OpenAI launched an AI-powered net browser referred to as ChatGPT Atlas this week, which makes me marvel: Is it lastly time to ditch Safari?
That information was on our minds as Max Zeff, Sean O’Kane, and I mentioned the browser panorama — together with some lesser-known alternate options — on the most recent episode of the Fairness podcast. Nevertheless it doesn’t sound like several of us will probably be making a giant swap quickly.
For one factor, Sean famous many corporations have tried and finally did not unseat the foremost browsers because of their lack of ability to generate profits on the browser alone. In fact, that’s much less prone to be an issue for OpenAI, with its more and more huge funding rounds.
Max, in the meantime, has really tried out Atlas and different browsers that promise AI brokers will do the give you the results you want, and he mentioned there’s a “slight effectivity acquire” at finest. At different instances, you find yourself watching the agent “click on round on a web site” — is that one thing regular customers are actually crying out for? Plus, there are important safety dangers.
Learn a preview of our dialog under, edited for size and readability.
Anthony: I’m nonetheless on Safari, however so far as the search engine, which is tied to browsers, I’ve really been making an attempt to experiment with non-Google [options] as a result of I’m simply bored with seeing all of the genAI stuff on the high of my search outcomes.
I feel additionally there’s this query of: If these AI browsers take off, what does that imply for the concept of the open net usually? You possibly can nonetheless go to net pages, however I don’t assume it might be loopy to counsel {that a} web site is simply going to turn out to be much less and fewer essential as an increasing number of of our looking is managed by these AI interfaces and chatbots.
Techcrunch occasion
San Francisco
|
October 27-29, 2025
Max: I feel that this has been a giant thought that folks speak about lots: What does the agentic net seem like? And I feel it’s a fascinating query. Individuals have tried to give you all these options to work towards this future that [they] really feel is coming.
And I feel that there’s a sure facet of it that jogs my memory of earlier tech waves the place it’s like, “Okay, however what’s the precise expertise? What’s the worth proposition to a client of utilizing one in all these instruments?”
And it’s simply not tremendous compelling right this moment. I’ve tried out ChatGPT Atlas and I’ve tried out Comet and probably the most beneficiant estimation of them is, it’s a slight effectivity acquire. It makes you barely extra environment friendly.
However more often than not that I’ve tried these items, you’re slowly watching it click on round on a web site, doing a little job that I’d most likely by no means do in the true world. I’d have it, like, lookup a recipe and add the entire elements to Instacart. I’ve by no means carried out that. I feel all of the tech bros all the time say that instance within the movies, and I’m like, “I don’t know if individuals are doing that that a lot.”
That is simply this big hole, within the face of the tech trade proper now [saying], “We’re constructing all these instruments for the agentic net,” however why would a standard individual use this? And I don’t know.
Sean: I’ve not used any of these [AI browsers] however that’s largely as a result of I’m nonetheless very a lot an outdated head in relation to search and looking usually — quite a lot of the work that I’m doing includes in search of paperwork, which simply naturally includes trying by way of totally different discrete elements of net pages that I’m aware of, a number of Boolean searches on Google. Perhaps I’ll strive these someday if Google actually does up and kill Boolean search, which it appears like is coming in some unspecified time in the future, nevertheless it’s not there but.
The factor that’s attention-grabbing to me about these AI browsers is that we’ve seen different corporations attempt to compete within the browser area and so they all the time lose as a result of it’s simply not possible to generate profits on a browser as a product. And a few have tried to cost up entrance for it, they’ll type of get by for a short while, nevertheless it’s simply finally not sustainable within the face of competing towards Safari or Chrome or Firefox, for that matter.
What’s attention-grabbing to me … is you lastly have these corporations that simply have infinite cash, to allow them to journey it out so long as they need, as a result of they’re not really making an attempt to generate profits on these items but. Ultimately they most likely will, however OpenAI doesn’t must generate profits on this factor within the subsequent 12 months or two, they’ll simply have it on the market and let it take form.