AI lets the US and other post-industrialized societies compete against subsidized supply chains and low cost of labor. It allows you to rebuild national security. Don’t lose sight of that.

Hate to say it but Napa Cabs (and adjacent) are the best wine. California wins. Châteauneuf-du-Pape and a good granache - malbec blend close second.

shift+tab and walk away

Sold $10 of SOL on Phantom, got $5.40 in ETH. Sold $9.90 of SOL on the CoinList mobile app, got $9.88 of ETH.

Nic has made some great calls, but this one may not age so well. VCs are bringing their token investments to public markets via these vehicles, because they are too illiquid in crypto markets to be exitable. It’s a form of asset maturation in a decentralized way.

Being able to take decentralized assets public opens up a world of liquidity that was previously not available at late stage crypto. The downstream effects is better VC fund dynamics creating more mature risk funding ecosystem. More late stage funds. Better early stage funds.

sfcompute has grown 3x since july 10th

people are notoriously bad at predicting these kinds of technology trees. 140 characters, web 2 led to giant adtech monopolies that built ai research divisions that led to the transformer. video gaming led to the company that creates H100 matmul machines at industrial scale

this is why “state capitalism” is fake, you cannot plan greatness, you cannot subvert the natural evolution of technocapital for too long, you cannot predict the industries that will matter today to climb the next branch of the technology tree

There's always a tendency to force a social experience onto an unproven format/utility too early before the desire or need for either has been established. ie. the first few VR social networks on the winning VR platform will be fun experiments, but too early.

San Francisco Compute always felt like a love letter in the form of cheap supercomputers. Inference feels like the flowers we sent to your door. To our locked-in friends building the future, here is our gift to you: 80% cheaper inference costs.

We’re only 25% through the mania. If you’re not in AI, crypto, or somewhere about to go public, you’re missing out on the greatest opportunity in history. How work gets done, how people transact and spend their time is all being reshaped right now.

A lot of people confuse useful with valuable. There are many useful things that are not really valuable. And many valuable things that aren’t entirely useful. You want to be both.

FYI you want to go public or launch a token by June 2026.

The most important “ingredient” for any startup is the quality of the people. The second is the product.

If you have your own chain, you can’t be a credibly neutral platform for token sales. You’ll always be conflicted.

You can give people an opportunity they would never have had access to, make them money, and they will still be very upset with you. Strange.

You're never too late to start exploring crypto.