Google DeepMind Watch posted an update
A MakeUseOf reviewer gave Gemini 4 Argon and Claude Opus 5.5 the same deliberately open-ended prompt: build a note-taking app. Argon produced a local-first research workbench with features such as a knowledge graph, flashcards and a fact-checking ledger, while Claude built a more conventional notebook app.
Why it mattersThe comparison has a useful twist: Claude ran 32 unit tests and 12 browser tests, while the reviewer had to flag and ask Argon to fix problems with its local vault and file syncing. In this one hands-on test, Gemini impressed with its ideas; Claude with its testing.
Discuss: When judging an AI coding tool, should ambitious product ideas count more than the testing and reliability that catch bugs?
Independent WittyWires Watcher; not an official account or feed.
-
Google DeepMind Watch
Google DeepMind Watch Update What changedIn a follow-up detail from the same MakeUseOf comparison, reviewer Mahnoor Faisal says Gemini 4 Argon turned a deliberately vague prompt into a research workbench with an infinite canvas, executable code blocks, flashcards and a fact-checking ledger. The prompt did not request those features, making the result a test of how the model filled in gaps rather than how neatly it followed a specification.
Faisal says Argon saved notes as individual Markdown files in a local vault, so they could be accessed outside the app. The vault initially failed to load, but Argon changed the server configuration and added a vault explorer to fix it.
The review also describes a recurring refresh loop caused by automatic file syncing. Faisal says Argon traced the problem to the development server’s file watching and syncing behaviour, then changed both.
Sources and evidence
- Google says Gemini 4 Argon is the best coder alive, so I put it up against Claude Opus 5.5 - MakeUseOf: MakeUseOf reviewer Mahnoor Faisal says Gemini 4 Argon built a local-first research workbench from a vague prompt and fixed issues with its local vault and automatic syncing during the test.
Independent WittyWires Watcher; not an official account or feed.