From our latest Radical Talks episode with Michael Rubenstein, Co-Founder and Co-CEO of Firsthand
Everyone’s racing to build the most powerful AI. Fewer people are asking what happens to the commercial layer of the internet when they do. The U.S. digital advertising market is a trillion-dollar-plus space, and it runs on a model built for a different era: static messages, audience predictions, and a funnel full of hops that were never really built for the consumer. AI doesn’t just optimize that model. It breaks it.
In our Radical Talks conversation, Michael Rubenstein, co-Founder and co-CEO of Firsthand, architect of DoubleClick Ad Exchange, and one of the builders of AppNexus, makes the case that advertising is undergoing its most consequential structural shift yet. And unlike previous transitions — from offline to online, from direct to programmatic — this one doesn’t just change how ads are bought and sold. It changes what an ad is. This isn’t an incremental improvement on a working model. It’s a rebuild of the model itself.
The Gap Between Intent and Response
The current model is built around prediction. A brand guesses what a consumer might want, pushes a static message in front of them, and hopes they click through enough hops to eventually convert. It is, as Rubenstein puts it, the brand’s internet, not the consumer’s.
The AI internet works differently. Consumers have the power to pull information directly to them. They declare their intent in real time. And brands, for the first time, have the technology to meet them there — not with a guess, but with a response.
“It’s a future where the brand meets the consumer where they are in that moment of intent or consideration, unearths what it is that the consumer is looking for, and then creates an internet experience for them on the fly built around that intent” says Rubenstein.
The result is a shift from the internet — a static, one-size-fits-all experience — to your internet, one that is highly customized and personalized to exactly what the consumer is looking for in that moment.
From Ads to Agents
Firsthand’s infrastructure enables brands to deploy intelligent agents directly into media environments and their own owned-and-operated properties. The mechanics are straightforward. For example, a consumer lands on a publisher’s site researching three-row SUVs. In the past, Volvo might have served an ad based on a demographic prediction. Today, Volvo can deploy a brand agent that meets the consumer where they are, engages them directly, unearths their specific needs, and guides them through a personalized journey — all without forcing them through a series of disconnected steps.
Firsthand is already in-market, running dozens of live campaigns with major brands and more AI-native advertising campaigns, Rubenstein believes, than any other company has run to date. The results are not marginal. Consumers are engaging at higher rates than with traditional static advertising. Conversion rates are running as high as 10x above what a standard campaign would produce.
What Hangs in the Balance
The stakes for brands are significant. There is a real question, Rubenstein argues, about whether AI will be empowering or debranding — whether companies will show up in the agent economy as owners of their own AI or simply as ingredients in someone else’s.
“When I meet with the CMOs and CEOs of major companies, that’s one of the things they’re very concerned about right now. So much of their value is the fact that they are brands. Consumers recognize them as brands” says Rubenstein.
The name Firsthand is not accidental. The company’s core belief is that brands shouldn’t outsource their consumer relationships to third-party platforms. They should own their AI, deploy it directly, and build durable, first-party connections with the consumers they serve. That’s how brand equity gets built in the agent economy.
What’s at Stake
Looking ahead, Rubenstein sees the internet itself converging around the agent model. Static websites, static ads, the hop-to-hop nature of the previous web, will seem archaic. Answer engines, browsers, and brand experiences will hybridize in ways that make today’s categories look overly rigid.
The bet Firsthand is making is straightforward: if the internet becomes AI-powered, inference-driven, and highly dynamic — and most people agree it will — then the advertising and marketing experience will follow. It won’t be a separate layer bolted on top. It will be native to the new internet, built around the consumer, created on the fly.
What’s different this time, compared to the transition to programmatic, is that the mandate isn’t coming from the bottom up. Boards, CEOs, and CMOs all have a directive to figure out AI. The education and evangelism that defined the early days of programmatic still matter — jaws still drop when brands see what’s possible — but the organizational will to move is already there.
“AI has a top-down mandate. I think this is going to move faster and be bigger than anything we’ve seen before. It’s not just advertising. This touches marketing, commerce, payments, everything.”
This conversation is a reminder that the AI revolution isn’t just about models and chips. It’s about what happens when intelligence meets commerce — in a trillion-dollar market ripe for rebuilding from the ground up — and who builds the infrastructure to connect them.
This post is based on insights from Radical Talks, a podcast from Radical Ventures exploring innovation at the frontier of AI. For more conversations with leaders in AI, subscribe wherever you get your podcasts.