On July 13, 2026, Stanford's Digital Economy Lab published an open letter about AI replacing jobs. It is 88 words long. Four sentences. More than 200 economists and AI researchers signed it, including 16 Nobel laureates: Daron Acemoglu, Joseph Stiglitz, Paul Krugman, Ben Bernanke. People from inside the industry signed too. Jack Clark, who co-founded Anthropic. Sarah Friar, OpenAI's CFO. Jeff Dean at Google.
The sentence everyone is quoting: AI could drive a transformation of the economy "larger than the Industrial Revolution, but unfolding over a vastly shorter time frame." The risk they name first is large-scale job displacement.
I make films for a living and I use AI on real productions every week. I read the letter twice. Here is what I think it says, and what I think it does not say.
What the letter actually says
Four sentences do not contain a policy program. The letter makes one claim and one request. The claim: AI may become radically more powerful over the next ten years, and the economic consequences will arrive faster than institutions usually move. The request: leaders should build the incentives, guardrails, and institutions needed to steer AI in a direction that "complements humans and benefits society."
The word worth stopping on is complements. Not replaces. Not automates. Complements. Two hundred economists, several of whom spent the last decade pushing back on AI hype, chose that word on purpose. Acemoglu himself, who shared the 2024 Nobel and spent years arguing the displacement panic was overblown, signed a letter that treats the risk as real enough to demand action now.
The question the coverage is not asking: whose job is it to make AI complement people instead of replacing them? The letter says leaders. Governments, institutions, companies. That is half the answer. The other half sits with anyone who opens one of these tools tomorrow morning and decides what to ask of it.
Will AI replace creative jobs?
This is the question people actually type into a search bar at two in the morning, so let me answer it from the only place I can: a career spent on sets and in edit rooms.
AI replaces tasks. It does not replace judgment. I have been the editor who lost nights inside a timeline, and I can tell you the hard part was never the cutting. The hard part was knowing which take had the energy, which pause could hold, which beautiful shot had to die for the scene to live. The tools I use now generate options at a speed that would have seemed absurd five years ago. Not one of them has ever known which option was the right one. That decision still walks in with a person.
Here is the model I use to think about this, and it is the reason I never panicked. AI is a modern child with hyperfocus. It absorbed a ton of information, but it cannot pull something new and unplastic out of all that alone. Its hyperfocus is unfolding the problem you hand it. Hand it a poor idea and it will amplify garbage everywhere you look. Hand it something good and rich, and it amplifies a unique idea into unique work. Everything starts from the author.
That is not a comforting slogan. It is a job description. The work that survives this decade is the work of feeding the machine something worth amplifying.
The Industrial Revolution comparison, seen from a film set
Cinema is a child of the last industrial wave, so this industry has run the experiment before. Sound arrived in 1927 and ended careers in two years: title-card writers, house musicians, actors whose voices did not match their faces. It also created entire departments that had never existed. Sound design, dialogue editing, the mixing stage. The jobs did not vanish. The job descriptions did, and the people who crossed over were the ones who treated the new machine as an instrument instead of an enemy.
The letter's real warning is not the comparison. It is the clock. The Industrial Revolution unfolded across generations; a farmer's grandson became a factory worker. The economists are saying this one may compress into a single career. Same transformation, no generational buffer. That part I believe, because I have watched it happen inside my own craft in under three years.
The amplifier, not the replacement
So I will say the uncomfortable part plainly. Some companies will read this moment as an invitation to cut people and let the machine run. The letter, signed by the CFO of OpenAI among others, is telling you that even the people building these systems think that path needs guardrails.
My position is harsher. Not because the models will stop improving. They will not. Because a tool that amplifies whatever you feed it is worth exactly as much as the people feeding it. Replace those people and you have built a very fast photocopier of your own averages. Whoever still believes AI replaces people will end up replaced themselves, by better ideas, held by teams that kept their humans and armed them.
What a working creator should do this year
Nobody reading this controls institutional guardrails. Here is what is actually in reach:
- Get your hours on the tools inside real work, not tutorials. I wrote about how the decision, not the tool, stays with the director when George Lucas called AI inevitable. Inevitable tools still need someone who knows what to point them at.
- Keep the paper trail of your judgment: treatments, references, rejected versions. Institutions from Cannes to the copyright office are converging on one test, meaningful human involvement, and the evidence of your decisions is becoming part of the work itself.
- Invest in the one asset the machine cannot generate: taste built from lived experience. The gatekeeping around craft skills is falling, as I argued when design tools stopped requiring a design degree. What is left when the gate falls is the quality of what you have to say.
The clock is the message
The letter is four sentences because it is not trying to predict the future. It is trying to start the preparation before the future arrives. On that, the economists and this director agree completely. Prepare. Not by hiding from the tools, and not by worshiping them. By becoming the kind of author whose ideas are worth amplifying, and by insisting, in every room where you have a voice, that the machine exists to serve the people doing the work. The transformation is coming either way. Who it complements is still a decision. Decisions have owners.
