Right. It probably needed to be general because there hasn't been enough time to work out sane specific responses, and even if they had, getting buy-in on specifics is a recipe for paralysis by indecision. A credible group of people simply pleading for policy makers, researchers, et. al. to take this seriously will lead to the project approvals, grant money, etc. that will hopefully yield a more sophisticated understanding of these issues.

Cynicism is understandable in this ever-expanding whirlpool of bullshit, but when something looks like it has potential, we need to vigorously interrogate our cynicism if we're to stand a chance at fighting it.

Reading the comments here is helping evolve my thinking on the issue for sure. Here's a comment I made in another thread:

> As I mentioned in another comment, the listed risks are also notable because they largely omit economic risk. Something that will be especially acutely felt by those being laid off in favor of AI substitutes. I would argue that 30% unemployment is at least as much of a risk to the stability of society as AI generated misinformation.

> If one were particularly cynical, one could say that this is an attempt to frame AI risk in a manner that still allows AI companies to capture all the economic benefits of AI technology without consideration for those displaced by AI.

If policymaker's understanding of AI is predicated on hypothetical scenarios like "Weaponization" or "Power-Seeking Behavior" and not on concrete economic disruptions that AI will be causing very soon, the policy they come up with will be inadequate. Thus I'm frustrated with the framing of the issue that safe.ai is presenting because it is a distraction from the very real societal consequences of automating labor to the extent that will soon be possible.

My own bit of cynicism is that regulating the negative impacts of technology on workforce segments in the US is a non-starter if you approach it from the technology-end of the issue rather than the social safety net end. Most of these automation waves that plunged entire employment categories and large metropolitan areas into oblivion were a net gain for the economy even if it was concentrated at the top. I think the government will temporarily socialize the costs of the corporate profit with stimulus payments, extended unemployment benefits, and any other thing they can do to hold people over until there's a comparatively small risk of triggering real social change. Then they just blame it on the individuals.

Agreed. To pile on: one true existential risk is the continued economic/social polarization of the US and other large nation states (not to mention climate changes).

We seem perfectly resigned to the expansion of an already huge underclass—-with or without the help of LLMs.

Might not AGI help convince our tech savvy aristocracy that one fundamental problem still is better balancing of opportunities and more equitable access to a good education? I see the probability of that happening as precisely 4.6%.

> will lead to the project approvals, grant money, etc.

In other words, a potential conflict of interest for someone seeking tenure?