AI Capabilities Are Doubling in Power Every 196 Days

“In April, the AI lab Anthropic announced it had developed a new system, called Mythos, that it said was too powerful to release to the public because of its ability to find gaping holes in online security systems.” Read it at the Telegraph.