Elon Musk is worried about advanced AI.

One thing he’s concerned about specifically is the potential for “contemporary” AI systems to “flood our information channels with propaganda and untruth.” That’s according to an open letter co-signed by Musk and published Wednesday to quite a bit of fanfare.

I assume readers don’t need my assistance when it comes to identifying the (somewhat tragic) irony in Musk lamenting a prospective “flood” of “propaganda and untruth” on “information channels.” And I anyway try to avoid abrasive political editorializing where it’s not necessary. So, I’ll refrain here other than to gently suggest that as things currently stand, there are bigger misinformation threats than AI, although I readily acknowledge the vast potential for AI to enhance the propaganda capabilities of malign actors, state-backed and otherwise.

With that out of way, Musk, co-signed by a who’s who of people who ostensibly know a lot about this (for lack of a more precise way to put it), think we’ve reached a dangerous precipice. Of course, all precipices are dangerous since you can fall off of them, but according to Musk, Yoshua Bengio, Andrew Yang, Steve Wozniak and, notably, Yuval Noah Harari, whose wildly popular books are replete with pseudo-warnings about AI’s potential, the race to “develop and deploy digital minds” is now “out of control” and poses a risk to the species.

“We call on all AI labs to immediately pause for at least six months the training of AI systems more powerful than GPT-4,” the letter said. “This pause should be public and verifiable, and include all key actors.”

Then, Musk and co. suggested that if industry players don’t voluntarily enact a moratorium posthaste, “governments should step in.” I don’t know about that, particularly when it’s coming from Musk, erstwhile Democrat-turned hard-right libertarian standard-bearer.

To call the recommendations in the short letter far-fetched in light of the sort of legislative ineptitude that defines US public policymaking would be to understate the case. Here are those recommendations:

AI developers must work with policymakers to dramatically accelerate development of robust AI governance systems. These should at a minimum include: New and capable regulatory authorities dedicated to AI; oversight and tracking of highly capable AI systems and large pools of computational capability; provenance and watermarking systems to help distinguish real from synthetic and to track model leaks; a robust auditing and certification ecosystem; liability for AI-caused harm; robust public funding for technical AI safety research; and well-resourced institutions for coping with the dramatic economic and political disruptions (especially to democracy) that AI will cause.

Do note: US lawmakers are still trying to sort out basic parameters for regulating last decade’s tech. The idea of those lawmakers preemptively regulating the next generation of technological advances and doing so efficiently and effectively would be laughable even in an environment where there was a sense of shared purpose — this isn’t such an environment. Musk knows that better than anyone.

Not that this is exclusively a US concern. Indeed, the whole point of the letter was that it’s a concern for all of humanity. But the sort of strict oversight the letter calls for will be implemented in China by dictatorial decree, at least as it relates to private sector AI development. And no luminaries hailing from democratic locales will convince Xi Jinping to slow down the Chinese government’s AI development efforts.

Israel has bigger concerns than AI currently, and those concerns are existential too. Paris is on fire — literally. The UK is still backsliding economically towards something that looks too much like an emerging market for comfort. And the rest of Europe is just hoping for “peace for our time.”

The point is: It’s not obvious who’s going to take the lead when it comes to governmental oversight of AI development. And even if it were, we’ve proven, as a species, wholly incapable of crafting globalized standards for matters of shared human concern. Just ask anyone (or any company) who’s succeeded in evading taxes or any nation currently skirting their environmental obligations.

The letter (which you can read in full here) is likely to garner more attention going forward, particularly given how much money Microsoft has already invested in OpenAI, and also given that authoritarian governments (which in this context basically just means China) would view any “pause” as an opportunity to catch up.

Musk and Bengio asked several additional questions: “Should we automate away all the jobs, including the fulfilling ones? Should we develop nonhuman minds that might eventually outnumber, outsmart, obsolete and replace us? Should we risk loss of control of our civilization?”

Maybe we should ask GPT-4. Or a T-800. (“Affirmative.”)