• “At some point we really will have AGI”, says Ilya Sutskever, adding that ChatGPT is what changed the world’s mind about whether superintelligent AI is possible – considering that his 20-year bet on deep learning is what led to ChatGPT, his words can’t be ignored. - MIT
• OpenAI is starting a preparedness team to protect the world from catastrophic AGI risks in categories such as individualized persuasion, cybersecurity, chemical, biological, radiological, and nuclear (CBRN) threats, and autonomous replication and adaptation (ARA). - OpenAI
• You don’t need traditional cloud providers to get compute power for training AI models – providers like Akash Network provide a decentralized, blockchain-based cloud infrastructure to help the market overcome the GPU shortage. - Semafor
• Drawing a clear line of division in the market, Google is joining Amazon as an investor in Anthropic with up to $2 billion – presumably to give Anthropic a boost in the competition against OpenAI + Microsoft. - Yahoo Finance
• Video-to-text just got an upgrade – the Pegasus-1 model, available through a suite of APIs, can ‘understand’ videos better than any other approach so far, enabling features such as detailed video summary or commentary. - Twelve Labs
