Oct 13, 2026 · 5 min · 7 segments
Meta launches new agent-focused models, Microsoft expands its custom AI infrastructure, and an emerging protocol begins defining how personal agents interact with businesses.
The company introduced Muse Wonder and Muse Wonder Full, calling them its most capable models to date for personal and professional agent tasks.
Muse Wonder Full is available to Muse Code subscribers.
Meta says Muse Wonder will also power more experiences in its Muse personal agent, beginning with subscribers, but the announcement did not give a rollout date.
The models emphasize two features.
Active memory is designed to carry relevant details across conversations.
Active focus is designed to keep an agent working toward a goal, try another approach after a setback, and ask for clarification when necessary.
Meta also says the models perform strongly in coding and in fields including law, finance, and health.
Those performance claims still need independent testing.
For companies building customer or employee agents, the strategic point is persistence.
The model race is moving beyond one-shot answers towards systems that remember context and continue multi-step work.
That raises the value of good permission controls, audit trails, and clear approval points.
The company introduced Muse Wonder and Muse Wonder Full, calling them its most capable models to date for personal and professional agent tasks.
Muse Wonder Full is available to Muse Code subscribers.
Meta says Muse Wonder will also power more experiences in its Muse personal agent, beginning with subscribers, but the announcement did not give a rollout date.
The models emphasize two features.
Active memory is designed to carry relevant details across conversations.
Active focus is designed to keep an agent working toward a goal, try another approach after a setback, and ask for clarification when necessary.
Meta also says the models perform strongly in coding and in fields including law, finance, and health.
Those performance claims still need independent testing.
For companies building customer or employee agents, the strategic point is persistence.
The model race is moving beyond one-shot answers towards systems that remember context and continue multi-step work.
That raises the value of good permission controls, audit trails, and clear approval points.
The rest of this transcript — segmented and speaker-labeled, so you land on the exact moment something was said
Search every transcript — by keyword, by phrase, or by meaning, across every show Radar indexes
Trends — what is surging across podcasts, measured against its own baseline
Alerts — when a name you follow appears in a newly indexed episode
No account is needed to search Radar.