Dave Eggers on the machine that thinks for you: 'you're cooked as a species'

At Dave Eggers's suggestion, the interview begins on a boat. The American novelist has for years written much of his work aboard a small vessel moored on San Francisco Bay, and on 27 June 2026, sitting somewhere on the water, he used the occasion of his new novel's publication to argue that the most consequential cultural technology of the century is not a platform, a streaming service, or a phone — it is the software that now drafts sentences on a writer's behalf. "Once you have a machine think and write for you," Eggers told the Guardian, "you're cooked as a species."

The remark, blunt by literary standards, lands at the centre of a debate that has migrated from laboratories and venture-capital pitch decks into the working lives of editors, novelists, journalists and screenwriters. Eggers is not the first writer to sound the alarm. He is, however, one of the most public-facing American novelists to frame the issue as a generational handover rather than a tooling upgrade.

A new novel, an old argument

The interview is pegged to a book launch. Eggers uses the platform to argue that the next generation of writers — schoolchildren, MFA students, the teenagers he teaches in literacy programmes — need active defence against a default in which keystrokes are outsourced to a model trained on everyone else's prose. The worry is not that machines produce passable text. Plenty of them do, and the marketplace has noticed. The worry, in Eggers's telling, is what happens to a writer's interior life once the drafting muscle is rented out.

He frames it as a question of apprenticeship. Writing, he suggests, is a craft learned by repetition, failure and the slow accumulation of a private ear — the voice that tells you, on the fifteenth revision, that a sentence is finally honest. A machine can short-circuit the repetition. It cannot, he insists, install the ear.

The argument is at least partly biographical. Eggers built his career on independent publishing — McSweeney's, the quarterly he founded in 1998, and the associated nonprofit 826 Valencia, which teaches writing to school-age children in the United States and abroad. For nearly three decades the throughline has been the same: build the scaffolding that lets young people spend time with their own sentences. Generative software, he suggests, is the first technology that threatens to remove the scaffolding rather than extend it.

The Sam Altman exchange

Eggers does not confine himself to literary hand-wringing. He told the Guardian he has debated Sam Altman, the chief executive of OpenAI, the US laboratory behind the large language models that now populate consumer products from email clients to office suites. The substance of those debates, as paraphrased in the interview, runs along familiar lines: technologists argue that the models augment rather than replace writers; Eggers argues that augmentation is the trojan horse.

The asymmetry is real and rarely named in trade coverage. A model company measures success in queries per day, active users and enterprise seats. A novelist measures success in something far less legible: the residual quality of a paragraph after the writer has stopped touching it. The two metrics do not collide head-on, which is precisely why the cultural conversation has drifted. The economic case for generative prose is straightforward. The cultural case against it — what a society loses when the average citizen stops writing — is harder to put on a slide.

Eggers's contribution is to insist the cultural case is the only one that matters. He frames the present moment as one in which the act of writing is being reclassified, in real time, from a civic skill to a discretionary output. Once that reclassification is locked in by default settings in word processors, browsers and classroom software, the work of rebuilding it is the work of a generation.

What the counter-narrative looks like

Not everyone in the literary world agrees that the alarm is well-placed. Pro-tool writers — a small but vocal contingent — argue that prose assistants are the latest iteration of a long lineage: the typewriter, the word processor, the grammar checker, the predictive keyboard. Each provoked a hand-wringing essay. Each, in time, became invisible. The job of the writer, in this telling, was never the keystroke; it was the judgment.

There is a steelier version of the same argument that comes from the open-source and independent-model communities, where developers argue that the cultural critique is selectively aimed at large US labs while smaller, openly licensed models — many of them trained on permissively licensed corpora and released with weights that anyone can inspect — are folded into the same panic. On that reading, what is under threat is not the writer but a particular concentration of capital.

Eggers's response, implicit in his framing, is that the concentration and the cultural effect cannot be separated. If the dominant interfaces through which a generation encounters prose are owned by a handful of firms, then the question of who trains them, on what, and with what guardrails is not a technical footnote. It is the editorial question of the age.

The stakes, plainly

The cultural economy of writing is not large in the gross sense — novelists, journalists, screenwriters and copywriters together account for a modest share of GDP. But writing is upstream of almost every other cultural artefact. The screenplay feeds the film. The pitch deck feeds the company. The investigative report feeds the policy debate. The court filing feeds the ruling. The letter to a child feeds the next letter.

If the default citizen of 2040 writes less, and reads more machine-flushed prose, the consequences show up first in places that are easy to ignore: the deterioration of persuasive public letters, the flattening of journalism's voice, the loss of the private craft of correspondence. They show up later in places that are impossible to ignore: a public sphere in which the only people still writing in their own voice are the people paid to do so.

Eggers, for his part, is doing what writers in his position have always done when a new technology reorganises the trade: he is teaching, publishing and pointing at the next generation. The boat, the new novel and the interview are not a manifesto. They are an invitation to take the apprenticeship seriously while there is still time to arrange one.

Desk note: Monexus framed this around Eggers's own argument — apprenticeship, voice and the reclassification of writing — rather than around the OpenAI product cycle. The wire coverage treated the interview chiefly as a celebrity profile pegged to a book launch. The structural question, this publication's reading, is which cultural institutions step into the gap if default software does the drafting.

This article is based on a single source interview published on 27 June 2026. Claims about Eggers's career, McSweeney's, 826 Valencia and his reported debate with Sam Altman are drawn from that interview. No other wire reports of the conversation were available at the time of publication; the framing here relies solely on the Guardian piece cited below.