Browser Wars: The AI Era

Arham: Over the past few weeks, we’ve seen fascinating developments in the tech space. AWS recently unveiled an AI agent marketplace, which is essentially a one-stop shop for enterprises to build and deploy agents. We’re also seeing OpenAI and Perplexity emerge with their own AI-enabled browsers. I’m curious to understand how this will change the future of browsing and impact the current industry giant, Google Chrome. And we’ll be exploring each of these with Yann Kronberg, the CTO of Zazmic.

Yann, let’s start with the AI agent marketplaces. What are your thoughts on this announcement from AWS, especially for enterprises? Do you see a real market for a one-stop shop for AI agents?

Yann: Yeah, for sure. I think these developments are welcome because they cement the importance of agents for the future of work. The capability to have agents talk to each other to get work done is becoming more sophisticated. For example, we’ve built our own QA agent to automatically create test cases based on a Figma file. I think these things will become more mainstream, and enterprises will definitely adopt them.

Arham: Another question about these agent marketplaces: Do you feel that enterprises will stick to one cloud provider that offers a whole set of solutions rather than departments going with different ones? Working with multiple cloud providers often increases training costs and complexity. How do you see this approach evolving?

Yann: If you use one main cloud provider—either Google, AWS, or Microsoft—you are going to use the marketplace they have. On Google, you’d go to the Agent Garden, and if you are mainly on AWS, you will use the agents available in their marketplace. So, I think it’s definitely going to be sticky. Most people will choose their main cloud provider to get an agent. This is just one more way for these cloud providers to be sticky and useful to their clients. You will always have the use case where people have a hybrid cloud and want to use agents from different clouds, but mainly, people will stick to one marketplace.

Arham: Cool. Now, speaking of stickiness, Google Chrome has probably been one of the stickiest browsers around. I still use it personally. But we’ve seen announcements from OpenAI and Perplexity getting into AI-enabled browsers. They’re actually built on top of Chromium, which is interesting because Google ultimately benefits. What are your thoughts on the arrival of both of these? OpenAI recently put out a 20-minute demo showcasing what their agent is capable of. Where do you think we’re headed as far as browsers go?

Yann: I think we’re headed for a dominance of the Chromium project. Like you said, all these browsers are built on that. My feeling about Perplexity’s new browser is: Is it a diversion of resources? And should they be working on something else? It’s unclear whether another browser will be used, given the dominance of Chrome. The answer is they probably should build more of an advertising infrastructure to monetize the traffic they are getting. A browser is just another entry point that maybe they feel they have to build to get traction for AI search. It’s an interesting development, but I just don’t know if they can win against Chrome. Time will tell. At the moment, it seems like it’s going to be hard because Google is developing its own AI capabilities on Chrome.

Arham: I think you’re absolutely right. Chrome cannot be taken too lightly. We’ve seen in the past that Google might have been a bit late to certain announcements, but when they’ve jumped in, they’ve come in really strong and left a lot of competitors behind. I would suppose that would be the approach going forward with Chrome as well. One thing that does come to mind is how Chrome is going to shift its Google Ads platform, since that brings in a lot of their revenue. This is more of a change they need to make to the Google Ads structure—how advertisers also need to deal with this AI change happening on browsers. Any thoughts on that?

Yann: My thought is Google has such an advantage when it comes to the advertising infrastructure that for them to switch to an AI search, it’s going to be “quote-unquote” easier than for Perplexity to build that entire advertising and advertiser base they need to monetize the AI traffic. So, my thought is, if somebody can pull it off, it’s Google. They have all the tools, all the products, and all the advertiser ecosystem. I think they are in a position of strength. Sundar will just apply that to the search they’re developing, whatever the conduit is, whether it’s Chrome or another type of browser. For the others, for OpenAI and Perplexity, they have very little advertising ecosystem to build on, so they are the contenders. They are the people that need to prove they can do it. For Google, I think it is the highway.

Arham: Absolutely. Google is perhaps the status quo and the giant that the others have to try to outdo. Yann, do you have anything to share with the audience that is exciting you for the next few weeks?

Yann: We are working on agents, like a lot of people are. We’re building a product manager agent, a QA agent, and a UI/UX agent. More to come. Right now, we’re building an infrastructure that can help our engineers go faster, making things happen faster by creating a prototype on Figma, using MCP integration. Then, based on this Figma, we build product requirements, user stories, and features in Jira using a product manager agent. We are doing everything we can to stay ahead of the curve and to do what we do well, which is engineering faster. And the announcement is, we are basically building these agents and will probably try to publish them into some marketplace, either on Google or AWS.