This last few years, the browser market has become quite a saturated space and hot topic. Everyone wants a slice of that market share, and the wealth of data that comes with it. And of course, the best way to do that is to capitalise on a trend.
But even after Dia, Edge, Comet, and now Atlas - users are still pretty confused. What problem are these browsers even looking to solve?
Let's start at the beginning.
The concept.
The concept seems simple on paper. A browser that can browse for you, cut through all the noise and help you do things at lightspeed. The ability to do a task or understand a concept as simply as you would explaining it to someone.
Sounds great on paper, but in practice it's really not that simple.
The core problem.
To put it simply, the core problem is that these browsers are positioned wrong, to where they are a solution looking for a problem.
They're all desktop browsers, first off. If I'm at my PC, I'm usually here to do work. Why would I want to hand off tasks when I'm here to do them? If this were a mobile app, sure - I could kind of see it. But if I'm already there to do a task, there's a good chance I'll actually do it faster than an agent that has never seen the site before.
Another really important problem is stripping away web design. Design is a core part of the experience, it's not some simple optional extra that you can just.. toss away. Would you trust your favorite sites the same as you do today if you stripped them of everything but their text? When you reduce every site to its core, the bar on trust and credibility moves. And while I'm not against LLM websets or expanded search, it should not be trying to replace websites.
They also are notoriously bad at showing their value. Most of their pitches showcase things that were already not a big deal, like organizing travel, or something a regular modern LLM can do just fine. There's very little gain for the data sacrifice of moving your whole life to their browser.
ChatGPT's new Atlas browser is a really good example of a lot of these issues. Let's put them into perspective.
Atlas.
OpenAI took their sweet time with their Atlas browser, as it was something that had been long rumored for years before its launch. However, it launched to be nothing special almost immediately.
First, it was macOS only. Apple-first releases are normal for OpenAI, but for such a big product, and one backed on Chromium, they really had no excuse. Especially when some of their biggest competitors were on Windows even during their betas.
Once you're in the browser, you feel very.. underwhelmed. They somehow managed to make a browser less impressive than Comet, with a way bigger resource pool. It actually just feels like Chrome if you tacked ChatGPT onto it. The memory stuff isnt really new either, and just uses your browsing history most of the time.
To be this late to the race only to release such a frankly pointless product is not a good place to be in, and I have to assume they did it because they wanted to have a foot in the door. Because if I was in their shoes I would not have shipped this at all.
What's next?
Good question - and we're already starting to see some of it.
Perplexity and ChatGPT are both launching integrated shopping already, and I feel like this is just the start of advertising and "intentionally poisoned" results.
If everything looks the same, ads are going to blend in better. Meaning they have to fight less, meaning more conversions. Even better if you have your browser vouching for them.
It's likely going to be Google Ads 2, and I fear it's no longer a case of how, it's a case of when.