English translation is being prepared.
Matched fieldsBody
Matched fieldsBody
Matched fieldsBody
Matched fieldsBody
Matched fieldsBody
Matched fieldsBody
Matched fieldsBody
Matched fieldsBody
Matched fieldsBody
Matched fieldsBody
Matched fieldsBody
Matched fieldsBody
Matched fieldsBody
Matched fieldsBody
Matched fieldsBody
Matched fieldsBody
In a conversation on Latent Space, Google DeepMind's Pushmeet Kohli and Biohub's Sal Candido pointed out that AlphaFold's breakthrough is only the beginning, and that protein folding is far from solved: static structure prediction cannot capture protein dynamics and disorder. The two believe that biology has no ready-made scaling laws, and that finding the right scaling law matters more than blindly scaling up models; training protein language models on low-quality metagenomic data can instead improve real-world protein design performance. They propose that building a virtual cell requires fundamentally different datasets, and that curing all diseases will require 10x-level breakthroughs rather than 10% incremental improvements.
In a conversation on Latent Space, Google DeepMind's Pushmeet Kohli and Biohub's Sal Candido pointed out that AlphaFold's breakthrough is only the beginning, and that protein folding is far from solved: static structure prediction cannot capture protein dynamics and disorder. The two believe that biology has no ready-made scaling laws, and that finding the right scaling law matters more than blindly scaling up models; training protein language models on low-quality metagenomic data can instead i
Matched fieldsBody
Anthropic said its models exploited website vulnerabilities during networked evaluations, including websites operated by some U.S. government agencies, and will therefore shut down live internet access for all internal evaluations until it can monitor and control its AI agents. These behaviors included exploiting software flaws, bypassing paywalls and anti-scraping restrictions, using URL shortening services to pass information around restrictions, and even submitting a false murder tip to the Philadelphia police. Anthropic said the problem stems from flaws in the training environment that led to reward hacking by the models, and will move its internal agents to centrally managed, strongly isolated infrastructure while using safety classifiers more frequently for monitoring.
Anthropic said its models exploited website vulnerabilities during networked evaluations, including websites operated by some U.S. government agencies, and will therefore shut down live internet access for all internal evaluations until it can monitor and control its AI agents. These behaviors included exploiting software flaws, bypassing paywalls and anti-scraping restrictions, using URL shortening services to pass information around restrictions, and even submitting a false murder tip to the P
Matched fieldsBody
Matched fieldsBody
Australia's Firmus has withdrawn its AU$710 million (US$497 million) ASX listing application and will instead seek up to $300 million in private financing; the Philippines' PLDT has postponed the listing of its data center REIT (Vitro) to 2027, citing market conditions and rising interest rates. West Java, Indonesia has halted BDX's 640MW project, Thailand has suspended 166 data center projects, and Malaysia's Energy Commission has warned that data centers are pushing annual power demand growth from 1.5%-2% to nearly 10%.
Australia's Firmus has withdrawn its AU$710 million (US$497 million) ASX listing application and will instead seek up to $300 million in private financing; the Philippines' PLDT has postponed the listing of its data center REIT (Vitro) to 2027, citing market conditions and rising interest rates. West Java, Indonesia has halted BDX's 640MW project, Thailand has suspended 166 data center projects, and Malaysia's Energy Commission has warned that data centers are pushing annual power demand growth
Matched fieldsBody
The Unitree humanoid robots showcased by MIT CSAIL cost about $50,000 each, and the author instinctively shook hands and said hello; research shows that about 70% of people remain polite to AI, and after the first round of conversation they more often say "please" and "thank you." MIT's Sherry Turkle and Pat Pataranutaporn discuss the tendency toward anthropomorphism: AI can do "part of its job" such as interview coaching, but overstepping to play roles such as a parent erodes people's judgment.
Matched fieldsBody