none of these labs are actually slowing down...

grok 4.7 is expected this week, and 4.8 will finish training this week

openai's Astra successor is close to solving another millennium problem

google is rumored to be using a bit of RSI in its model development

zhipu wants AI helping build its next-gen model