Zosoforum

I seem to have left twitter in an inopportune time. Because now, seemingly everyone has become an AI expert the past few days taking a side in affairs they knew nothing about until maybe weeks ago.

You are seeing people who otherwise talk about completely different things now talking about AI Safety or Effective Altruism and their opinions are about what you'd expect. Complete propagandised garbage regurgitating what has been repeated endlessly

One thing I will say is that EAs and the general Safety cohort brought this upon themselves. They wanted more people to care, and now they will.
I don't think it was entirely a good strategy, mind. Effective Altruism is likely to get far bigger but also far less "itself" if that can be understood. It might get subsumed by other ideologies depending on where it ends up ultimately

It is funny to see the internal AI crowd try to appeal to outsiders with the outsider terminology. Lumpenspace (fat) once helped write a thread accusing an EA company of being behind the recent cyber attack and swarm incidents. It was completely incorrect, but the thread OP kept mentioning that the company was Israeli; obviously an attempt to taint the movement so that outsiders associate it with diddy blud netanyahu or something. Very funny.

'Effective altruism' soon to become the next 'empathy'. I hail the vacuous concept.

I can't believe I'm writing this, but I agree with Yudkowsky here; wanton want of politicisation of the EA cause, trying to put the issue alongside e.g. climate change in terms of near-term issue and political allegiance, has been destructive for their short-term outlook. Anthropic being declared as a "supply chain risk" by the DoD was the canary in the coal mine of this, but Anthropic getting involved with contracts to the federal government—especially the DoD—should've made predictions of this not too surprising in the first instance.

I was at a Right-DSA AI Regulation event the other day where they were of the view that it was always going to eventually bifurcate into a political issue(With Republicans taking the pro-business anti-regulation position) and I suspect they're correct. It's who they are as a party, so when handed a new issue they try to approach it with what has historically worked for them.

The EA/Safetyist types seem completely hamstrung by their own ineptitude to effectively translate their thoughts and opinions outside their natural bay area techy rationalist social bubble, for example making the term 'p(doom)' a core part of their lexicon rather than picking something immediately graspable to the average layman reading it. Add into this the grandiose self-importance they have and they seem to read every instance of friction with, or inability to reach the public as persecution by the uneducated and wicked.

The inability for the anti-GOP camp to formulate a positive alternative vision for AI than what Trump and crew are doing means that opposition is likely to settle around angry kneejerk luddite positions. The luddites and safetyists are going to squander a lot of time and energy bickering for the foreseeable future while the current uncontrolled pace of AI development continues without unified opposition.

America as a society seems to increasingly view AI as something on par with the arrival of a new world religion, its going to completely shatter the old order and unleash a new alien paradigm in its wake. If they think thats good, they want their own cliquey in-group to be its priesthood and make the God-machine in their own image, if they think its bad, they view it as basically the Antichrist imposing satanic totalitarianism on them.

>>7

It's weird because currently there is no way that AGI/human-level AI can be developed in near-term. It's just impossible with current architecture. They have no sentience, yet americans think it's gonna be terrifying scifi-level terminator AI.

People need to stop taking scifi stories as a source. Muh Chatgipiti is like ultronerino or something like that

>>3

"Israeli" has had an interesting semantic drift as of late, just means evil and nefarious regardless of any involvement of the actual state of Israel (see "spiritually Israeli" or "I'm an artist from Tel Aviv and this is my art" slander). Now I think the company in question was actually Israeli but nonetheless, not like they caused the recent incidents, that's mostly negligence on part of AI labs AND the affected companies