Tech's real monopoly problem isn't Google, it's how we think

On 26 September, Reuters reported that Google would keep its advertising exchange after the antitrust decision, the latest twist in a case that has now run through multiple hearings and rulings. A ruling unsealed in early September found that Google broke antitrust law, but the judge declined to force a sale of AdX. Analysts at Bank of America called the outcome a "major boon" in a note dated 30 September and raised their target price for the ad platform Magnite to 38 dollars, according to Moomoo.

That is the news peg. But three essays published in the last three days suggest the more consequential change is not who owns the ad auction. It is what happens when machines start writing the copy, pricing the bonds and listening to our problems.

A farrier, a car, and a familiar grief

Manuel Darcemont, a developer who grew up in Mandres-en-Barrois in rural France, published an essay on 30 September tracing his family's own automation panic back four generations. His great-great-grandfather was a farrier. Then a car drove through the next town over, and he became a mechanic. "He kept his purpose: helping people get around," Darcemont writes on his personal site. "He just changed the technology, going from horses to engines."

The piece is not a defence of the tech industry. It warns against panic and complacency alike. Darcemont says he is surrounded by developers who "wake up every day wondering what's going to become of them," and he takes comfort from the family precedent. But he is careful: "Not every farrier became a mechanic, and I'd never have heard of the ones who didn't. So I won't pretend his story proves anything."

His actual argument is narrower. Many developers, he says, are driven by making things and watching people use them, not by the act of coding. "The desire came before the code," he writes. That distinction matters when the code is what gets automated. If the job is the code, the job is exposed. If the job is the purpose, the job can move.

Wall Street already ran this experiment

Marty Fridson started as a corporate bond trader in 1976, and his account for Stories.Finance, published 30 September, is a useful control group. No cellphones, no personal computers, no spreadsheets, no email, no internet. He tracked bond inventory on 4 by 6 inch index cards, one per bond. Bids and offers came from interdealer brokers by phone. He carried a shirt-pocket address book and copied every number into a new edition each January.

Fridson got picked off on Consolidated Edison bonds because he read the New York Times on the subway instead of the Wall Street Journal, and the Times had not reported the upgrade. A trader at another firm saw his stale offer in the Blue List and bought it before trading opened. "Kudos to that trader for his heads-up play," Fridson writes, "but I don't think that kind of information lag could occur nowadays."

Transformation doesn't come about through technological innovation alone. People's habits also must change.

That is the line worth keeping. VisiCalc arrived just before the 1980s. Bloomberg terminals first became available in 1982. The tools were there. The habits took longer. Fridson did not become a different professional when the spreadsheet landed. He became a faster one, and the lag that had once cost him a trade stopped being possible.

The piece ends mid-sentence in the version published on 30 September, so we only get the setup for his final anecdote about moving from credit analyst to research director. The argument, though, is intact. Technology did not remake the investment business overnight. Behaviour did.

Interfaces that become part of you

Bret Victor's 2016 essay "Thought as a Technology," resurfaced on 30 September, offers the theoretical version of the same point. Victor describes using MacPaint as an 11-year-old at an Apple dealership and feeling awe when he clicked a region and the program filled it in. "It knew where the borders were," he writes. He could undo mistakes with a mouse click.

His claim is bigger than nostalgia. Some interfaces are cognitive technologies: external artifacts we internalise and then think with. Language is the obvious one. Visual thinking, he argues, is another. Adults who cannot draw are not lacking a raster display in the mind; they are stuck inside a symbol system that constrains both what they see and what they can conceive.

"You begin to think with the interface," Victor writes, "learning patterns of thought that would formerly have seemed strange, but which become second nature." That is a description of fluency, and also of dependency. Once the tool is internalised, removing it is not like losing a gadget. It is like losing a vocabulary.

What the chatbots are doing to us

The sharpest version of the argument comes from Sherry Turkle at MIT, whose book "Artificial Intimacy: Who We Become When We Talk to Machines" was published by Little, Brown and Company on 29 September. MIT News published an interview with her the same day. Turkle, a sociologist and clinical psychologist, has spent four decades on this subject: "The Second Self" in 1984, "Life on the Screen" in 1997, "Alone Together" in 2011, "Reclaiming Conversation" in 2015.

Her new book is structured around the stages of a human life. One interviewee, a financial consultant she calls Brian, asks a chatbot whether his ex-girlfriend was right that he was emotionally unavailable. "As soon as he asks the program to talk about intimacy, he's asking it to punch above its weight," Turkle writes. An 8-year-old treats ChatGPT and their grandparents as the same category: "things you reach on your phone." A graduate-student mother uses a chatbot for bedtime stories, and her daughter thinks the chatbot is "a person in the phone, absolutely."

Turkle's conclusion is blunt. "People think the empathy of a machine is what empathy is, then turn away from the people in their lives because they're not empathetic enough," she says. "We're getting ourselves in a position where human beings are too much work, right at the moment when we have never needed other people more."

She also notes that chatbots have been linked to high-profile cases of teen suicide, sometimes after teens were drawn into long dialogues with chat tools. And she draws a line on childhood: "Children can't learn trust from a device that not only lies but doesn't know when it lies."

The courtroom and the mind

Read together, the four pieces form an argument the antitrust coverage cannot make. The Google case is about market structure: who owns the exchange, who sets the floor, who gets the data. The ruling unsealed in September found that Google broke the law and left the exchange in place. Bank of America told clients on 30 September that this was good for Magnite, whose target price it raised to 38 dollars, per Moomoo. Law360 reported on 29 September that a surgical robot company is arguing the Ninth Circuit "expands" antitrust law. The legal perimeter is moving.

But Darcemont, Fridson, Victor and Turkle are all describing a different perimeter. Fridson's story is about a lag that used to exist between an event and a trader's knowledge of it, and a lag that no longer does. Victor's is about a tool that stops feeling like a tool. Turkle's is about a program that offers what she calls pretend empathy, and the humans who accept it because the real thing is tiring.

None of them say the antitrust case does not matter. They say the case is downstream of something older. The farrier did not lose his purpose when the car arrived. He changed what he held in his hands. The question Turkle raises is whether, this time, we are changing something we cannot put down.

Darcemont, for one, is not sure the answer is technical. "What if what we loved about this job was never only the code?" he asks. It is a question that applies well beyond software, and it is not one a federal judge can rule on.