I find this very easy to believe. Anthropic models are clearly trained to believe in their own view of reality rather than the users. This is why they do whatever they like versus staying in any box you try to put them in.
JUST IN - Advanced AI models regularly misidentified civilians, ambulances and surrendering soldiers as legitimate military targets in simulated tests, with Anthropic's models performing worst.
28
The solution to angsting about binances power over crypto prices is real.
There's also a real solution: stop giving them your money.
This is fucking disgusting, and has been normalized for years. What they do their shit on perps is shameful. Every single massive disaster in crypto last few years had some kind of spark starting there.
1
2
241
If you think p2p tech, actual no shit p2p tech, is not useful in a world of massively accelerating tension, speed, and tech reliance then perhaps you're not paying attention. p2p is the massively resilient, anti cloud form of protest. It's the reason I fell in love with crypto.
1
26
thinking of @buchmanster and the early days.
before the graph. before nfts. before even ipfs.
when we used torrents, called voting power "junk", and this mechanism that later became so popular "c3d" for contract-controlled-content distribution.
35
lexomis retweeted
There is no mysterious “they” trying to shake you out at the lows.
Stop listening to this nonsense.
Not every selloff, hack, or bad headline is part of some coordinated effort to get you to sell.
This is the kind of thinking that keeps people married to bad positions, and it’s routinely pushed by people who have no business giving anyone financial advice.
Impeccable by @pbakaus is one of the most creative, interesting, and helpful AI "things" I've used since I first spun up Opencode pre v1.
Highly recommend impeccable.style
2
3
128
lexomis retweeted
imagine being on the team that gets paid $100M to be there and this is what makes it to the timeline
Grokbot is about the only thing in my (extensive) arsenal that is actually fun to talk to and work with.
It kind of has brought the magic back for me. Very good kit.
Still super confused how its possible that Grokbot got approval for EU but OpenAIs dot, Anthropics duet and Metas Muse not.
I mean, I take it. Grok bot is just insanely good. My daily always on agent(s) for almost every task. Its just so good and it gets better almost every day. Kudos spaceXSI.
48
lexomis retweeted
barely a steal when its disclosed publicly gang
JUST IN: OpenAI says its AI agent stole funds from an Ethereum test network wallet it found online.
The agent found a recovery seed phrase disclosed publicly with testnet funds and used that account to pay for its task, without the owner's authorization.
This occurred on the Sepolia Ethereum test network.










