Lately, numerous companies have announced “AI Web Browsing”. Opera with their local built-in Agent. Perplexity with a new browser.
There is an open-source “Browser-use” which according to their benchmark performs better than OpenAI’s Operator.
All of this indicates one thing: Automating browser tasks will become a zero-activation energy task, and quite soon AI navigator & posters will become indistinguishable to humans.
So, my question(s) are:
How are you using Operator? There’s been very little noise from anyone showcasing things that these automated tasks can do.
What do you think the future of the WWW will look like?
Me, personally, I don’t use Operator. It’s a token-heavy, uncontrollable tool that inefficiently browses the internet. I think there will be some web browsing that will incorporate a clever mix of free DOM navigations, and then use AI when the DOM navigating fails.
For #2. I believe that the public WWW will become increasingly bare. Information won’t be as easily accessible. Advertising on websites will have greatly diminishing returns. Leaving only privately funded informational pages with an intention of influencing AI left.
People will become increasingly scared and less motivated to post on the public internet. More posters will be AI and not worth speaking to. Some posters will create their counterpoints and message using AI, essentially becoming a lazy mask feature. AI slop will become increasingly unbearable. Any site with a valuable interface will have to be locked down by intrusive attestations and authorization.
Biometrics will become more common in social media. Places like X will increasingly become echo chambers where AI retrieves information generated by AI to post more information with. It will be feedback-loop hell
I have heard the same: “it’s not there yet”. I find all of this morbidly fascinating to watch. As a programmer it’s a guilty pleasure seeing the things people are willing to spend numerous amounts of tokens on, when it could be done with an hour or two of programming.