This weekly roundup thread is intended for all culture war posts. 'Culture war' is vaguely defined, but it basically means controversial issues that fall along set tribal lines. Arguments over culture war issues generate a lot of heat and little light, and few deeply entrenched people ever change their minds. This thread is for voicing opinions and analyzing the state of the discussion while trying to optimize for light over heat.
Optimistically, we think that engaging with people you disagree with is worth your time, and so is being nice! Pessimistically, there are many dynamics that can lead discussions on Culture War topics to become unproductive. There's a human tendency to divide along tribal lines, praising your ingroup and vilifying your outgroup - and if you think you find it easy to criticize your ingroup, then it may be that your outgroup is not who you think it is. Extremists with opposing positions can feed off each other, highlighting each other's worst points to justify their own angry rhetoric, which becomes in turn a new example of bad behavior for the other side to highlight.
We would like to avoid these negative dynamics. Accordingly, we ask that you do not use this thread for waging the Culture War. Examples of waging the Culture War:
-
Shaming.
-
Attempting to 'build consensus' or enforce ideological conformity.
-
Making sweeping generalizations to vilify a group you dislike.
-
Recruiting for a cause.
-
Posting links that could be summarized as 'Boo outgroup!' Basically, if your content is 'Can you believe what Those People did this week?' then you should either refrain from posting, or do some very patient work to contextualize and/or steel-man the relevant viewpoint.
In general, you should argue to understand, not to win. This thread is not territory to be claimed by one group or another; indeed, the aim is to have many different viewpoints represented here. Thus, we also ask that you follow some guidelines:
-
Speak plainly. Avoid sarcasm and mockery. When disagreeing with someone, state your objections explicitly.
-
Be as precise and charitable as you can. Don't paraphrase unflatteringly.
-
Don't imply that someone said something they did not say, even if you think it follows from what they said.
-
Write like everyone is reading and you want them to be included in the discussion.
On an ad hoc basis, the mods will try to compile a list of the best posts/comments from the previous week, posted in Quality Contribution threads and archived at /r/TheThread. You may nominate a comment for this list by clicking on 'report' at the bottom of the post and typing 'Actually a quality contribution' as the report reason.
Jump in the discussion.
No email address required.
Notes -
So, it looks like Anthropic has finally gotten the backing to give OpenAI a serious run for their money when it comes to foundational models, via a Series C round of $450 million. There are a few core differences between their approach, and intelligent AI pundits seem to think that these could have dramatic impacts down the road on who gets to control the future of AI. My personal thoughts are still muddied - I'm not sure how much the technical side of things will matter in the age of hype and massive digital marketing. ChatGPT's first mover advantage may have already cemented them the crown of AI for the foreseeable future - unless of course they make a massive blunder.
What's most amusing to me in this whole situation is the way the landscape played out. Even though there were many firms hoping to advance the capabilities of AI, the two that are currently at the top originated as AI Safety outfits. Just like OpenAI, Anthropic sniped a lot of the top engineering talent pretending to care about safety,, then shamelessly pivoted to capabilities enhancement once they became "increasingly convinced that rapid AI progress will continue rather than stall or plateau."
While there are many condemnations and arguments over the AI Safety movement as a whole in the corner of the internet that cares, I think most people that discuss the future of AI don't take these signals strongly enough. The fact that the two largest, arguable most influential 'startup' orgs focused on AI Safety have already flipped to the other side, OpenAI having led the charge, should give anyone who believes in the Eliezer Yudkowsky doomer movement pause. If most people with power in that camp genuinely believed that doom was a given without a long period of AI restriction and alignment research, we should've seen massive departures and drama from OpenAI and Anthropic. The fact that their employees, stakeholders, and most of the AI ecosystem seems content to swallow their facile window-dressing as to why they've switched sides is proof to me that there's a massive lack of conviction in the AI Safety space.
Whatever happens, we're certainly in for an interesting few years. Whether AI continues to advance rapidly or stalls out, the world has already been changed at a level equivalent at least to the birth of the Internet. Now we're just waiting for the dominos to fall and the dust to settle.
Anecdotally, when I chat with Anthropic employees, they seem more genuinely concerned with safety than OpenAI employees, for better or for worse. OpenAI folks seem more or less lip service from top to bottom, while an Anthropic person has told me they struggle to sleep and have nightmares many nights about paperclipping (though that is an outlier). Hopefully the piles of money will help pay for a comfier mattress and therapy.
Obviously that's a pretension among actual decision makers. I wonder how long until we get a third AI company who positions itself as the one that truly, super duper deeply cares about safety and starts poaching employees to get started and raise VC money.
More sympathetically to the idea of working for an AI capabilities company despite being genuinely concerned about safety, suppose you think there's a 10% or even 50% chance of doom. It's very clear at this point that the genie is out of the bottle, and it's unlikely anything you're going to do is going to cause or meaningfully accelerate doom. Might as well make lots of money in the meantime and have a chance at godly amounts of money if doom doesn't happen.
I remain skeptical simply because these are, for better or worse, privately held companies and thus are really only concerned with profitability. And the problem is that safety is rarely profitable. This has happened thousands of times with thousands of products. Flaws, even potentially fatal flaws are not reasons that a product isn’t shipped absent a strong fear of lawsuits or strong regulation. And I’m not sure how hard it would to sue a company for making a dangerous AI, especially if the flaw wasn’t perfectly obvious before release.
One plausible argument for corporate safetyists (which we'll undoubtedly be seeing more of) is that we need strong regulation so that the only AIs created will be the hellful, harmless ones created by big, responsible corporations. The cat may be out of the bag, but perhaps it can be contained in a small room.
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link