Thinking Machine Labs raised $2B in the largest seed round ever.

Founded by former OpenAI executive, Mira Murati, Thinking Machine Labs’ seed raise dwarfs the next two closest rounds, Yuga Labs ($450M in 2022) and Lila Sciences ($200M in 2025).

The money was raised at a $10B post-money valuation in a round led by a16z.

Murati’s goal is to build general purpose AI systems that are customizable and capable for enterprise use first, then consumers.

According to the Financial Times, an investor who passed on the deal said Murati’s pitch offered no information about a product or financial plans. She was also given a board vote equal to all other votes combined plus one.

For context, OpenAI’s most recent post-money valuation was $300B, and the company reportedly hit $10B ARR earlier this month.

Kalshi vs. Polymarket

Just as reports came out that Polymarket was in the process of raising $200M at a $1B valuation, Kalshi announced their $185M raise at a $2B valuation, signaling the market’s preference for a regulated prediction markets platform.

• Kalshi’s raise was led by Paradigm with participation from Neo, Sequoia, BOND Capital, and Citadel CEO Peng Zhao

• Polymarket’s raise is reportedly being led by Founders Fund

Sidebar: ironic that Paradigm, one of the largest crypto focused VCs, chose Kalshi (not a crypto play) over Polymarket (a crypto play). Kalshi is one of three Paradigm investments without a crypto angle. Not a good look for Polymarket!

The big difference in the two companies’ valuations can primarily be explained by Kalshi’s status as a CFTC regulated platform, which allows them to operate in the U.S. Polymarket is cryptocurrency-based, and is forced to restrict usage to non-Americans. There’s probably some optimism at Founders Fund that Thiel can pull strings in D.C. over the next few years and get past the regulatory hurdles. In the meantime, Polymarket is focused on capturing international market share.

Prediction markets are now an emerging asset class.

With roughly two hundred mil in hand, Kalshi plans to scale engineering, partner with brokerages, expand liquidity, and develop new products for traders on the platforms.

In other news, Michael Burry has drastically altered the composition of his portfolio.

Burry is most notorious for his impressive prediction of the 2008 housing bubble that paid off well for him and his investors.

He cut his portfolio from 13 positions to 7, with all but one consisting of put options. The legendary “Big Short” investor is once again flashing red on the market, this time aiming squarely at two overheated zones: AI-fueled U.S. tech and Chinese internet giants.

His top position? A massive $97.5M bet against NVIDIA, representing nearly half of his $199M portfolio.

Strikes and expirations aren’t disclosed in the 13F, but if history is any guide, Burry’s not making this move lightly. He’s calling bluff on the AI boom—and hedging hard against China’s shaky footing.

So what’s the probability that Burry is vindicated again?

Probably higher than you think, for a couple of reasons.

One reason is geopolitical instability.

But that seems to be a long shot.

Markets seem to have grown numb to geopolitical risk. Case in point: Last Saturday, U.S. involvement made an Israel–Iran escalation feel imminent.

By Monday, the markets shrugged and the S&P barely blinked.

So unless there is a full-fledged conflict that poses a real threat to the US economy and / or markets, it doesn’t seem to be likely.

The more probable scenario? The AI house of cards folding in on itself.

If you’ve spent any time in startup circles lately, you’re well aware most “AI-powered” companies are glorified ChatGPT wrappers. Rather than building proprietary models, they rent a room in OpenAI’s high rise.

This creates an interesting codependency.

These startups rely on ChatGPT to function—but OpenAI relies on them to a certain extent because these wrappers are OpenAI’s distribution layer.

If they don’t convert users to paid plans quickly enough, they die. And when they die, ChatGPT loses a small part of its viral growth engine.

Now add in the most important node in this ecosystem: NVIDIA. It doesn’t just supply the chips–it controls the entire AI supply chain.

From model training to inference, it all runs on NVIDIA’s silicon. Any stumble there—logistical, political, or competitive—would send tremors through the whole stack.

NVIDIA, at the moment, seems to be both the linchpin for public markets and private venture backed companies.

If any type of disruption in the company’s supply chain or regulatory action that adversely effects it occurs, this could not only send a severe correction to the public equity markets but also cause a massive reverberation in the startup world.

In the midst of AI hype, Burry seems to be the lone ranger that thinks the ground was never solid to begin with.

Headlines

• Jane Street co-founder says he was duped into funding AK-47s for Sudan coup. The Independent

• Meta in talks to acquire voice cloning startup. TC

• Why a16z believes Cluely is the new blueprint for AI startups. TC

• Bill Ackman pledges to back any mayoral candidate that can beat Zohran. Fortune