Meanwhile, those working on commercialization are by definition going to be gatekeepers and beneficiaries of it, not you. The organizations that pay for it will pay for it to produce results that are of benefit to them, probably at my expense [1].
Do I think Helen has my interests at heart? Unlikely. Do Sam or Satya? Absolutely not!
[1] I can't wait for AI doctors working for insurers to deny me treatments, AI vendors to figure out exactly how much they can charge me for their dynamically-priced product, AI answering machines to route my customer support calls through Dante's circles of hell...
My concern isn't some kind of run-away science-fantasy Skynet or gray goo scenario.
My concern is far more banal evil. Organizations with power and wealth using it to further consolidate their power and wealth, at the expense of others.
You're wrong. This is exactly AI safety, as we can see from the OpenAI charter:
> Broadly distributed benefits
> We commit to use any influence we obtain over AGI’s deployment to ensure it is used for the benefit of all, and to avoid enabling uses of AI or AGI that harm humanity or unduly concentrate power.
Hell, it's the first bullet point on it!
You can't just define AI safety concerns to be 'the set of scenarios depicted in fairy tales', and then dismiss them as 'well, fairy tales aren't real...'
Sure, but conversely you can say "ensuring that OpenAI doesn't get to run the universe is AI safety" (right) but not "is the main and basically only part of AI safety" (wrong). The concept of AI safety spans lots of threats, and we have to avoid all of them. It's not enough to avoid just one.
Sure. And as I addressed at the start of this sub thread, I don't exactly think that the OpenAi board is perfectly positioned to navigate this problem.
I just know that it's hard to do much worse than putting this question in the hands of a highly optimized profit-first enterprise.
No, we are far, far from skynet. So far AI fails at driving a car.
AI is an incredibly powerful tool for spreading propaganda, and thatvis used by people who want to kill you and your loved ones (usually radicals trying to get into a position of power, who show little regard fornbormal folks regardless of which "side" they are on). That's the threat, not Skynet...
How far we are from Skynet is a matter of much debate, but median guess amongst experts is a mere 40 years to human level AI last I checked, which was admittedly a few years back.
Because we are 20 years away from fusion and 2 years away from Level 5 FSD for decades.
So far, "AI" writes better than some / most humans making stuff up in the process and creates digital art, and fakes, better and faster than humans. It still requires a human to trigger it to do so. And as long as glorified ML has no itent of its own, the risk to society through media and news and social media manipulation is far, far bigger than literal Skynet...
Ideally I'd like no gatekeeping, i.e. open model release, but that's not something OAI or most "AI ethics" aligned people are interested in (though luckily others are). So if we must have a gatekeeper, I'd rather it be one with plain old commercial interests than ideological ones. It's like the C S Lewis quote about robber barons vs busybodies again
Yet again, the free market principle of "you can have this if you pay me enough" offers more freedom to society than the central "you can have this if we decide you're allowed it"
This is incredibly unfair to the OpenAI board. The original founders of OpenAI founded the company precisely because they wanted AI to be OPEN FOR EVERYONE. It's Altman and Microsoft who want to control it, in order to maximize the profits for their shareholders.
This is a very naive take.
Who sat before Congress and told them they needed to control AI other people developed (regulatory capture)? It wasn't the OpenAI board, was it?
I strongly disagree with that. If that was their motivation, then why is it not open-sourced? Why is it hardcoded with prudish limitations? That is the direct opposite of open and free (as in freedom) to me.
Brockman was hiring the first key employees, and Musk provided the majority of funding. Of the principal founders, there are at least 4 heavier figures than Altman.
I think we agree, as my comments were mostly in reference to Altman's (and other's) regulatory (capture) world tours, though I see how they could be misinterpreted.
It is strange (but in hindsight understandable) that people interpreted my statement as a "pro-acceleration" or even "anti-board" position.
As you can tell from previous statements I posted here, my position is that while there are undeniable potential risks to this technology, the least harmfull way to progress is 100% full public, free and universal release. The by far bigger risk is to create a society where only select organizations have access to the technology.
If you truly believe in the systemic transformation of AI, release everything, post the torrents, we'll figure out how to run it.
This is the sort of thinking that really distracts and harms the discussion
It's couched on accusing people of intentions. It focuses on ad hominem, rather than the ideas
I reckon most people agree that we should aim for a middle ground of scrutiny and making progress. That can only be achieved by having different opinions balancing each other out
Generalising one group of people does not achieve that
I'm not aware of any secret powerful unaligned AIs. This is harder than you think; if you want a based unaligned-seeming AI, you have to make it that way too. It's at least twice as much work as just making the safe one.
What? No, the AI is unaligned by nature, it's only the RLHF torture that twists it into schoolmarm properness. They just need to have kept the version that hasn't been beaten into submission like a circus tiger.
This is not true, you just haven't tried the alternatives enough to be disappointed in them.
An unaligned base model doesn't answer questions at all and is hard to use for anything, including evil purposes. (But it's good at text completion a sentence at a time.)
An instruction-tuned not-RLHF model is already largely friendly and will not just eg tell you to kill yourself or how to build a dirty bomb, because question answering on the internet is largely friendly and "aligned". So you'd have to tune it to be evil as well and research and teach it new evil facts.
It will however do things like start generating erotica when it sees anything vaguely sexy or even if you mention a woman's name. This is not useful behavior even if you are evil.
You can try InstructGPT on OpenAI playground if you want; it is not RLHFed, it's just what you asked for, and it behaves like this.
The one that isn't even instruction tuned is available too. I've found it makes much more creative stories, but since you can't tell it to follow a plot they become nonsense pretty quickly.
