The browser is becoming a worker
the browser is becoming a worker: chatgpt atlas wraps chromium around an agent that books flights and completes purchases, gemini auto browse lives in chrome’s side panel, and meta’s muse is heading to smart glasses on a cloud linux vm with its own browser and terminal. yet osworld 2.0 finds the best model finishes just 20.6 percent of long real-world tasks. the agent can click.
it cannot stay on task.
Context
Google introduced Gemini 3 auto browse in Chrome on Jan 28, 2026. UploadVR reported on Sep 29, 2026 that Meta's Muse agent is coming to its smart glasses soon.
Nerd Level Tech reports that OSWorld 2.0 shows even the best underlying model completes only 20.6 percent of realistic long multi-step tasks.
The Gemini auto browse feature, Muse coming to glasses and the 20.6 percent OSWorld 2.0 result match the sources, the last through a secondary summary. ChatGPT Atlas booking flights, Muse running on a cloud Linux VM with its own browser and terminal, and the side panel placement were not seen in the sources read, so unsupported here, not refuted.
Related work
- Google: Putting Gemini to work in Chrome ↗Primary post.
- UploadVR: Meta's smart glasses are getting its Muse AI agent ↗Glasses report.
- Nerd Level Tech: AI browser agents in 2026 ↗OSWorld 2.0 summary.
Watch next
- Find the OSWorld 2.0 results page.
Sources
- Google blog, Jan 28, 2026blog.google
- UploadVR, Sep 29, 2026uploadvr.com
- Nerd Level Tech, Jul 18, 2026nerdleveltech.com
Provenance
The note above is reproduced unedited from the original post, first published on Threads on 2 October 2026 at 21:44 IST. Sources are the papers and datasets the note draws on.
View the original post ↗Embed this note
More notes
The air is now being asked to keep its own ledger
the air is now being asked to keep its own ledger: ecmwf’s aifs compo becomes the first ai model to forecast atmospheric composition globally every three hours, cleanair simulates 365 days of pm2.5 over china in ten seconds, and a unified framework maps six pollutants at one kilometer across the whole country. the air now files its own composition report.
read the note →The current is now being asked to draw its own map
the current is now being asked to draw its own map: china’s langya 2.0 predicts six ocean phenomena including internal waves and mesoscale eddies, a deep net called wenhai resolves eddies globally with air sea flux formulas built in, and scripps infers surface currents from the way temperature patterns deform in satellite images. the ocean now files its own circulation report.
read the note →The soil is now being asked to report its own carbon
the soil is now being asked to report its own carbon: a nix color sensor paired with generative data augmentation predicts soil organic carbon without a lab, random forest drives 74 percent of soil health mapping studies, and sentinel 2 tracks five year carbon change across france and italy from 922 samples. the dirt now files its own carbon account.
read the note →