AI LLM models evolved very fast over the past few years to the point where they enabled the creation of revolutionary professional tools with huge performance improvements which are impacting business decisions and creating the risk for huge unemployment impact.
Over the past few weeks though we've been seeing an increasing number of users complaining that Anthropic's Claude had lost performance after new models were announced. There's the suspicion that the company is facing a infrastructure crisis where they don't have the computing resources to keep their AI performing as before. Anthropic assumes some of the blame for a few bugs that they're fixing but splitting the blame; they're stating that users may be overloading the models with too much information. But there are other signs too that show that a different kind of limit may have been discovered.
There are a some guides that recommend care when talking to an AI to submit questions or delegate work to do.…