Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Microsoft lays off one of its responsible AI teams

platformer.news

Post reply on HN
Print view View original

121–130 of 466 posts

Re: Microsoft lays off one of its responsible AI teams

#121

Post by TedDoesntTalk »

Eisenstein wrote: post #91
xyzzyz wrote: post #48

Earlier quoted context omitted.

This is the case with literally all kinds of professional ethicists. Thanks to them, we do not routinely do scientific challenge trials and plenty of other valuable experiments, and in many cases are reduced to collecting significantly worse data and doing it very slowly and expensively.

Much better off when were literally traumatizing babies to see how they would react to adverse stimulus. * https://en.wikipedia.org/wiki/Little_Albert_experiment

And the Tuskegee Syphilis Study … The goal was to observe the effects of syphilis even though a cure was available

https://en.m.wikipedia.org/wiki/Tuskegee_Syphilis_Study

Re: Microsoft lays off one of its responsible AI teams

#122

Post by mc32 »

nradov wrote: post #110
mc32 wrote: post #102

Earlier quoted context omitted.

The solution is often making it impossible for people to commit mistakes. Just like you can't "accidentally" launch an ICBM. I admit some of these may be difficult to accomplish with training protocol, but often these firms that come in and advise on how to be "safer" are just pounding harder. You know what, making someone glaze over a checklist isn't going to make them be safer. It takes the right personality and at…

Please explain how your make it "impossible" to run over a co-worker with a forklift. Have you ever worked in a warehouse? Checklists have been proven to improve safety in many environments. They aren't a panacea and some workers will inevitably ignore them, but they do help.

We're getting a bit off the rails here. My point is a lot of these cottage industries that pop up to address whatever the current zeitgeist is in the workplace is 90% useless and the time (money could be better spent).

You could re-design warehouses. Add sensors so that proximity activates a power kill switch, etc. But this is not the main thrust. The main one is most of these expert advisors are closer to extortionists than problem solvers.

I will believe them and take them seriously when they volunteer their commitment to improve things. People go to those time management trainings and the same wasteful meetings continue but with some added nonsense twist. Harassments training but managers still can't keep their hands to themselves. SOX and people still misappropriate funds. Group collaborations with some idiotic legos exercise and people still protect their silos.

Re: Microsoft lays off one of its responsible AI teams

#123

Post by im3w1l »

nradov wrote: post #110
mc32 wrote: post #102

Earlier quoted context omitted.

The solution is often making it impossible for people to commit mistakes. Just like you can't "accidentally" launch an ICBM. I admit some of these may be difficult to accomplish with training protocol, but often these firms that come in and advise on how to be "safer" are just pounding harder. You know what, making someone glaze over a checklist isn't going to make them be safer. It takes the right personality and at…

Please explain how your make it "impossible" to run over a co-worker with a forklift. Have you ever worked in a warehouse? Checklists have been proven to improve safety in many environments. They aren't a panacea and some workers will inevitably ignore them, but they do help.

I would go as far as impossible, but those proximity sensors cars use work pretty well.

Re: Microsoft lays off one of its responsible AI teams

#124

Post by atleastoptimal »

pixl97 wrote: post #17
monero-xmr wrote: post #8

If Microsoft and Google don’t release, some startup will beat them to it. Open source means raw AI will become widespread. It’s just math anyway, and you can’t stop progress.

"If I don't make the world eating machine, someone else might make the world eating machine first!" No wonder we don't see aliens filling the universe. Intelligence is its own great filter.

"I'm gonna steal the Declaration of Independence"

Re: Microsoft lays off one of its responsible AI teams

#125

Post by dang »

daydream wrote: post #100
karpierz wrote: post #76

Earlier quoted context omitted.

The (productive) point of the ethical AI team is to have internal people to consult when launching an AI product. Embedding ethical AI people on teams makes about as much sense as embedding legal counsel. Their services won't be used on a single team nearly enough to justify their inclusion. Of course, it can be window dressing if the company doesn't actually have product teams care. Just like a CDO position. But you…

The product teams don’t care. And the ethical AI team doesn’t want to sit back and wait for people to come to them. They want a seat at the table for all important AI products. More in my comment on the other submission about this news https://news.ycombinator.com/item?id=35146611

(We've merged the threads now. Thanks for linking, not copying!)

Re: Microsoft lays off one of its responsible AI teams

#126

Post by nradov »

ssalka wrote: post #89
anonylizard wrote: post #19

Earlier quoted context omitted.

This. People should understand AI safety != AI ethics. AI safety has two branches, AI ethics and AI alignment. Those two branches despise each other with intense passion. The alignment branch (responsible for developing RLHF->ChatGPT) believes AI ethics is completely trivial, and distracts attention away from existential risks. The ethics branch (responsible for all the media attention about racist AIs/bias etc) view…

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.

There is no evidence that any group is capable of bringing AGI to life.

Re: Microsoft lays off one of its responsible AI teams

#127

Post by dangond »

zug_zug wrote: post #29
monero-xmr wrote: post #20

Earlier quoted context omitted.

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

Well, if the first AGI is made in a closed controlled simulation honeypot (which it believes isn't a honeypot) and it immediately starts hacking everything then we'll at least have confirmation how immediately dangerous it can be.

Since the other commenters haven't made clear what the problem with this is, here's my question to you: How do you know that this AGI is 1) an AGI, and 2) that it believes it is not in a honeypot? In other words, at what point do you believe an AGI is safe to release into the wider world if you're testing it using this method? Is it safe after spending a day without any visibly nefarious actions? What will you do when you release it and it turns out it had been pretending to not know it was in a honeypot?

Re: Microsoft lays off one of its responsible AI teams

#129

Post by Dylan16807 »

pixl97 wrote: post #112
teolandon wrote: post #90

Earlier quoted context omitted.

Your original comment didn't talk about ethics or morals, but was talking about intelligence being its own great filter. More intelligent AI consuming us could be unethical and immoral in your view, but it would still be intelligence spreading, and not filtering itself. Just not specifically human intelligence.

I mean, killing all the people on the planet seems unethical to me. Maybe I have a deep misunderstanding of ethics.

But you weren't talking about the ethical ramifications. You made a claim about a great filter, and that claim doesn't hold up.

If I say that using humans as batteries in the matrix doesn't make sense, that's not tacit approval of the rest of the scenario.

Re: Microsoft lays off one of its responsible AI teams

#130

Post by 1attice »

Former MS AI Platform IC here.

The main value of this team was likely internal, as a sort of gauze bandage on the conscience of the progressives (like myself) that MSFT has tried to court.

There was a time, not too long ago, where working at the new, 'refreshed' Microsoft meant ingesting an endless stream of news releases highlighting zero-carbon this, compostable cutlery that, and, of course, ethical AI. When Tegru was let go from GOOG we hosted a talk by her. That sort of thing. The point was to make you feel like you were one of the good guys -- your TC in effect included the intangible bonus of sleeping soundly at night, knowing that the Aethyr Council or whatever was doing all the handwringing for you. So, you know, no need to worry about social impact, just keep coding.

What this really means is, "the market is so bad right now for SWE that we no longer have to worry about the qualms of our staff, if you don't like us, the door is over this way"

Post reply on HN