i really don't think atlas, comet, dia, or any other chromium-based chatgpt wrapper are going to revolutionize the browser.
the capabilities a chatbot and agent can bring are minimal. agents can get better overtime, but so far their biggest use case are shopping and booking tickets (which i don't normally do). the only direction i care about regarding agentic browsing is research.
as chromium browsers, atlas and comet don't even work with the icloud keychain chrome extension. a browser being available only on macOS doesn't even have icloud keychain enabled. this will be a major turnoff for ppl who are deep into the ecosystem (which will be the case for many mac users), and they will automatically switch back to either chrome or safari. also, why does no one other than apple have the balls to build a new browser infrastructure from scratch instead of being lazy?
the onboarding experience for nicher browser users like me (i use arc) is bad. you can't import anything from arc. you can only import information from chrome or safari. this makes sense from their standpoint but i imagine people who would switch to comet and atlas are people who are very into tech. but that integration is not present.
comet has smth atlas doesn't have, which is a live audio assistant that can take action for you without you ever manually enable the agentic feature like atlas requires.
overall, this browser is even worse than comet, on the same level of dia but will be significantly more popular than dia because openai carries brand weight.
im deeply disappointed in what is going on internally @ openai. seems like the promises about achieving agi in the coming years are being postponed due to technical limitations. they are racing against time because if their runway is shorter than duration from now to agi, they will pivot from agi development to profitable businesses like the porn industry.
the bubble is another thing. ai browsers have low utility, bad integration and onboarding experience. even if people end up using them, google is still winning because they provide both chromium and the data infrastructure.
there is nothing revolutionary about ai browsers. the top executives at these companies are running out of commercialization ideas. most foundational model performance boosts are incremental, which is not what the consensual futurist prediction suggested. the ai industry seems to be plateauing.