Every: what human work looks like after automation
Published in August 2026 as the introduction to Every’s Thesis Statements collection, Dan Shipper’s piece gathers predictions from 100 practitioners — founders, researchers, and writers — on what human work looks like when AI handles generation at scale. The collection is called “Thesis: 2027” and is designed to be testable against reality as the predictions age.
What the article is about
Shipper frames the collection as a response to a specific problem: AI discourse tends toward either wholesale adoption or blanket rejection, neither of which is useful for people trying to figure out how their work actually changes. The 100 predictions come from people who use frontier models daily in their actual work, not analysts writing about the technology from the outside.
Context
The predictions cover a wide range of roles and disciplines. Shipper selects and introduces contributions from company founders, researchers, and practitioners who have published their thinking about what AI changes about creative and knowledge work specifically. The collection is anchored to 2026-era capabilities — models that can draft, analyze, synthesize, and reason — rather than speculative future systems.
Key takeaway or method
Several recurring themes run through the predictions. Questions become more valuable as answers become abundant — a formulation that appears in multiple forms. Anne-Laure Le Cunff, founder of Ness Labs, states directly: “Answers will become abundant and questions will become the hard part.” For writers, this reframes the task. If AI can produce serviceable prose on any topic given a prompt, what is the writer’s contribution? The predictions lean toward: setting the frame, deciding what is worth writing about, and maintaining a perspective that is accountable to real experience rather than training data.
Karri Saarinen, CEO of Linear, argues that design — and by extension creative judgment — becomes AI’s primary challenge. The implication for writing is similar: the harder human contribution is not the execution of a format but the judgment about what should be said and how it should land for a specific reader.
Other contributors emphasize that unglamorous foundational work — accurate sourcing, clear reasoning, editorial accountability — may become more differentiating rather than less, because AI makes everything else cheaper and faster. The question of what is left for humans to do keeps arriving at the same answers: interrogation, curation, and accountability for what gets published.
Who it is useful for
Writers, editors, and content strategists thinking about how to position their work as AI takes over the production layer. The article is more useful as a framing document than a tactical guide — it does not describe specific workflows or tools. It is well-suited for teams trying to articulate what the human role in an AI-assisted content operation looks like, and why certain editorial capabilities remain valuable when drafting becomes inexpensive.