A Conversation with Frazer Anderson

Frazer Anderson is a Managing Director at Link Ventures. The following is a transcript of our recent conversation, recorded on January 21, 2026, which has been lightly edited for clarity.

Davos and GPU Diplomacy

FA: [I’m] live here from Davos with Alex Wissner-Gross. Alex, you are very much a digital native. You live very much online, you spend a lot of your time building software, creating agentic systems. What has gotten you out of your digital lair in Massachusetts to come to Davos? Why is it so important that you’ve bestirred yourself to be here?

AWG: Good excuse to step outside my normal comfort zone and spend a lot of time in Paris and Germany and Austria and Liechtenstein. So really I’m just here for the change of scenery. This is all icing.

FA: Tremendous. Have you heard anything new here? Have you heard anything that was particularly interesting to you? Have you heard anything that you thought was completely ludicrous?

AWG: I had a wonderful conversation earlier—and I don’t think this got recorded—with US Under Secretary of State Sarah Rogers, who I think is doing a brilliant job of thinking about “GPU diplomacy“ in an era when compute is very much not evenly distributed around the Earth’s surface. I would love to have many more conversations like that here.

The Dyson Swarm and Realpolitik

FA: Okay. Can you double click on that a little bit? I think everyone is aware of the rise of sort of realpolitik as a way of doing business, I’d say probably over the last, oh I don’t know, 18 months or so. But I think while we all see the economy changing in a very dramatic way, there’s a lot more to come. And the things that can actually be moved places—GPUs, the sort of physical layer of intelligence—are going to have much more influence over the way countries are built, the rate at which they develop, their ability to maintain strategic trade advantages against each other. How are you thinking about the world? How are you thinking about some solutions to ease the pain from the Singularity?

AWG: Sure. Well, let me get the clichés out of the way first. Cliché number one: GPUs are the new oil, or GPUs are to the 21st century what oil was to the 20th century. Cliché number two: The world is dividing up into zones or spheres of influence, each with its own independent tech stacks. Clichés are now out of the way. Let me try to say something a little bit more interesting. I think the Dyson Swarm is going to happen. I think at this point, barring some wild sci-fi left turn in civilization, there’s so much momentum at this point for the potential upside to building data centers in space and diverting this explosion of intelligence infrastructure that we’re building up to space. I think it’s all but a foregone conclusion that the realpolitik that one might speak of actually becomes “realspacepolitik“ and not land-based. And that adds literally a whole new dimension to it. A classic example: a friend of mine runs one of the first lunar data center companies. And realpolitik on the Moon looks very different than on the Earth. For example, you have to contend with problems—also low Earth orbit problems—like, well, there’s really not quite the same defense perimeter that one can establish on the Moon. It’s very much a Wild West. If you’ve seen the Brad Pitt movie about this, Ad Astra, I think it depicts in beautiful cinematic detail what happens when commercialization—albeit not AI-oriented commercialization—hits the Moon. It’s a Wild West. Like, what does that look like? That’s what I think we’re racing toward.

AI Compute Futures

FA: Okay. And if we stay macro, you know, I think everyone has seen the sort of rate at which Nvidia’s market cap has grown enormously. I think there’s a general intuition that anytime there’s a supply squeeze, it is inevitably followed by a glut. There is absolutely no evidence that that is happening yet with Nvidia’s GPUs. Google’s making immense progress. What do you think we’re going to see in the GPU market over the next year? Or does it even matter because the Dyson Swarm is happening at such a fast rate that the individual players are just going to be absorbed by this aggregate demand?

AWG: I’m glad you asked me for a crystal ball. So as it happens, I have a financial interest in a company named Ornn that specializes in being a crystal ball for GPU and more broadly compute futures. So my meta non-answer is: ask Ornn. Or pay Ornn to establish a futures contract. Like, build a market around this and let the masses—who have far more insider info than I hopefully ever will—build some future contracts and some options around establishing what the future of compute is. That said, I’m obviously a long-term bull on the intelligence explosion and I want intelligence too cheap to meter. Whether that’s ultimately hyper-inflationary or hyper-deflationary, or ultimately very good or bad for a single player like Nvidia, I’m completely agnostic. I just want to tile the Solar System with compute at this point.

What New Startups Should be Building

FA: You’re the best person I know at calls to greatness. Any calls for startups? Any calls for businesses? Any calls for problems to be solved before we end this interview?

AWG: Very kind. I would say: assume the window to do transformative things is on the one hand closing, and on the other hand has never been wider. So an example of the sort of [new] company that I’m now not very enthusiastic about is SaaS for “fill in the blank,” for a variety of reasons. The example of the sorts of companies that I’m now very enthusiastic about, that I think can become multi-trillion dollar companies, are companies like: using AI to redesign coastlines and create new islands; AI for solving physics; and AI for changing the way physical service economy-based labor categories work. I think these are all prime at this point for innovation, for entrepreneurship. A rule of thumb is: if you saw it in a science fiction novel and it hasn’t been built yet, now is the time to build it. So, call to action, paraphrasing “It’s Time to Build.” I would say it’s time to build the sci-fi future because we’re catching up with the sci-fi future very rapidly. I think we have like 10 years or so before we hit Star Trek level future, max. So mine sci-fi wisdom, mine sci-fi for future ideas, and go build them now.

Favorite Science Fiction

FA: Three sci-fi novels to read to mine for ideas?

AWG: That’s a softball. Okay, so my three favorites—glad that you asked—are: Accelerando by Charlie Stross, Rainbows End by Vernor Vinge, and of course Diaspora by Greg Egan. Which respectively depict a single family going through the Singularity, the pre-Singularity times as seen from a family’s perspective, and then post-Singularity times as seen by post-families. So I think read those three books if you haven’t already. This will give you a pretty good sense of the pre-, trans-, and post-Singularity, and this is all happening over the next few years so you won’t have to wait that long to experience it.

FA: Thank you, Alex.

AWG: Anytime.

(Disclosure: I have financial interests in Ornn, PSI, and 021T.)