Why I Didn't Sign the AI Open Letter

Instead, I edited it.

My colleague, co-author, co-founder, and great friend Erik Brynjolfsson reached out to me last week to see if I wanted to sign “We Must Act Now,” the “statement on AI’s transformation of the economy” he helped organize. And I wanted to want to, because it’s an obviously important topic. And because I respect Erik’s thinking on it so much.

Erik and I have been obsessed with interested in the topic for 15 years: writing about it, thinking about it, arguing about it, and talking about it publicly and privately, alone and together, all around the world. It has been a great ride, and I can’t imagine a better partner for it. Erik is crazy smart and well-informed, open to new ideas and persuadable by evidence, and as his recent profile in The Atlantic points out, a really good human being. If there’s a bad bone in his body, I haven’t come across it.

So I read the statement with keen interest, which quickly turned into feelings of disappointment and frustration. Let me walk you through how those feelings mounted over the course of We Must Act Now’s three admirably short paragraphs. The text of the statement is bolded below, followed by my internal monologue as I read it:

1. AI may become radically more powerful over the next 10 years.

Yep! No problem at all here. In fact, I’d probably use a stronger verb than “may”.

2. This could drive an unprecedented transformation of our economy, larger than the Industrial Revolution, but unfolding over a vastly shorter time frame…

I agree with the first statement, while again finding the verb too timid. It is pretty clear that AI is a general-purpose technology, which means that the transformation is somewhere between likely and certain.

That final claim, meanwhile, is really strong. What’s a vastly shorter time scale — years instead of decades? Is AI going to transform the economy in an unprecedented way by 2030? I truly don’t think so.

… It could bring risks, including large-scale job displacement, as well as opportunities such as major gains in living standards.

Dammit.

Why lead with the bad stuff here? Doesn’t AI already have enough of a PR problem? Aren’t politicians from both sides of the aisle, in a rare show of bipartisanship, increasingly targeting AI? Is coverage of AI in elite publications like The New York Times not negative enough for the authors of “We Must Act Now?”

As a group, economists strongly believe that technological progress, like free trade, is a good thing (both clearly raise standards of living, for example). But I heard Alan Blinder ruefully admit several years ago that economists had lost the public debate about the benefits of free trade. Do they really want to lose the debate about technological progress too? Do they want to contribute to that loss?

I’m surprised, and not in a good way, that this paragraph didn’t even gesture toward the conventional economic wisdom that the benefits of tech progress outweigh its drawbacks.

3. Economists, policymakers and technology leaders must act now to understand the economics of transformative AI…

Hell yeah! We have to understand this stuff. Erik and I believe this so deeply that we cofounded Workhelix, a company that helps organizations understand how they’re being transformed via AI.

…and to build the incentives, guardrails, and institutions needed to steer AI in a direction that complements humans and benefits society.

Noooooooooooooooooooooooooo!

Reading these final words I felt like a World Cup competitor watching their teammate blast a shot on goal late in the match to equalize the score and cap a great comeback, only to see the ball ricochet off the crossbar.

We almost had it. We were so close.

I got disappointed because this is a statement — a clear one — of the philosophy of upstream governance. Even though we don’t know what AI’s impacts are going to be, “we must act now” to build new and/or more incentives, guardrails, and institutions to deal with them.

But maybe those impacts won’t be so big and bad. As Erik and his colleagues point out, there are “canaries in the coal mine” about labor force displacement due to AI. But there’s also evidence in the other direction: Software job postings are up lately, unemployment is pretty low for younger workers, real US median income continues to rise in the AI era, AI-adopting companies are adding workers faster than their low-adopting peers, and so on.

Or maybe our existing incentives, guardrails, and institutions will be enough to deal with them.

Or maybe some of our interventions will be bad ideas. It’s been known to happen. Remember the call in 2023 to “pause giant AI experiments?” It attracted more than 2000 initial signatures. I wonder how many of them now think that pause would have been a good idea.

There’s also more than a whiff of dirigisme in those final words: “economists, policymakers, and technology leaders must act now… to steer AI.” (emphasis added) I’m surprised by this. Economists and business leaders tend to be more of a laissez-faire crowd.

Editor to the Letter

Okay, it’s easy to nitpick and criticize. Can I do better at writing a short letter that captures the importance of this moment in the AI era? You be the judge. Below are my edits to We Must Act Now. I stuck with that letter’s three-paragraph structure (flat-out copying the first one!) and even saved a couple words.

My edits to the statement are textually minor. I feel like it came so close, but bounced off the crossbar and didn’t help our side — those who believe in the twin power of tech progress and Econ 101 — win.

I hope the below resonates with people who believe that even as powerful as AI is, it doesn’t yet require us to turn away from permissionless innovation and economic freedom and embrace higher levels of upstream governance and dirigisme.

1. AI is likely to become radically more powerful over the next 10 years.

2. Like previous world-changing technologies, AI will bring major gains in living standards. But it will also bring new risks, harms, and disruptions. And because of its extraordinarily fast improvement, AI’s benefits and shocks might come quickly.

3. So economists, policymakers and technology leaders must act now to understand the economics of transformative AI, and to build the capabilities needed to respond quickly and effectively to the challenges it will bring.

Is my letter closer to a statement of your beliefs, or are you with the signatories of We Must Act Now? Let’s talk about it