Many users saw ChatGPT Atlas's quick discontinuation as proof AI browsers are outdated and risky due to privacy and lock-in issues, while a few welcomed the shift toward extensions and on-device models.
Based on 8 visible X reactions from 34 accounts; directional sample.
Ask a question below.
Published answers will appear here.
@simonw @corbtt Yeah... said that myself in a slightly different way. MO needs to be whitelist (opt in) not blacklist (opt out - often only as a response to a negative consequence). Imagine all those lovely active secrets lurking in the browser's local storage just ripe for the picking.
@signulll every platform shift, someone tries to port the old surface forward. desktop sites got crammed onto mobile before native apps won. ai browsers clearly are backwards. even ai using apps as primary surface in a few years will look outdated.
@signulll Atlas dying 9 months in is the tell. The browser was never the product, the agent is. Once it books and buys on its own, tabs and chrome turn into background plumbing
@signulll I was very surprised AI labs didn't just build Chrome extensions and went to build full-on browsers instead. We're finally in a good plae.
@simonw This is another rung on the lock-in ladder for frontier labs. I'm going to do my best to keep using agent-agnostic tooling
@simonw the privacy issues are genuinely unsolvable as long as ai sees exactly what you see
With Atlas being retired in favor of the browser embedded in the ChatGPT app I wonder if the whole category of AI-enhanced browsers is coming to a close The security/privacy issues remain unsolvable IMO - I want my AI to use its own separate browser and stay out of the one I use
https://x.com/signulll/status/2075590206896456154/photo/1 https://twitter.com/polymarket/status/2075584061427339310
Many users saw ChatGPT Atlas's quick discontinuation as proof AI browsers are outdated and risky due to privacy and lock-in issues, while a few welcomed the shift toward extensions and on-device models.
Based on 8 visible X reactions from 34 accounts; directional sample.
Ask a question below.
Published answers will appear here.
@simonw the privacy issues are genuinely unsolvable as long as ai sees exactly what you see