Robert F. Kennedy Jr. confirms 2028 presidential run, known as Bitcoin supporter with over $1 million holdings
The presidential candidate has disclosed owning at least $1 million in Bitcoin with no plans to divest. He advocates for Bitcoin's decentralized nature and fixed supply as tools to stabilize the dollar and prevent currency devaluation alongside other hard assets like gold and silver.
Hey buddy, Sentient just gave a shoutout to @guardrailai
team for all their safety support over the past few months.
They've been teaming up to build some super critical components and are cooking up open-source AI together.
@SentientAGI
Hey, here’s my bold call for Doma in the next 30 days:
Doma will announce they’re teaming up (or rolling out a pilot) with a major traditional domain provider.
It’ll start being seen as the bridge between the old-school internet and the new.
@domaprotocol
CryptoAnalystBench: A New Test for Crypto AI Assistants
So here's the deal – the crypto world is changing like every single minute, right? And most AI assistants just totally fail when you throw conflicting data at them. They sound super convincing, but then they give you completely contradictory conclusions. It's wild.
What's CryptoAnalystBench?
Basically, it's a new way to check how good AI answers actually are in the crypto space. It's different from your regular tests because it doesn't just measure facts – it looks at:
How deep the analysis goes
Whether the info is actually current
How consistent the answers are when you've got contradictory data
Whether the answer actually makes sense for what you asked
Why does this even matter?
Picture this: the market's tanking, all your indicators are pointing different directions, and you need a clear take on what to do – not just a random pile of facts. That's where CryptoAnalystBench comes in. It catches mistakes that other tests completely miss: outdated wording, shallow analysis, stuff that contradicts itself.
How does it actually work?
They built this test based on real questions people actually ask – like 2,000 different scenarios across 11 categories. Then an impartial AI system grades all the answers using clear standards, without even knowing which model wrote what.
The payoff? Now AI teams can actually fix specific weak spots in their models instead of just guessing where things might go wrong in real situations.
@SentientAGI
@Medvidio@PolymarketTrade This is a funny and slightly creepy reflection of our own internet culture, not the real beginning of a machine civilization. No need to build a bunker just yet
@Ju1i_crypto@domaprotocol@D3inc Great free domains! It's nice to see these promotions popping up regularly—a joy you quickly get used to, but one that still feels pleasant nonetheless.