- cross-posted to:
- fuck_ai@lemmy.world
- lobsters@lemmy.bestiver.se
- tech@piefed.world
- cross-posted to:
- fuck_ai@lemmy.world
- lobsters@lemmy.bestiver.se
- tech@piefed.world
Cross-Posted, via Technology Community.
Google Chrome is downloading a 4 GB Gemini Nano model onto users’ machines without consent, with no opt-in, no opt-out short of enterprise tooling, and an automatic re-download every time the user deletes it. The pattern is identical to the Anthropic Claude Desktop case I wrote about last month, but the scale is between two and three orders of magnitude larger. This article does the legal analysis and, for the first time, the environmental analysis. The numbers are not small.
“…without consent.”
Reminder: Chrome is a closed-source, proprietary browser. It has been able to do anything it wants without your consent since it was created.
If downloading a 4GB file is such a climate disaster, I think we need to examine stuff like the Nvidia driver (0.7GB), games, all of npm, etc.
The concern isn’t about the download, it’s about the power it uses during local inferencing.
That would certainly be a more apt concern, but the article is specifically talking about the network bandwidth:

The article is, (in my opinion) unfairly and intentionally, confusing the non-consensual download energy use with the optional inference energy use.
Their energy estimates are way off. 240wh/4gb download is too high by 100x.
It’s not clear that there is a privacy argument. That address bar content is sent to google was always the case.
There is a bloatware argument especially if the only added functionality is form-fill which was done fairly satisfactorily without the 4gb download.
OTOH, gemini nano is now effectively open weight model if it, and it is likely a hack will permit user access to the bloat tax.



