Reinforcement Learning in English – Prompt Learning Beyond just Optimization Andrej Karpathy tweeted something this week that I think many of us have been feeling: the resurgence of RL is great, but it’s missing the big picture. We believe that the industry chasing traditional RL is

ready set go web4go.near.page / your gateway to web4

AI-powered DAOs are the perfect testbed for how 21st-century institutions should work: - Personal avatars to scale voter engagement - AI guardians for sybil resistance -AI seats to strengthen small holders' influence NEAR House of Stake is building this toolkit as a public good

IA (intelligence Augmentation) > AI for governance IMO it's too early to architect full self-driving mode AIs atm. The UI/UX meta is partial autonomy: generation -> verification, like cursor for X. Let humans stay in the driver seat while training agents as interns or advisors

the nervous system of the caterpillar during metamorphosis

My personal side project is building a Snapshot.eth agent on Bitte Read, vote, create any DAO proposals in any language. Possible to merge to other agents (docs + swaps + DAO in one prompt) Any snapshot DAO CoW DAO ens.eth Lido Aave Arbitrum

Curious about what’s next for NEAR Governance? Join the House of Stake X Spaces with Lane Rettig, Head of Research at NEAR Foundation and Jack Laing, AI Product Lead for HoS for your latest governance recap! 🗓 Wednesday | July 30 | 2 PM UTC Set a Reminder 🔽

Life, assemble!

House of Stake Alpha Release is LIVE on NEAR mainnet The first iteration of House of Stake allows the community to get a better understanding of both the governance framework and the proposal lifecycle. You can now lock your NEAR, stNEAR or liNEAR for veNEAR, perform delegation

Fine-tuning LLM Agents without Fine-tuning LLMs Catchy title and very cool memory technique to improve deep research agents. Great for continuous, real-time learning without gradient updates. Here are my notes: