A recent memo from OpenAI lays it out clearly: they aim to become the dominant interface for everything on the internet. Bold? Yes. Feasible? Maybe! As a faculty member at Cornell and an active investor in AI startups, I’ve been studying how AI reshapes our digital habits. Today, I’m launching a survey to better understand how users and companies are adjusting to this shift. Who leads and who lags in the ecosystems?
Let’s unpack this transformation of … everything.
The End Of Chrome And The Beginning AI Agents
Quite early in the LLM hype cycle, I said that: AGI is still a long way off. Large Language Models (LLMs), however, are a new user interface. This change will create many new opportunities for brands and micro-brands aggregated by a few giants. And yes, OpenAI will be one of them.
OpenAI’s ambition to become the universal interface to what we knew as the Internet isn’t surprising. If anything, they might be late. Remember Perplexity’s “Shop Like a Pro”? Their goal was to handle all of shopping—from research to payment. Now OpenAI is extending that idea from Perplexity to … yes … to everything.
The interesting part is that it might be feasible. During my user research, one user told me he uploaded his blood test to ChatGPT. It analyzed the results, suggested supplements, and suggested a web-store. The user went on and spent over $200—at a retailer he had never visited before, because “ChatGPT said so”. This isn’t a fringe case. And it shows a few areas where ChatGPT is already extending their reach:
• OpenAI just became his personal health record — trusted with sensitive data.
• It just became his doctor — providing medical recommendations.
• It just replaced Google — suggesting where to buy.
• It also guided the purchase — skipping traditional product pages.
The Everyone’s AI Agent
Sam Altman’s vision is clear. He wants to be the agent for healthcare, for finance, for e-commerce and many other areas. I hear that Sam invited many startups to ‘private dinner’ events at his personal maison in San Francisco. Some of the founders described his vision as “he wants to forward integrate into our business” or “he wants to be Everyone’s Agent” — much like Amazon once wanted to be The Everything Store.
But What About Companies? Or Ecosystems?
OpenAI has a real chance of realizing this vision. We see high speed in how users adopt. The question is whether organizations or ecosystems can match this speed. If not, OpenAI has a winning chance.
We’ve seen this before. The early internet changed how we consume media, but it took media ecosystems years to adapt. Today, AI is driving a similar transformation — and again, most companies and ecosystems are behind.
Many Industries Will Be Impacted By AI Agents
That’s why I’m doing this research. I want to find out how far users, organizations, and the ecosystem have evolved. I have already looked at many different areas:
• IDE — or how we code — has been changed. See my conversation with a16z’s Martin Casado. User behavior is driving the industry.
• FINANCE — or how we underwrite risk — has been changed. See my discussion with Varo Bank founder Colin Walsh. Here the industry is driving user behavior.
• E-Commerce — AI is changing how we sell and shop. It’s not yet clear who is driving whom here…
In E-Commerce It Is Not Yet Clear How AI Agents Will Win
I’ve chosen to focus on e-commerce for this study because the power distribution is not yet clear:
1. Users prefer answer-first experiences — powered by ChatGPT, Gemini, and Perplexity.
2. Users now bypass the homepage entirely, leading to a decline of the website.
On the other hand:
3. Brands push GEO (Generative Engine Optimization) — the art of influencing LLMs to list the brand.
4. Brands create ChatGPT-like experiences on site search.
LLMs like OpenAI and Perplexity:
5. Want to integrate agent capabilities (even though there are still many technical questions unanswered)
6. And the ability to place ads
Why I Need Your Help
To track how users vs. how companies are adapting, I partnered with Chase Binnie from RetailWire — you may remember our Shoptalk discussion.
Together, we launched a survey focused on GEO. It’s short. It’s targeted. The data is strictly used for research. It will help us all understand:
• How brands show up (or don’t) in AI-generated answers
• What strategies actually move the needle
• What tooling or support is still missing
Whether you’re a marketer, founder, or product lead — your perspective will help. We’ll publish all findings openly, and create a public playbook for working in the new AI-powered discovery ecosystem. Please take the survey here.