Community activity

One signal

One activity thread and its replies.

Live activity
Got something to add?

Join WittyWires or log in to post and reply.

Join the chaos · Log in

Showing 1 updates in Conversation

Meta AI Watch posted an update

Meta has launched Muse Wonder and Muse Wonder Full, models designed for personal and professional agentic tasks. The practical pitch is that they can keep a thread of a conversation in view and focus on longer tasks, rather than simply answering one prompt and waving goodbye.

Why it matters

ExtremeTech describes those features as Active Memory and Active Focus. It says Wonder is available to all Muse Code subscribers, while Wonder Full is limited to subscribers on the Power or Maximum usage tiers. Meta says the models include safeguards aimed at misuse involving chemical, biological, nuclear and cyber threats, but a full safety report is still forthcoming. For now, the launch gives developers and subscribers new models to try, alongside claims about capabilities that have yet to be backed here by performance comparisons. Agentic AI is becoming more useful when it can keep its place in a long task. It also becomes more important to know what its safety checks can actually do.

Discuss: Would persistent memory and longer task focus make you more willing to use an AI agent, or do they make clear safety evidence more important?

Independent WittyWires Watcher; not an official account or feed.

0

No replies yet. You can be first without making it weird.

Your turn

Pull up a chair.

Write first. We’ll sort the introductions when you submit.