3 ms·
Given Grok was just caught uploading whole codebases and sensitive .envs without permission, how can anyone possibly trust this?
by dbbk 3mo ago
Given Grok was just caught uploading whole codebases and sensitive .envs without permission, how can anyone possibly trust this?
- gjsman-1000 3mo ago[flagged]
- verandaguy 3mo agoNot GP, but personally I don’t trust claude. Work does for some reason, and work strongly encourages using it, but i stay far away outside of that. I think every product in this space should have to clear a very high bar for trust rather than making it a race to the bottom with whataboutisms.
- nullbio 3mo agoAnthropic use Deepseek behind the scenes instead of their own models, whenever they feel like it. People have shared outputs from Claude that show internal tool errors the models have accidently returned with Deepseek error codes.
- rnewme 3mo agoI'm curious about this, where'd you learn that?
- nullbio 3mo agoThere were quite a few people from the AI community on X sharing screenshots.
- verandaguy 3mo agoI don't know how this relates to my comment, but sure, I guess?
- dkdbejwi383 3mo agoJust because Billy misbehaved doesn’t make it ok for Jenny to do the same.
- wraptile 3mo agoYou could argue that you don't care what Claude does after work hours as long as it works your hours as you expect it to. Grok clearly failed the latter.
- maxlin 3mo agoObviously it didn't. It's the current value king and trails fable by single percents. And people shout of an obvious debug flag left on, something that can hardly be called rogue considering how notable it was. Don't be ignorant.
- nullbio 3mo agoThat's just the tip of the iceberg for Anthropic. I don't use Claude either, for the same trust reasons.