Anthropic’s Capacity Pinch Cracks the Door for OpenAI: Dave Lee

When Anthropic PBC announced that its new model, Mythos, would have only a limited release because of the vast security risks, it reverberated across businesses and governments as the potential tipping point that many had predicted and feared.

But some skeptics wondered if something else might be going on. Was Anthropic limiting the use of this model because it simply couldn’t handle the additional demand?

Anthropic has said the restriction was motivated purely by safety. But the accusation has grown louder as external analysis has suggested Mythos’ advance in capability might not be as fearsome as billed and Anthropic’s reliability issues have become an industrywide talking ...