Andy
73.7K posts


On the @theallinpod this week, @chamath asked @nvidia CEO Jensen Huang about decentralized AI training, calling our Covenant-72B run "a pretty crazy technical accomplishment." One correction: it's 72 billion parameters, not four. Trained permissionlessly across 70+ contributors on commodity internet. The largest model ever pre-trained on fully decentralized infrastructure. Jensen's answer is worth hearing too.

✅Etherfi buybacks are back on (no clue why they were off for a few days) ✅ Final investor vest sent out. No more VC overhang. ✅ Daily spend at ATH using cards ($2.5M/d) ✅ Staking TVL back above $6b ✅ Upbit listing yesterday ✅ Adding ~300 new cards a day ✅ CAC going down, ARPU going up, Payback period compressing sallygazzy mode note: dcf cap seeded ethfi

“Bittensor is recreating a lot of the early interest people saw during DeFi Summer, where multiple projects were launching.” Andy asks Synth founder James Ross for his bull thesis on Bittensor. Hundreds of millions of dollars could flow into the Bittensor ecosystem, and TAO could go 10x, 20x, or even 50x from here. We saw something similar with Templar, which generated major interest and helped drive TAO up roughly 30–40%. Will we continue to see this growth across the Bittensor ecosystem?


Frontier model pre-training is increasingly concentrated The compute required for state-of-the-art training doubles every 6 to 10 months, pricing out anyone but the major labs: OpenAI, Google, Meta, and a handful of others with billion-dollar GPU budgets Templar is emerging as one of the most credible challenges to that concentration and they’ve just completed the largest decentralized pre-training run, EVER You NEED to pay attention to open sourced decentralized alternatives @NousResearch also released its Hermes Agent which is being dubbed “the next OpenClaw” Decentralized model training will become bigger than you think; now’s the time to educate yourself We published a @KhalaResearch report on Bittensor last month which covered Templar (I’ll link below)



LATEST: ⚡ Visa Crypto Labs has launched a command line tool that lets AI agents make card payments directly with the interface.

Nauman Sheikh on what the SEC commodity classification actually unlocks: "Now these tokens can be used as collateral." "CME is piloting Bitcoin, ETH, stablecoins as collateral against derivative trades." "It makes the market more capital efficient. Brings in new institutional capital." "There was a huge divergence in volume. All in Asia, offshore markets. Now volume can come back to the US."












