The diagnosis behind the AI backlash
The exchange started with the investor Gavin Baker arguing that Amodei's messaging skews negative. Amodei disputes the premise on its own terms: risks and benefits have been roughly balanced across his writing, and Machines of Loving Grace exists because he thought the industry was underselling the upside.
One clarification before going further, because the coverage blurs it. Both posts are a reply to Baker, not to Meta. Zuckerberg's essay is a separate text published five days earlier, and the two men have not addressed each other directly. What follows is a comparison of two documents, not a transcript of an argument.
Back to the reply. Amodei concedes the part that matters and refuses the causal story attached to it. Public opinion on AI is negative, and that is a serious problem. What he does not accept is that he or any other AI leader talking about risk is the main reason. The alternative explanation is institutional: people do not trust companies, governments, or the tech industry, and assume any new thing is a new way to be taken advantage of. AI inherited that.
He then names the criticism he thinks is the accurate one, and it is aimed at his own side. The strongest charge against AI companies including Anthropic, in the view of one of their CEOs, is that they have not yet delivered on the promises they made — and the fix for that is delivery, not narrative.
"I do not agree that my messaging has been disproportionately negative. In fact it has been about equally balanced between risks and benefits: I’ve written one major essay about each"…／"In fact, I wrote Machines of Loving Grace because I didn’t feel the AI industry was painting an inspiring enough picture of how the technology could radically transform the world for the better."／"I do agree that the public has a negative view of AI (and that this is a big problem), but I don’t think it is primarily caused by me or any other AI leader warning about AI’s risks. I think it is fundamentally a crisis of trust."／"I think by far the most accurate criticism of AI companies including Anthropic is that we haven’t yet delivered on our big promises to benefit the world. That is totally on us, and I think it’s the criticism you should be making, instead of all this stuff about messaging and marketing."— from Amodei's post on X
Regulation or open weights: where the split is
Set the two positions next to each other and the shared premise appears immediately.
| Zuckerberg | Amodei | |
|---|---|---|
| Core danger | Superintelligence centralized in a few institutions | AI concentrating power structurally |
| Remedy | Put the technology in everyone's hands | Not open weights alone — they shift concentration to whoever has compute and chips |
| On regulation | Necessary, but restricting capabilities is the centralizing path | Baker's either/or framing is a false choice |
Zuckerberg's essay opens on the question of who gets access to superintelligence and whether it stays restricted to a few institutions, and builds a three-part philosophy on individual empowerment, invention, and balance of power. On policy it is not anti-regulation: it calls government policy necessary and asks for close cooperation between frontier labs and government. What it treats as the centralizing move is restricting capabilities. It also turns a question back on the risk-focused camp: if you believe AI will take most jobs and much of humanity's relevance, why hurry to build it?
Amodei, replying to Baker rather than to Meta, attacks a framing rather than a set of values. Choosing between regulatory concentration and wide distribution is, in his words, a false choice — and outside Silicon Valley, regulation is commonly read as the thing that constrains corporate power on ordinary people's behalf. For the live version of that argument in Congress, see the AI kill switch bill and the fight over who pays for data center power.
"The defining questions of our age are who will have access to superintelligence and what will we direct it towards. Will it be centralized and restricted to a few institutions, or will it be a tool that empowers everyone?"／"We propose a philosophy based on individual empowerment as the source of prosperity, invention as the primary purpose of superintelligence, and balance of power as the foundation of safety."／"I do not understand why anyone who believes that AI will eliminate most jobs and much of humanity’s relevance would rush to build that future."／"Restricting capabilities leads us down the path of centralization and lack of checks and balances, so we should be extremely careful about this — especially if other nations pursue less restrictive paths."／"Government policy is necessary to ensure a positive future and mitigating the risks discussed above."／"We should be skeptical of any proposals that lead to centralizing superintelligence, and we should favor policies that promote a healthy balance of power that favors people."— from Meta's published essay
"First, on regulation, I think that “either concentrate it in the hands of a chosen few companies and politicians via regulation or distribute it widely” is a false choice."／"Many people outside this bubble think of regulation as something that constrains corporate power and benefits ordinary people."／"Overall my view is that AI is *structurally* a technology that tends to concentrate power, for reasons that have nothing to do with regulation (more to do with the extreme implications of the scaling laws)."— from Amodei's post on X
Both texts are long and both got flattened into single quotable lines within a day. Pulling the essay and the two posts into markdown side by side is the quickest way to see how much of the disagreement is actually about remedy rather than premise.
Why the optimistic case did not land
The reaction to Zuckerberg's essay went after the author rather than the argument. On TechCrunch's Equity podcast the hosts described it as missing the mark for a lot of people, and the reason given was history: the same person promised that social media would give everyone a place to talk to their friends, and what arrived was ragebaiting and advertising. One host read the essay as positioning its author as an anti-Dario.
This is Amodei's thesis arriving as evidence. If the problem is that promises have not been delivered, then a new promise is the least persuasive thing available, and it will be measured against the last one.
"Meta CEO Mark Zuckerberg published a 6,500 word essay this week declaring that “The Future is for Everyone” and painting an optimistic picture of a future powered by AI…"／"It feels like he’s positioning himself as almost like an anti-Dario." (Rebecca Bellan)／"You look back to the social media days, he was saying that he wants to make sure that everyone has an outlet for talking to their friends and having a social network. And what do we have instead? We have ragebaiting and advertisements, and not connection." (Rebecca Bellan)／"Yeah, I think that the manifesto didn’t strike the right note for a lot of people." (Kirsten Korosec)— from TechCrunch's report
What follows from the diagnosis
Treating the backlash as a communication failure implies one program: better essays, better campaigns, a better story. Treating it as a trust failure implies another: ship things that visibly work and let the record accumulate. The two executives are running the first and the second respectively, and only one of them conceded that his own industry has not yet earned the benefit of the doubt.
