Silicon Valley is pouring billions into AI agents that shop on your behalf. Ron Johnson, the man who built Apple’s retail stores alongside Steve Jobs, thinks the industry has the scale of that shift wrong. In an interview with TechCrunch AI, the 66-year-old retail veteran argued that AI will make online shopping better but won’t change where or how people actually buy.
“AI is a new technology that will improve the online shopping experience,” Johnson told TechCrunch AI. “But I don’t know that it’s going to change which way we shop.”
That’s a direct challenge to the agentic commerce thesis, and it comes from someone with a track record of calling retail shifts correctly.
The bet he’s pushing back on
The biggest names in tech are building the plumbing for agents to handle shopping end to end. According to TechCrunch AI:
- Google is pushing its Universal Commerce Protocol, a standard meant to carry AI agents from product discovery all the way to checkout.
- OpenAI is turning ChatGPT into a shopping destination where users research, compare, and in some cases buy without leaving the chat.
The pitch is simple: tell an agent what you want, and it finds it, compares it, and pays for it. Less friction, fewer tabs, no store visit.
Johnson’s response when asked if someone would let an agent pick and buy a $1,000 or $2,000 laptop sight unseen: “Honestly, nobody’s going to do that.”
Why he thinks the physical store survives
His argument rests on one idea: some purchases are too personal to delegate. Buyers want to feel a laptop’s weight, see the display, and figure out which size fits their life. “AI will never be able to have you physically experience a product,” he said.
Where AI does fit, in his view, is upstream of the store. Agents narrow the options and educate the buyer. “They’ll just become more informed shoppers when they come to the store.”
What stands out here is that Johnson isn’t a skeptic of AI itself. “I’m a real believer in AI. I’m an AI optimist,” he told TechCrunch AI. He’s skeptical of a specific claim: that the last step, the decision and the transaction, moves to a machine for high-consideration goods.
Why his opinion carries weight (and where it doesn’t)
Johnson joined Apple in 2000, right as online shopping was taking off and analysts were writing obituaries for physical retail. He built a store network that became central to how Apple sells and supports its products. Competitors copied the glass, the open floors, and the Genius Bar, but as he put it, “The secret sauce for Apple has always been its people.” No commission, no pressure, just staff figuring out what the customer actually needs.
His record isn’t spotless, and he’s candid about it. His J.C. Penney turnaround in 2011 flopped within two years after sales collapsed. “I applied a startup mentality to what needed to be a turnaround transformation,” he said. His later startup Enjoy Technology, which delivered tech products and setup to customers’ homes, went bankrupt in 2022.
So take his prediction as a retail operator’s read, not a technologist’s. He knows what makes people trust a purchase. He’s been wrong about what makes a business model work.
What this means for the next two to three years
My take: both sides are partly right, and the split will happen by price and personality of the purchase.
- Low-consideration goods (consumables, refills, commodity electronics accessories) are heading toward agent-driven reorders. Nobody needs to hold a USB cable first.
- High-consideration goods (laptops, phones, furniture, cars) will see agents own the research phase while humans keep the final call and often the physical touchpoint.
- The store’s job shifts from selling to confirming. Customers arrive knowing what they want. Staff who can handle a well-informed buyer become more valuable, not less.
For AI practitioners and businesses, a few practical moves:
- If you sell expensive or tactile products, optimize for agent discoverability now. The shortlist is being written by machines even if the purchase isn’t.
- Don’t strip humans out of the last mile of a big-ticket sale. Johnson’s no-commission model is worth studying because agents will make pushy sales tactics even less effective.
- If you’re building agentic commerce tools, target the categories where delegation feels safe. Winning the reorder is a real business. Winning the laptop purchase is a much harder one.
Johnson’s book, Shop Different, revisits the Apple retail bets in more detail. His core message on AI, though, fits in one line he attributes to Jobs’ way of working: “There’s no substitute for human intuition.” The agent revolution will test that claim over the next few years, one shopping cart at a time. You can read the full interview at TechCrunch AI.