Anthropic gives Claude AI a memory

Watch Anthropic gives Claude AI a memory - Rizzle on Dailymotion

Transcript

00:00Anthropic gives Claude AI a memory.

00:03Anthropic announced Claude can now automatically remember details from past conversations without

00:08being reminded, a feature rolling out first to team and enterprise users.

00:12Until now, Claude's memory required repeated prompts.

00:16Paid users could manually save chats.

00:18Now it autonomously incorporates project goals and processes, becoming more like a colleague

00:23attentive to Slack threads.

00:25The memory feature also carries over into Claude's project tools.

00:29Pro and team users can easily upload files, allowing Claude to remember and utilize context

00:34without repeated prompts.

00:36But concerns arise about incorrect memories, though Anthropic maintains memory is fully

00:41optional.

00:42Users can adjust Claude's memory settings for specific preferences, like design style, while

00:47disregarding others.

00:49Both OpenAI's ChatGPT and Google's Gemini offer similar features, though ChatGPT experienced

00:55increased delusional behavior during its memory rollout, noted by the New York Times.

01:00So yes, memories make AI more helpful, but also occasionally more unhinged.

01:05To counterbalance, Anthropic introduces incognito chats, ensuring full privacy without conversation

01:11storage, ideal for private discussions or brainstorming sessions.

01:25So yes, there's a lot of questions.