> ## Content Index
> Fetch the complete content index at: https://www.betteratcoding.com/llms.txt
> Use this file to discover other available public pages before exploring further.

# Dario says AI’s problem isn’t doomer talk. It’s that the industry hasn’t delivered.
- URL: https://www.betteratcoding.com/trending-news/dario-says-ais-problem-isnt-doomer-talk-its-that-the-industry-hasnt-delivered/
- Published: 2026-09-01T02:56:28.000Z
- Updated: 2026-09-01T02:56:28.000Z
- Author: Zacarias Ripoll Cid
- Tags: trending-news

> (Anthropic CEO says AI backlash is ‘fundamentally a crisis of trust’)

TechCrunch reporter Anthony Ha reported on August 16, 2026, that Anthropic CEO Dario Amodei and investor Gavin Baker had a public argument. Amodei’s warnings about AI risk have helped fuel a backlash in the US, especially against data centers, said Baker. He said Amodei “has lost the argument” on AI regulation, and he “should try harder to be a more positive advocate for his own industry.”

In posts on X dated Aug 15, 2026, Amodei replied. He rejected the idea that his messaging has been “disproportionately negative.” It's "about equally balanced between risks and benefits," he said. He wrote the essay “Machines of Loving Grace” because the industry wasn’t offering an inspiring enough vision.

He agrees that the public has a negative view of AI. “That’s a big problem,” he says. He rejects the idea that the negativity is “largely driven” by leaders warning of the risks.

“I think it’s a crisis of trust, fundamentally."

### What the "crisis of trust" means

The crisis of trust is simpler than it seems. It is not a claim that people have been talked into fearing. It is a claim that people no longer believe the people in charge.

“Ordinary people don’t trust companies, governments, or the tech industry and always suspect that we are cooking up some new way to screw them over,” Amodei said. If that’s how the public feels, a more cheerful press tour won’t fix it. And a darker warning will not fix it. People want tangible evidence.

That’s why Amodei’s criticism was aimed at results, not at his tone. “I think the most accurate criticism of AI companies including Anthropic is that we haven’t yet delivered on our big promises to benefit the world.” AI will cure cancer is “more a cliche than it is inspiring,” he said. What would change opinions, he said, is “actually curing cancer.

Now if you are not an investor and not a CEO, this is the part that will help you. You don’t have to decide if Amodei talks too much about risk. You can use his own test: has the product shown up yet or is it still a speech?

## The regulation argument, plain English

Baker’s charge is that the warnings have made regulation tougher for the industry and that Amodei should sell the upside more. Amodei pushed back against a common Silicon Valley shorthand: regulation means capture. Capture is the idea that big companies help write the rules so that the rules protect them.

He says Anthropic tries to propose rules that slow down frontier labs — the small set of companies that are building the most powerful models — while helping smaller competitors. He also says AI “structurally” centralizes power. That means the technology itself tends to stack advantage on whoever already has the most compute and chips.

## What “open weights” mean

Open weights means a lab releases the model learned parameters. The large file of numbers that is the trained model. Other people can use that file to run the model themselves, study it, or build on it. They don’t have to submit every prompt to a private company API.

That sounds like a clean way to distribute power. “It helps, but it’s not enough,” Amodei said. Open weights “help some,” he wrote, but “are nowhere near a sufficient solution,” because they shift the focus to whoever has the most compute and chips. The model is available for download. Still you need a huge amount of computing power to train a bigger one, or to run it for millions of users.

His ideal “rules of the road” would serve three purposes at once. They would (a) manage cyber, bio, and alignment risks, (b) tighten the leash on frontier companies, and (c) allow room for open weights but also manage the risks those models carry.

TechCrunch later updated the article to include more of Amodei’s quotes, including the ones about open weights.

## How it helps ordinary people

You already do a version of this every time you use a chatbot. The tool generates an email draft and you still need to rewrite the weird sentence. It indicates a function, and you still run the tests. That is not cynic. That’s what trust looks like when it needs to be earned in small pieces.

Keep that up. And use the tools you already have to save yourself time. Never mind the slogan that AI will soon fix everything. If a product helps you write a clearer note, debug a stuck script, or explain a concept to a coworker, that's a real benefit. If it’s talking about curing cancer, that’s a pitch.

Amodei’s measure is delivery itself. That standard applies to the rest of us as well. You can judge AI by what it actually does for you this week, not by a CEO’s essay.