it was Matt Rogers. It was just like an incredible group of people all came for free to talk at the Commonwealth Club. I don't know if you know this, Cody, but Selena Tabakiwala is a multi time entrepreneur, started evite in the 90s, Glaxo was at Survey Monkey. She started a company called HomeBoost, which is a company to help consumers basically figure out how to make their homes more comfortable and efficient. She started that because she came to that conference, figured out she could take her product design shops, consumer product experience and apply it into the climate space. And so to me, that's a massive. I'm going to put a W on the board on that one.

[A]

Amazing.

[B]

I think what we saw was a massive wave in 2020-2022. Cool to do it. A lot of people wanted to do it and so there's a lot of interest. I think it's waned since then, just to be frank. I think that the realities of life, it's hard to build climate tech companies. It isn't the fastest path to a massive valuation in a short period of time. Plus, the boom in AI has pulled a bunch of people back over to AI. I still meet people either in their late career or people, especially students, everything from Undergrads, Master students, MBAs on a weekly basis who just decided this is where I want to work. Because they're like, look, I want to align, I want my work to affect the future. And I think this is a lot of our job. Cody is back to the zero trade offs. I like to make things zero trade offs. I want to make more companies out there that are great careers, a place where someone can go build a family, buy a house, do all the things they want to do with their lives. And by the way, all the work they do is aligned towards improving the way the climate and the way we live in the world. That is the mission that sort of really got me to do this. I think we can do that and we are doing it. And that is going to just unlock the next wave of talent.

[A]

Tell me more about what you're doing now at gigascale.

[B]

It's a lot of what we've been talking about. It starts with a simple premise, which is we believe there are businesses out there that fundamentally are great businesses, meaning they build a product that has massive pull, customer wants it or is just cheaper. That's the sales pitch. We're doing this, it's cheaper. And the reason it is is because it's actually more efficient in some form. And so it happens to align with, we emit less carbon, we use less power, and that's why we're cheaper or better or whatever it may be. And so those two things align, which is commercial interest and climate impact. And we look for the best entrepreneurs we can find in Those fields. We've assembled a team of operators. So I've started companies, I've run companies. Victoria who joined with me, she was in the solar industry. In the 2000s she started a company, she'd been invested for 10 years. Evelyn joined us as well. She had a YC backed biotech company, also deep tech investing. And then Evelyn is a chemical engineer with another chemical engineer, mechanical engineer. I'm a super nerd. So we try to really take deep technical expertise and commercial viability and put those two things together and say, hey, where are the places where there is room to build trillion dollar companies that will improve the way we live in the world? That's what we're doing. And then the thing I think that makes us unique is we are able to take all that operational experience once we pick a company and back them, we're able to provide them a lot of help and it's back to that. It could be as simple as like, man, I remember fundraising and that sucks. Or I remember having to rename my company. We raised a series B and I had to do a layoff right after. Let me talk you through what that's like. And just trying to skip people over what I call the obvious mistakes. There's a lot of parts of building a company, they're kind of the same. No matter what your company is, how you do performance reviews, how you hire plus or minus 10%, you're not going to innovate. You want to spend all your energy innovating on your go to market your consumer and your technology. And so we try to help them fix all those problems so they can focus on the thing that's really important and that's what we do. It's a lot of fun.

[A]

Pre seed seed Series A, general early stage is your wheelhouse.

[B]

Yeah. So I'll give you a where we are now and how we got here. So where we are now is quite simple. We're pre seed to series A. That's sort of our wheelhouse. I think we've helped companies like Selena get started, Pasture Biosciences and another company that's working on a MRNA vaccine to help cattle reduce methane emissions. It's a company we help get started. So it's everything from get the company started to series A. That's our sweet spot. Because I came out this fresh and didn't have any portfolio or anything. I was able to do what I love to do, which is high frequency experiments. And so we have done a bunch of later stage investing because people want our expertise and so form Energy, Commonwealth, Fusion, Systems, Mill. We're all Series C+ companies that we've invested in. I've done a bunch of other angel investments, so we've got actually bigger portfolio of companies that we support. After doing that, realize that our sweet spot is really pre seed to Series A.

[A]

Of all the ways you could come at this problem, you've got resources from your time and success at Meta. Why is Early Stage Venture the channel that you decided to go down?

[B]

This was more of a journey than a. I nailed it on the first try. I kind of like had a little bit of a wander through the woods again. I started with philanthropy. I thought maybe that was it and I was like, okay, well maybe I'll just take a bunch of the money or investing and move it over to climate funds. Okay, let me try. Okay, let me do some angel investing, see what that's like. Again, a uncharitable version is like a random walk in the woods. A charitable version is I frequently experiments to find your optimal point. The advice that I give to people is like, you're trying to find this perfect alignment of impact you want in the world and what are the unique skill sets you bring to bear to that problem? And there's some optimal intersection of those two. And it just turns out that ability to believe in technological trends, understand things from a first principles to pick entrepreneurs, part of their being 14 years as an executive at Meta, it's like I watched people go from just out of school to like VPs and C level executives and I watched a lot of people not succeed. So I have a pretty good pattern matcher for like who's going to fail and who's not. That aligns better with pre seed to Series A from a skill set perspective is just where our sweet spot is. Then you have to ask yourself the question of, okay, from an impact perspective. I think there's this point of view of like, we've got all the companies out there we need. That's just wrong. Entrepreneurs every day who come up with crazy ideas like, whoa, that's a really good idea.

[A]

If the Wildfire problem was solved, it would be solved. It's not solved as an example.

[B]

And the good news is, Cody, like, I think the problem with climate is it's really easy to get overwhelmed very quickly. Just, oh my God, it's such a big problem. Oh my God, there's ag, there's buildings, there's power. And then you just realize like, look, if you actually look at history, it's a small number of people and a Small number of companies that fundamentally transform how things happen. Back in the early days it was Henry Ford. What would electric, what would EVs look like without Tesla? What would the GPU market look like without Nvidia? Where's the Jensen of basically power of ag of buildings? And so we're just looking for those people and one of them in a major sector can be completely transformative. And this goes back to my son experience in a big company. The best way to get a big company to innovate is to kick their butt with a startup. If you're frustrated with how a company isn't being carbon climate sensitive, certainly lobby do all that stuff. My plan is I'm going to kick their butt in the markets with a startup and then at some point they're going to be forced to reckon with the fact that their competitor is fundamentally cost advantage to their old school way of doing things and then they're going to have to change or die. That's my point.

[A]

You certainly lived that on the inside at Facebook as it grew up and completely transformed multiple industries around it from advertising to media and publishing to video games, down multiple pathways there as you have invested, are you starting to get more and more thematic in your approach? One thing at MCJ we've got a sizable portfolio now of 100 or so companies that I've started to see in our world are what I've been starting to call families of innovation. Meaning you invest in one thing and then you see how it's actually has a need in the market for this other thing over here or its customers are looking for this other thing over here and you see another startup that's working in this area and you invest in that. And then it unlocks a third synergy that you find another startup and they're all collaborating together to go after a space. And I'm curious, I know we've seen it in EVs where there are these families of innovation that navigate between charging and access to charge points and this, that and the other. Curious how you've approached that so far.

[B]

My favorite way of operating that we don't always get to do is we just go deep in one area for some calendar period of time. Pick geothermal. We're going to like meet every company in the geothermal space, read everything we possibly can and come to a point of view. Because I think in a lot of these things if you really again, first principles, tailwinds, there are better and worse approaches to physical things. You can say maybe this won't work or there's high technical risk. You can usually winnow down the set of possible solutions to something, and then once you do that, you now have a strike zone of, okay, this set of things is really interesting. And then, wow, we just met this entrepreneur who's incredible, who's working on this particular thing. Bullseye. We can run really fast. That's probably my favorite. But sometimes it gets you down a path you didn't discover. We mentioned Doc Cycle before. You've had them on the show. We kind of got there. Not because I was trying to figure out ethylene. We were basically looking at what I'd.

[A]

Call power to X. Sarah Lamiasson, that's the CEO, right?

[B]

Yeah, Sarah. We were basically saying we spent a bunch of time looking at sustainable aviation fuel on E jet and saying, like, okay, if power's going to get cheaper, can you use that to produce jet fuel? And the more we started looking at it, we kind of came to a funny conclusion, which is like, if I had cheaper electrons in like a box to make things, I wouldn't make jet fuel. There's a bunch of other things I'd make first as we were doing that, because they're easier to commercialize, they're higher margin, a bunch of other things. We met Sarah and it was like, she's doing exactly this and she's incredible. It was a very fast process to get to conviction, and she and that company are just crushing it.

[A]

You've invested quite a bit in storytelling. I mean, that's the thing we know a little bit about here at mcj. You brought on my dear friends Kristen and Meryl, who are running an agency called the Regenerates. They're now part of the Gigascale team. How does storytelling and helping founders that you've backed get their message out fit into your strategy? And why did you choose to make that a fairly core part of what you do?

[B]

I love that you asked this question, Cody, and it's one of the things that I'm inspired by the work you do in this program. Humans convey knowledge through stories. This is how we learn things, as nerdy and data driven as we are. And so many of these things are ideas and a story until they're not. And the way you're going to get customers and investors and employees or future employees to join is by having a very clear story. Here's the change we're going to see in the world if we succeed. This is what happens, and this is why it's, like, worth coming in on Saturday or staying a Little bit late because it really does matter. And my experience in running a lot of high performance teams is that's actually what really gets teams to execute. When it's clear that your effort is going to matter and it's going to matter to something you care about, that's when I'm at my happiest. My happiest moments, Cody, have been like working massive hours every week and being exhausted at the end of it because I'm like, this really matters. And my least happy is like it doesn't matter what I do. The results are going to be the same. That is a depressing state of the world for humans to be in. And so I think part of what we're trying to do is these amazing ideas that groundbreaking can change the world. How do we tell that story in a way that everyone can understand so they can join up as an investor, as a new employee, as a partner, as a customer. And it's just so critical to get that really, really right. And back to talent. Shout out to Kristen Amel, like we were lucky enough to bump into these folks and work with them now for a while. And the difference in skill set between them and the average person doing this is just incredible. In writing, in marketing and positioning and that really matters. And this is the other back to humble moment. This is something I counsel. We work with a lot of super deep tech founders, PhDs, scientists, plasma physicists. I'm a super nerd. I spend a lot of time thinking the nerds ruled everything. And it's really easy to dismiss all the other professions and basically be like, oh, finance people, deal, whatever, anyone can do it. No, every single position in your company, there's a 10x version of that. There's a 10x engineer, there's a 10x deal person, there's a 10x marketer. If you don't understand that, you're going to be in trouble. Because if you can hire those people, that can be the difference in like, wow, I've heard about your company. It's amazing. I'm in and like wait, what do you do? Chemical, what's ethylene? I don't know. Sounds complicated. Those two states of the world couldn't be incredibly more different. So anyway, we're lucky enough we have a 10x team here and we apply to our startups.

[A]

You also just looking at your website and I haven't met these folks, but it looks like you've also been investing in talent in particular in terms of trying to bring resource to your startups where you can Help them get the best hires inside their companies. Maybe share a little bit about your talent strategy at gigascale too.

[B]

Cody this is the fun part of I didn't start out like, I want to start a vendor firm. I was like, I want to get 20, 30, $50 trillion climate tech companies out of orbit. Where are the levers that I could jump on to make that happen? And it turns out, like, finding those companies and backing them didn't have a like, well, venture firms have X, Y and Z, so I must do X, Y and Z. We just gathered a lot of data, talked a lot of entrepreneurs kind of figured it out. And then two, two things right away was like what we just talked about, which is all these companies need to fundraise, get customers, get employees. That is messaging and marketing. And so we're going to do a plus 10x service for that other thing. All these companies need is talent. They all need to hire people. And so lucky enough, Adam is someone I happen to work with at Meta and his co founder Janet is amazing and they have been working with us to help the talent strategy. And this isn't just, hey, I need this wreck. A lot of it is, what's your talent strat? The number of times someone comes and says, we need you to do a hire for this. And we're like, no, no, no, no, you don't actually need that. You need this other thing instead. That's so much higher value than, okay, cool, I'll run your process and go to LinkedIn and figure it all out. Is strategic advice. And that's same with the messaging and marketing side. And my M.O. cody, is really simple. It's like, I'll tell an entrepreneur, I'm only gonna send someone to you in a topic area with one criteria is if they and I disagree on a topic, take their advice. That's how good they are. And that's part of the value of working for 25 years. It's like I've just seen so many people, I've worked with so many people that we can like just say, no. This person really knows what they're doing. They're going to save you a lot of time. And so we do that in marketing, we do that in talent. We've done this in the people side. I've worked with people, I've worked on the people side to help with tough career growth and other issues. That is the secret sauce of us is we're just going to really accelerate people's scale out of orbit.

[A]

Mike we started the conversation reflecting on your time coming to work right in the guts of the dot com crash in 2000. And we talked about how clearly the vision of that time was correct. The Internet was going to take over commerce and communication and media and how we live our daily lives. We just got overhyped and then it fell on its face and then it took some time to recover and get to where we were. And you timed your second wave into that world perfectly. You were at meta during the absolute growth phase of the consumer Internet with this whole world of energy tech, climate tech, whatever. We've been through the boom and bust cycle of Clean Tech 1.0. We've been through the boom and bust cycle, frankly, of fracking and the oil and gas version of that same thing. Where do you think we are right now? Coming out of the trough of disillusionments and into the slope of adoption, the slope of enlightenment, all that stuff.

[B]

I think we definitely had a hype bubble a little bit in the last couple of years. I think we've come down that hype bubble. I don't know that we've hit bottom yet in terms of shaking these things out. And so I think that like everything's going to happen tomorrow is not true. This is what happens.com thing. It's like everyone's going to be on the Internet tomorrow. It took a little longer than we thought. Still faster than anything else. If you look at naval art has a great set of slides out there. And my favorite slide in his deck is fastest electrons to 100 terawatts. If you like graph solar growth versus every other form of power generation. How fast do they get to 100 terawatts? Solar just crushes everything. The reality is the difference between now and Cleantech 1.0 is solar is 10 times cheaper per cell and getting cheaper still. Lithium ion batteries are 90% cheaper than when they were first introduced. At 99% in the market, they're still going down in price. And so again, my business is premised on batteries are going to get cheaper and so they're going to be used for more stuff. I think we're going to be okay.

[A]

Do you draw a parallel between that and like the intel chip Moore's law curve in terms of what it enabled from a compute perspective?

[B]

Yeah, similar as like if you were counting on compute power increasing or dollar per watt, dollar per compute getting better, then yes, you were sort of banking on a relatively safe trend there. And so like again, I think it's tailwinds in a market you want to be investing in tailwind. So it's like we can screw up a bunch of stuff, but we still got momentum in our favor. Never. Everything's going to go right. One of the things that puzzles me is the debate about electric vehicles, because the math is just so overwhelming here. Already in China, 66% of the cars in their category, the cheapest version of that car is electric. That's off the lot. As soon as you drive it off the lot, if you're cheaper, you're saving money the day you drive it home. And then battery prices continue to go down 10%. And the battery pack is the most expensive part of the ev. Just like, wait a second, we're like at parity or cheaper and the most expensive parts getting cheap. Gas engines are not getting cheaper. 10% a year. They're not. So we may have a bad year. Maybe there's some change in policy or people get frustrated or whatever it is. And so maybe 26 is a bad year. But the long term trend, which is what really matters, is sort of overwhelming category. And we can just make stuff faster. I mean, we haven't even talked about like AI for design. We haven't talked about 3D manufacturing. You just go talk to. People have been building things for 20, 30 years. They're like, yeah, 30 years ago, we'd have to hope we got it right. Now I make six of them on a 3D printer. I run them all through a test and be like, all right, that one works, let's take that one. Or I model it in software, which is even better. It's not the same. It's like, why Instacart works now versus Webman didn't. In 2000, everyone had a fricking smartphone and so the density of deliveries made sense. That was it. That's the difference. Otherwise, it's the same business.

[A]

Same business. This has been great. There's more we could talk about. We haven't even talked about all the work you do on the philanthropic side in climate with additional ventures and all of that. If there are a handful of more things you think we should jump into, let's jump into it. Otherwise, I feel like I've learned an absolute ton. And I really appreciate you taking the time to come on here.

[B]

I think we've covered a lot. Maybe a word of thanks and a word of encouragement. Thanks to you, the MCJ team, for all you've done. I think one of the biggest surprise pleasures of working in the climate space is the amount of collaboration people have had. Even though we might be in competitive environments because everyone sort of cares about the outcome. And that just like makes life more fun because we're sort of on a similar journey. And then for anyone listening, the biggest lesson for me is climate can be overwhelming because it's huge. The good news there is that you can flip that around and say like, wait a second, there's a lot of different things that can have impact and there's a lot of different skill sets that are needed. So if you're listening to this and you're like, wait, I'm like a biz dev person or I'm really good at financing, there is a role for you. We need you badly. All the engineers too. We need all of you. We need all of it. And just stacking these companies with talent is one way to help us make sure that we can get these next set of trillion dollar companies, which I firmly believe will be on the energy transition. I mean, I'll just end with this Cody, which is sure you've seen this data yourself. The inputs into what I can estimate is about $12 trillion worth of industry, a $2 trillion power industry, $4 trillion chemicals industry, a trillion dollar steel industry and buildings is another trillion dollar. Like these things are all getting massively disrupted via climate demands, be it volatile energy prices. That's bad news in one way. From a startup perspective, that's fantastic news disruption equals opportunity for startups. So I hope if someone listens to this and gets inspired to either get into climate or stay in climate, that would be worth our time.

[A]

Well, that is a good sentiment on which to end. Srep, thanks for joining us.

[B]

Thank you.

[A]

Inevitable is an MCJ podcast. At mcj, we back founders driving the transition of energy and industry and solving the inevitable impacts of climate change. If you'd like to learn more about mcj, visit us at MCJ VC and subscribe to our weekly newsletter at Newsletter MCJ vc thanks and see you next episode.