Who is the AI ​​Browser for?

OpenAI this week released an AI-powered web browser called ChatGPT Atlas. This made me wonder, “Is it finally time to ditch Safari?”

This news was on our minds when Max Zeff, Sean O’Kane, and I discussed the browser landscape (including lesser-known alternatives) on the latest episode of the Equity podcast. But I don’t think any of us will make a big transition anytime soon.

First, Sean pointed out that many companies have tried to usurp the dominant browser position, but ultimately failed, because browsers alone cannot make money. Of course, that’s unlikely to be a problem for OpenAI, whose funding rounds are getting bigger and bigger.

Meanwhile, Max said he’s tried Atlas and other browsers that say AI agents do the work for him, but there are at best “marginal efficiency gains.” You might also see an agent “clicking on a website,” but is that really what a normal user would want? Plus, there are significant security risks.

Read a preview of the conversation below, edited for length and clarity.

Anthony: I still use Safari, but I’m really experimenting with more than just Google when it comes to search engines associated with browsers. [options] Because I’m tired of seeing all the genAI stuff at the top of search results.

You may also have questions such as: If these AI browsers become widespread, what does that mean for the whole idea of ​​the open web? You can still access web pages, but I think it’s not unreasonable to think that websites will become less and less important as our browsing becomes increasingly controlled by AI interfaces and chatbots.

Max: I think this is a big idea that people talk about a lot. What is Agent Web? And I think that’s a fascinating question. People have tried to come up with all kinds of solutions for this future. [they] The feeling is coming.

And I think there are certain aspects of this that remind me of previous waves of technology. “Okay, but what is the actual experience? What is the value that we can provide to the consumer by using one of these tools?”

And it’s no less convincing today. I’ve tried ChatGPT Atlas and Comet, and the most generous review I can give is that they are slightly more efficient. It will slightly increase your efficiency.

But most of the time I’ve tried these things, I’ve been slowly clicking around on a website and watching myself perform tasks I’d never do in the real world. For example, look up a recipe and add all the ingredients to Instacart. I’ve never done anything like that. I think my tech friends always give examples of that in their videos, but I’m like, “I don’t know if people do that.”

Considering the current technology industry, this is just a huge gap. [saying]It says, “We’re building all these tools for Agent Web,” but why would the average person use this? And I don’t know.

Sean: I’ve never used any of them. [AI browsers] But that’s mainly because I’m still very old-school when it comes to searching and browsing in general. A lot of the work I do involves searching for documents, which naturally requires looking at various individual parts of web pages that I’m familiar with, and doing a lot of Boolean searches on Google. If Google really gets rid of Boolean search, I might try these out someday. I feel like that will happen someday, but it hasn’t happened yet.

What’s interesting to me about these AI browsers is that I’ve seen other companies try to compete in the browser space and always lose because it’s impossible to make a profit out of a browser as a product. Some people try to charge you upfront. You can get by for a while, but ultimately it’s not sustainable given the competition from Safari, Chrome, and Firefox.

What’s interesting to me is that we finally have companies that have infinite money, and they can ride it out for as long as they want. Because they’re not really trying to make money off of these things yet. That will happen eventually, but OpenAI doesn’t need to make money from this in the next year or two, it just needs to get it out there and make it happen.