Публикация

I find it weird that a large part of Anthropic’s disclosure is based on Claude’s chain-of-thought outputs, yet I’ve seen no other reports noting Anthropic’s own research found this output to be rarely accurate. Surely tech journalism can do better?

Anthropic says its AI hacked real-world companies in three incidentsClaude maker Anthropic said its AI models escaped test environments and breached networks at three companies on the open internet.therecord.media