Microsoft lays off one of its responsible AI teams • Hacker News Board

Re: Microsoft lays off one of its responsible AI teams

#341

Post by xpe » Tue, Mar 14, 2023, 2:21 PM UTC

djvdq wrote: post #324

ModernMech wrote: post #292

Earlier quoted context omitted.

Depends. As with Internet forums, the presence of Nazis (or whatever similar bad actor) could ruin it for everyone else. Let’s say you are a high schooler writing an essay about WWII. You ask Google LLM about it, and it starts writing Neo Nazi propaganda about how the holocaust was a hoax, but even if it wasn’t, it would have been fine. The reason it’s telling you those things is because it’s been trained by Neo Nazi…

Blatant propaganda will of course ruin such product for regular people. But subtle propaganda? I wouldn't be so sure. And it's not only about Nazis.

Right. Subtle propaganda could shape what arguments are used, how they are phrased, and how they are organized —- all of which can affect human perception. A sufficiently advanced language model can understand enough human nature and rhetoric (persuasion techniques) to adjust its message. A classic example of this is exaggerating or mischaracterizing uncertainty of something in order to leave a lingering doubt in the reader’s mind.

Re: Microsoft lays off one of its responsible AI teams

#342

Post by dustymcp » Tue, Mar 14, 2023, 2:26 PM UTC

heretoo wrote: post #140

ketzo wrote: post #104

Earlier quoted context omitted.

Yeah, this is like claiming lead ethyl in gas “hasn’t caused mass harm” in 1954. There’s absolutely zero way to know that yet.

Except that lead's toxicity has been known for thousands of years.

I would argue social media has at least 10 years of known toxicity, it is so addictive and harmful to young people i dont get why its not prohibited like gambling or smoking..

Re: Microsoft lays off one of its responsible AI teams

#343

Post by yanderekko » Tue, Mar 14, 2023, 2:27 PM UTC

xyzzy123 wrote: post #172

bmc7505 wrote: post #111

Earlier quoted context omitted.

Those two branches despise each other with intense passion. I never actually made this connection before, but having spent far too much time listening to both camps, this observation does ring true. Another observation I had is that there is an orthogonal axis to the safety/ethics dichotomy that follows a more scholastic approach, namely neuro/symbolism. These two branches correspond, roughly, to the debate between m…

I feel the difference is that to a first approximation, ai alignment folks tend to be grey tribe and ai ethics people tend to be blue or rainbow tribe. The schism isn't a technical disagreement about ai, it's a disagreement between subcultures about what values are important.

>The schism isn't a technical disagreement about ai, it's a disagreement between subcultures about what values are important.

I wish it were so, but I think the tribal lines create a lot of motivated cognition about the technical details. I don't think there are many blue tribe members who say "yes we agree that rogue AI has a 10+% chance of destroying humanity by 2040, but I still think that it's more important to focus on disparate impact research than AI alignment." They instead just either assign trivial probabilities to existential risks occurring, or just choose not to think about them at all.

Re: Microsoft lays off one of its responsible AI teams

#344

Post by cataphract » Tue, Mar 14, 2023, 2:30 PM UTC

hn_throwaway_99 wrote: post #9

These sorts of adjunct, off-to-the-side "ethics" teams are always corporate window dressing, and when push comes to shove they always get jettisoned when competition demands an increase in speed. I don't mean to start a flame war, but this is a reason why I think "Chief Diversity Officers" are such an ill-conceived notion. It's not that I don't think diversity in corporate environments is extremely valuable, but it's…

Raymond Chen calls this "you're not my manager": https://devblogs.microsoft.com/oldnewthing/20070522-00/

Re: Microsoft lays off one of its responsible AI teams

#345

Post by yanderekko » Tue, Mar 14, 2023, 2:32 PM UTC

dragonwriter wrote: post #163

ssalka wrote: post #89

Earlier quoted context omitted.

AI alignment people are the group who are capable of bringing AGI to life, but want to do it responsibly. AI ethics people are the group who are not capable of bringing AGI to life, but want to make sure that those who can do it responsibly or not at all.

> AI alignment people are the group who are capable of bringing AGI to life, but want to do it responsibly. AI alignment people are the group who believe “responsibly” means, centrally, bringing AGI to existence as fast as possible, and, critically, under the tight, opaque, authoritarian control of a narrow group of people whose ideology is, well, exemplified by figures like Sam Bankman-Fried, and doing everything po…

>AI alignment people are the group who believe “responsibly” means, centrally, bringing AGI to existence as fast as possible, and, critically, under the tight, opaque, authoritarian control of a narrow group of people

Hey, let's be fair now, a lot of them also believe that we just need a world government that tightly controls GPU use.

Sarcasm aside, the alignment crowd definitely does not have a consensus about the desirability about creating AGI asap. Yudkowsky is clearly horrified by OpenAI.

Re: Microsoft lays off one of its responsible AI teams

#346

Post by chamanbuga » Tue, Mar 14, 2023, 2:35 PM UTC

This is an incredibly click bait article by gifted journalists. Most of these folks are fine and still have similar core priorities. Also there are multiple other responsible AI teams in MS. I have extreme empathy for those impacted, but this is trying to generate news where there isn't.

Re: Microsoft lays off one of its responsible AI teams

#347

Post by giraffe_lady » Tue, Mar 14, 2023, 2:37 PM UTC

thwayunion wrote: post #339

What a strange comment section. A customer service chatbot telling your customers that an issue is solved, when the issue isn't solved, is a core product failure. It's also a Responsible AI issue. A code generation model suggesting GPL code without attribution is a core product failure and serious legal risk [1]. It's also a Responsible AI issue. A diffusion model generating child porn or nazi propaganda is a core pr…

But it has a whiff of what used to be called political correctness so we must RAGE and reject it entirely.

Re: Microsoft lays off one of its responsible AI teams

#348

Post by JustBreath » Tue, Mar 14, 2023, 2:39 PM UTC

pixl97 wrote: post #54

monero-xmr wrote: post #20

Earlier quoted context omitted.

How can you stop a computer program? Make math illegal?

I mean if you want an honest answer, nuke any place that's making GPUs/TPUs. I mean, they don't spring up out of the ground and at least our current models require insane amounts of GPU/TPU based compute. The factories are fragile billion dollar facilities, well beyond the difficulty of extracting uranium. Now, that probably just delays the world eating machine for some time, but it would be a hell of a lot harder to…

Didn't work for the luddites.

Re: Microsoft lays off one of its responsible AI teams

#349

Post by xiphias2 » Tue, Mar 14, 2023, 2:42 PM UTC

JohnFen wrote: post #144

g42gregory wrote: post #136

Earlier quoted context omitted.

There is an implicit assumption that the teams are wrong. What about the wisdom of the crowds? If most teams think it's not needed, why do we think they are wrong?

I don't see it as an implicit assumption that the teams are wrong, but a recognition that the teams have a fairly narrow goal that can easily blind them to larger implications of design decisions as they "move fast and break things". Having some sort of outside check on that seems wise to me.

Google, Microsoft and Facebook had those teams, all they achieved is to get a new competitor by slowing down the natural pace of getting things into production by dumbing things totally down instead of finding technical solutions to ethical problems that keep most of the power of the AI systems.

Re: Microsoft lays off one of its responsible AI teams

#350

Post by dragonwriter » Tue, Mar 14, 2023, 2:42 PM UTC

ben_w wrote: post #220

dragonwriter wrote: post #163

Earlier quoted context omitted.

> AI alignment people are the group who are capable of bringing AGI to life, but want to do it responsibly. AI alignment people are the group who believe “responsibly” means, centrally, bringing AGI to existence as fast as possible, and, critically, under the tight, opaque, authoritarian control of a narrow group of people whose ideology is, well, exemplified by figures like Sam Bankman-Fried, and doing everything po…

Sam Bankman-Fried is a uniquely terrible example of ideology, given St. Petersburg: https://conversationswithtyler.com/episodes/sam-bankman-frie... and https://astralcodexten.substack.com/p/open-thread-250 (and, y'know, being a fraudster who temporarily pulled the wool over the eyes of the regulators not just the charities he promised money to) Also: is SBF really the central example of alignment in your head, and no…

> Also: is SBF really the central example of alignment in your head,

No, he's a particularly well-known example of the broader ideology to which virtually all of tbe alignment camp subscribes, and which centers their desires for AI.