Mark Zuckerberg says every AI lab has the incentive to build safely

“Every lab has the responsibility and incentive to move at the pace required to train its models safely, and the ability to take its own actions to ensure that happens,” said Mark Zuckerberg in a post on X on Tuesday.

In the context of these days, when everyone is talking about “responsibility”, “the danger of AI”, and slowing things down, Meta’s chief executive answered Dario Amodei’s call for the industry to slow down and asked Washington for an antitrust waiver so that rivals could coordinate on safety.

Moreover, Sam Altman and Elon Musk were the first to publicly endorse the plan, while Palantir’s Alex Karp was less impressed, saying that a pause is unrealistic when China is unlikely to join one.

“Meta delayed shipping Muse for several months to focus on safety and security. We didn’t call for everyone else to do this before we would.

We just did it as part of our day-to-day work because it was clearly the right thing for people and for us. I’m proud of the security foundations we’ve built,” continued Mark Zuckerberg in the same post.

Muse was finally launched in the United States on 8 September, after being postponed from April. According to the same report, Meta has admitted to encountering reliability issues during internal testing, including instances where the agent retrieved personal photos it had not been asked to retrieve.

The White House does not seem to be very willing to accept a slowdown either. At his Doonbeg resort in Ireland on Monday, President Donald Trump described the supporters of the slowdown as “negative forces” and said the main concern was staying ahead of China.

OpenAI has stated that it will provide independent evaluators with access to its systems at the level of an employee, whereas in his post, Zuckerberg maintained Meta’s current arrangements and gave no indication that Meta intends to take part in the scheme that Amodei has outlined.

“I believe the key to building a positive future for everyone is maintaining the right balance of power. This is within our power to do,” closed Meta’s chief executive.

After a couple of days of reactions and warnings about a technology that is advancing too fast and could pose a danger to humanity, we still don’t have the big picture of what will happen and what this slowdown will look like for companies and people, as this “right balance of power” should already exist from the beginning.