Comment by dmix | Hacker News Reader

Comment by dmix | Hacker News Reader

Anyone paying close attention to the output of generic "AI ethics" people (especially on Twitter) should not be surprised that companies that when push comes to shove the businesses getting actually useful AI tools that are being tested by real people for real things generating real data, are starting to question the utility of the bulk of the crowd who skirted off hypotheticals for years to build their reputations as "experts". Once the hypotehtical meets the meat of the problems you usually can quickly root out the people not providing any value and merely pushing trendy FUDy positions.

The key here is that the costs are still as high a ever, Microsoft is very aware of the consequences of ignoring the legitimate AI ethics issues. So it's not likely it's merely ignoring risk.

The real useful AI ethics people will be born out of real world applications. Not pearl clutching by people making grand projections, often with a limited grasp of what the technology actually means in practice, in a competitve open market where you the cat is out of the bag even if you don't like it.

The legitimately talented AI thinkers will still be highly employable at the end of the day. Even if a small subset of talent got burned by the output/management of their previous team they'll be okay long term. Hell, even the non-talented ones will still be in demand considering the demand in the media for stories/politically convient takes, and the general tolerance for mediocrity in mega-tech-firms.

I think there is a lot of unjustified hatred for the AI ethics guys. From what I have seen, they have raised manu important issues so far, and led to many improvements both around ethics and AI in genetal.

For example, the cops wanted to arrest people based on input from image recognition systems. Someone demonstrated that the leading system flagged multiple members of Congress as dangerous fugitives and that was the end of that.

> Someone demonstrated that the leading system flagged multiple members of Congress as dangerous fugitives and that was the end of that.

That's not a question of ethics, it's a question of the model performance.

Although it is an attempt to circumvent unethical use of AI

If the model performed so poorly it would be an ethical concern to deploy it, as many departments were so eager to do.

The performance wasn’t uniformally poor across all demographics, there were some subsets of people which this performed worse then others, An expert in ethics would know where to start testing a model for suspected biases. A simple QA of the model—particularly if the QA team is not very diverse—is more likely to fail at testing for these biases.

Yeah I think the question is are these people just philosophers pontificating about AI doom, or are they AI researchers creating techniques to make AI decisions less opaque and biased (or less biased in undesirable ways at least)?

The latter are definitely useful but the former is something you start so you can look good and then disband a few years later when they produce little useful output and cost a ton of money.

“On the other hand, everyone involved in the development of AI agrees that the technology poses potent and possibly existential risks, both known and unknown. Tech giants have taken pains to signal that they are taking those risks seriously — Microsoft alone has three different groups working on the issue, even after the elimination of the ethics and society team. But given the stakes, any cuts to teams focused on responsible work seem noteworthy.”

I read your take as "ethics doesn't provide immediate economic value or model improvement along traditional metrics of performance, so companies are right to can them if they choose". I see ethics as looking at problem before they happens which is sort of an economic externality that companies won't pay for since the cost isn't something they have to bare, but in a more concerned world would.

This, 100%.

> Hell, even the non-talented ones will still be in demand considering the demand in the media for stories/politically convient takes

Language models can probably do a better job soon.

> Microsoft is very aware

thanks, that is comforting

> the cat is out of the bag even if you don't like it

thanks for bullying us around

> the legitimately talented AI thinkers

who exactly authorised you to project legitimacy in a space that has not seen any regulation?

>thanks for bullying us around

Who is bullying who exactly?

>who exactly authorised you to project legitimacy in a space that has not seen any regulation?

He's using the second meaning of 'legitimate'

   able to be defended with logic or justification; valid.

>Who is bullying who exactly?

People telling that we shouldn't worry about such topic, because companies are aware of them. Because companies never did bad things just to earn more, right?

Seriously.

Why is it that in a forum skewed libertarian techbros, there's this nigh-religious faith that corporations will not only get AI right, but get its impacts on society right? You know, with their stellar track record and all.

To me this is just another of countless demonstrations of silicon valley ego/hubris.

Wait? you think you know better than thousands of other people who dedicate their lives to the topic, assume ideology and attack them with slurs and you think /they/ are the ones with too much ego and hubris? Dunning-Kruger effect much?

> thousands of other people who dedicate their lives to the topic

to make it plain to you, there were thousands of slave traders who dedicated their lives to the topic, including e.g. how to optimally fill-up the ship with bodies. what does this prove?

the idea that meticulous pursuit of a domain somehow gives it its experts the moral high ground or ensures that they will keep it safe for society is so bizarre and alarming it only reinforces the notion that a bunch of people have become completely unhinged

AI practitioners have already proved themselves untrustworthy by putting themselves in the service of entities that invaded privacy and engaged in large scale algorithmic manipulation of e.g. voting. This is not an assumption. Its a dire fact.

More broadly, corporate structures have repeatedly proved themselves untrustworthy, both in the small, with scandals and fraud and at-large, with regulatory capture that ensured their negative impacts on society could go unhindered for decades

Did I hit a nerve of yours? Maybe you see a little too much of yourself in my comment? If you think that the risk of getting technology catastrophically wrong is more than a passing philosophical diversion to your "thousands", when we've all heard "move fast and break things" ad nauseum, maybe you should be examining your DK blind spots instead of accusing someone else of the same.

Also, is "libertarian techbro" a slur now? Or are you just resentful that I compared faith in progress to faith in a deity?

Reply on news.ycombinator.com