ChatGPT and Claude are ‘becoming capable of tackling real-world missions,’ say scientists
The scientists developed a tool called "AgentBench" to benchmark LLM models as agents.
Archive context
Older archive item. Useful for background and entity history, but not a fresh market-moving signal.
The scientists developed a tool called "AgentBench" to benchmark LLM models as agents.
Why this matters
This cryptocurrency story adds another data point to the current market tape and is useful when read alongside nearby source coverage.
Original source
Read on CointelegraphRelated market context
Anthropic’s Claude AI Predicts a Wild Price for Chainlink (LINK)
Anthropic’s Claude AI predicts that in a peak bull-market setup, Chainlink (LINK) could hit $150–$250 by January 1, 2027, with a s...
Kalshi Reviews Three Sub-$100 Bets That Called Trump’s Press Secretary Pick Early
None of the three wagers topped $100, but each backed Katie Zacharia for White House press secretary before her selection was repo...
Anthropic Claude AI Predicts BTC to $1M+ Ahead of IPO
Ahead of its highly anticipated IPO, Anthropic’s Claude AI predicts that by January 1, 2027, Bitcoin could hit $500,000–$1,000,000...
Sam Altman ChatGPT AI Predicts Bitcoin to Hit Insane All-Time High in 2026
Sam Altman’s ChatGPT AI predicts Bitcoin could reach a new all-time high in 2026, nearly doubling its October 2025 high of $126,00...
Hong Kong brokers’ crypto trading commissions fall 13.5% in first half of 2026
The decline in crypto trading commissions may signal reduced investor confidence and liquidity, impacting market dynamics and regu...
OpenAI, Google and Meta battle for AI internet domains as crypto firms target .bitcoin and .wallet
OpenAI, Google, and Meta are competing for AI-related internet domains as cryptocurrency firms pursue their own digital naming rig...