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

Google’s AI-powered ‘inclusive warnings’ feature is very broken

vice.com

Post reply on HN
Print view View original

981–990 of 1001 posts

Re: Google’s AI-powered ‘inclusive warnings’ feature is very broken

#981

Post by account42 »

la6472 wrote: post #548
meetups323 wrote: post #11

I wonder if this is an example of not knowing your audience (or rather assuming one's self represents a wider audience than it does). The people vouching for this at Google are likely product marketing managers, public relations folks, social media managers, etc. All they do is write corporate garbage all day, and much like we have "nit"s in PR's for formatting, variable names, etc; they likely have similar reviews t…

I support it and I don’t work for google. I like this feature that encourages more empathy in this strife and hate filled world.

Neither hate or empathy are in words being written but in the intention behind them. If anything, corporate writing is all about feigning empathy while having none. Or do you actually believe that corporations care deeply about your privacy or all the other things they claim to do.

Re: Google’s AI-powered ‘inclusive warnings’ feature is very broken

#982

Post by leaflets2 »

npteljes wrote: post #960
madamelic wrote: post #181

Earlier quoted context omitted.

No, they are not. Teach someone with no previous cultural knowledge what "allow" and "deny" means, then ask them what white list and black list means, then ask them what allow list and deny list means. Black and white list requires previous knowledge of what those terms mean, allow and deny do not. Someone can intuit from just basic language knowledge what the feature does.

My pet peeve is that the new, inclusive terms are often more cumbersome. For example, allowlist is longer by a syllable. I know that it's hard, but I wish that people came up with terms that are just as handy as the old ones. Changing blacklist to blocklist is quite cool for one.

And "main" is shorter (better I think) than "master".

Although I don't think anyone ever seriously thought that "master" branch meant the "lord" branch.

Re: Google’s AI-powered ‘inclusive warnings’ feature is very broken

#983

Post by kevincox »

Reading this I don't get quite the same reaction as everyone else. I think this is actually a useful tool. However the main failing is that there is one global ruleset decided by Google. If this was customizable by each organization or department it could actually have value. I don't think shaping the way your company communicates is much of a problem, the problem arises when Google shapes how your company (and individual users speak). It would also be interesting if these different models could be published and shared.

Basically I think the core idea is fine. But applying the same model globally was a clear mistake.

Re: Google’s AI-powered ‘inclusive warnings’ feature is very broken

#984

Post by School-Cotton »

powersnail wrote: post #875
deanCommie wrote: post #868

Earlier quoted context omitted.

Quoted post unavailable.

That’s interesting. I just did this introspection, and I find that I say: Trump, Obama, Biden, Hilary Clinton, Kamala Harris (full name). But interestingly, I call Jimmy Carter the full name, too. Can’t bring myself to use just “Carter”. Though, why does the US culture so insist on taking the husband’s family name? It kinda feels like “Hilary” is the actual name of the person, and Clinton is imposed.

> why does the US culture so insist on taking the husband’s family name?

It doesn't really anymore. But it did when the Clintons got married several decades ago.

Re: Google’s AI-powered ‘inclusive warnings’ feature is very broken

#985

Post by darkerside »

noduerme wrote: post #980
darkerside wrote: post #977

Earlier quoted context omitted.

I got it, and I think my point still stands. Some of the problem here is Poe's Law: > every parody of extreme views can be mistaken by some readers for a sincere expression of the views being parodied I could tell you were being sarcastic, but the fact remains your said what you said. Sarcasm is great for tearing bad things down, but useless for building anything to replace it. I am not a fan of this Clippy shit eith…

But I just explained it, if not to you then to someone who didn't get it. I'm absolutely willing to delve into the machinery of anything to explain it. I'm not, however, willing to live in a world where everyone is encouraged to tailor their expressions to the lowest common denominator. The truth is that speaking to the LCD is pointless unless you're a populist who needs their vote, or otherwise need to manipulate th…

It's just ironic that you explained it to me, who already got it, because I asked you to clarify it for other people, who didn't ask. It's exactly my point.

I also don't want an anodyne and bland world, but I believe some communication should be that way. In fact, I believe the world would be a better place if the populist messaging were more lower common denominator, and our personal communications more dynamic. Not because it legally has to be, but because it helps preserve the status quo (which I selfishly like).

I think there are other reasons to care about the "lowest common denominator" (which really means, people who don't think like me) besides to manipulate or gain something from them. I believe everyone has value as human beings, and we'll build a better society if everyone understands the conversation.

Re: Google’s AI-powered ‘inclusive warnings’ feature is very broken

#986

Post by raxxorraxor »

rayiner wrote: post #711

It’s also inaccurate to call this language “inclusive.” In reality, you’re excluding most Americans who aren’t among the handful of mostly affluent, white, college educated people who talk like this. My Bangladeshi immigrant mom would have no idea what to do if a warning popped up telling her that she should use a different word than “mother.”

It isn't colleges that advertise this exclusively, it is also corporate consulting like McKinsey. Know them? Allegedly the Robin Hood of the poor and disenfranchised.

The author of the article still sees it as a necessity. They cannot leave people talking like they want to.

Re: Google’s AI-powered ‘inclusive warnings’ feature is very broken

#987

Post by raxxorraxor »

ajyotirmay wrote: post #836

I still wonder, how blacklist/whitelist or master/slave has helped or promoted slavery and racism. And how many devs who are from black communities offended by the use of these terms, at least before it was made to be an issue in mainstream. Because right now I feel like this is a "woke feminist" issue. And I'm not a white person. I've never thought otherwise of these words, but maybe that can be because of the mixed…

Probably severe nepotism with industry consulting or someone has some idealistic child in colleague and their guilt about neglected parenting made them force the issue as it has been adopted in some parts of the industry. At least official, never seen anyone use different terms by now and I don't even know the substitute. It is always something like this. But yeah, they certainly are white as the snow, so much is sure.

Re: Google’s AI-powered ‘inclusive warnings’ feature is very broken

#988

Post by ElFitz »

JaimeThompson wrote: post #462
ElFitz wrote: post #62

And now the rest of the world will be very happy learn that they will once more have to silently deal with the aftermath of America’s latest political trend. After banning world-famous paintings from Facebook because they dared show a breast, a few well-off engineers and product managers will now pick which English words should remain, and which should be let go. How nice of them. Don’t know what we’d do without them…

>banning world-famous paintings from Facebook because they dared show a breast Issues with breasts aren't exactly a new thing in the US, just ask Janet Jackson.

For context: https://en.wikipedia.org/wiki/Super_Bowl_XXXVIII_halftime_sh...

And this quote, from the wikipedia article, sums it all quite well:

> "I know many people in other countries are scratching their heads and thinking 'What in the world is the big fuss over there?'"

Re: Google’s AI-powered ‘inclusive warnings’ feature is very broken

#989

Post by raxxorrax »

limelight wrote: post #674
slg wrote: post #648

Earlier quoted context omitted.

>However, the net result is that words are driven out of the language even if everyone involved in the document wouldn't care. No one uses "fuck" in corporate communication either. Is anyone worried about that word being "driven out of the language"? I don't think corporate speech is as influential in overall language use as you are implying.

> No one uses "fuck" in corporate communication either. Swear words are specifically used to be provocative, so naturally they're not going to disappear. Words like "motherboard" or "whitelist" were historically neutral and primarily used in professional settings, so removal from corporate speech is correspondingly a much bigger factor. To be clear, I'm not particularly worried or concerned about this. I don't consid…

That is just enabling some pretty bad actors in my opinion who don't even care about minorities. People that do care tend to be very flexible about language use.

Not being tolerant here is a pretty clear sign of close mindedness in reality or just a hint of lacking life experience.

Re: Google’s AI-powered ‘inclusive warnings’ feature is very broken

#990

Post by la6472 »

Stupulous wrote: post #654
la6472 wrote: post #548

Earlier quoted context omitted.

I support it and I don’t work for google. I like this feature that encourages more empathy in this strife and hate filled world.

While I don't doubt that an intent of this is to promote empathy, I would need to see some evidence before I could be open to the possibility that that is its effect. Anecdotally, these things seem to incense anger and hatred- I've never heard anyone say that being language-policed made them a better person, but I have seen people behave in a way that suggests the opposite. Personally, I become less empathic when som…

To begin with you and your crew can stop downvoting comments you disagree with. Do it and you will have evidence of change.
Post reply on HN