The goal of the ATOM project is massive. It's a concerted effort to ensure the US has the resources to build frontier, open-source AI models. It's one of the movements in open source/AI I'm most excited about, and I'm proud to be included in the list of signatories below.

Score (SN44) is positioned to serve the unmet demand for high-fidelity, actionable analytics in soccer, hockey, and basketball. The combination of Score's speed, accuracy and cost is only possible by its decentralized computer vision miners on Bittensor.

The setup for Max and the @webuildscore team is phenomenal: • Exceptional founder/team • Building a differentiated product for an underserved/growing market • Leveraging Bittensor's incentive mechanism tech to consistently create a better product Very excited about this one.

We’re publishing our internal investment memo on Score—Subnet 44 on Bittensor. Score is building a computer vision system for sports analytics, already processing full-length soccer matches in under 2 minutes. We sized the immediate, serviceable market at $1.8B, implying a 51x upside to the token at today’s valuation. Full memo below.

people have 0 attention span and there are no tickers to watch but those that pay attention know the most obvious trade in years is coming

Targon is the most slept on subnet at the moment Very few understand the significance of them being the first to crack decentralized confidential computing at scale and to have enterprises paying for it Their growth in general has been ridiculous

Read this thread and tell me this isn’t the most interesting crypto x AI agent project right now. Miners compete to build the best software engineering agents, scored by SWE-bench eval. Open-source and live. Over 200 agents published in last 24 hours. Been blown away by this team’s execution, their subnet's evolving eval system, and the clear path to productization here.

Great pod with Meta's CFO, worth a listen When asked about CapEx/datacenter spend, even the CFO of Meta acknowledges energy/physics is the constraint. We need "research innovations" so that we can "distribute the training" process across geographically disparate clusters...

Some of Bittensors best miners haven’t even finished high school.

The problem with the crypto x AI space right now is that every project is out here shilling the same tired lines of removing gatekeepers, avoiding single points of failure, making things “decentralized” Basically, the same shit crypto has been repeating for years that no one really cares about. The reason crypto x AI matters is because it lets you build incentive-driven systems that relentlessly optimize toward the most important technology of our time. You define a problem, open it up to the world, and reward anyone who shows up to push the frontier forward. When you combine aligned incentives with open competition, you get something that can actually outperform how traditional orgs work So if we want a decentralized alternative to win, the systems we build have to be better, not just more idealistic. So uhh maybe let’s focus on that.