“There's a there's a project that's subnet 62. It's called Ridges AI... They spent about a million dollars in Tow like rewards and in 45 days, they hit 80% of…”
Jason is bullish on crypto-incentivized open-source AI and points to subnet 62 / Ridges, which lives on Bittensor. TAO is the network token whose emissions fund that flywheel, making it the liquid proxy for the thesis.
“I would not be shocked if you see Anthropic exiting this year at $80 to $100 billion in revenue.”
Gerstner walks through Anthropic adding Databricks-plus-Palantir of revenue in a single month and hitting its $30B exit target by March, then stakes a doubling-to-tripling to $80-100B by year end. The implied path is a much higher valuation than the $380B last round.
“I think if you're not an incumbent, so specifically, who do I mean? Meta, Google, and SpaceX. I think those three people who have all three of them... What…”
Chamath argues the AI race converges to a compute war that only capital-rich incumbents can win, and he names Meta first. Among the three he listed only Meta and Google are public, and Meta most purely turns the thesis into a buyable equity.
“just the top four hyperscalers, $350 billion of expected CapEx this year. On its way, I think Jensen said one trillion by 2030.”
Hyperscaler CapEx $350B this year heading to $1T by 2030, and AI revenue is finally validating the spend
“I'm a buyer of the shares today, notwithstanding all of the vibes that you describe.”
Against the chorus of OpenAI FUD (departures, secondary marks below the last round, the New Yorker rehash), Gerstner makes a present-tense buy call, betting the imminent Spud model and Codex ramp put OpenAI firmly back on the wave.
“do not count out Meta, right? I think Meta is absolutely in this game. Google is absolutely in this game.”
Fortress balance sheet and full-stack compute keep Google firmly in the frontier-model race
“I think the market's probably being a little too pessimistic with respect to at least some of these software companies... there's probably a little bit of the…”
Sacks thinks the market is overpunishing enterprise software on AI-disruption fear and that value buys remain. With no single ticker named, the IGV software ETF is the cleanest way to express a basket-level long on the beaten-down group.
“I guess you could say that Palantir is already one of them, right? It's an application company that's been turbo charged by these model capabilities.”
Application-layer winner already turbo-charged by frontier models as AI value capture climbs the stack