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

341–350 of 1001 posts

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

#341

Post by SpicyLemonZest »

scythe wrote: post #301
timmg wrote: post #27

Earlier quoted context omitted.

As someone who also works at Google, I agree that there is a vocal minority who want Google to be a promoter of progressive values, but: I think you nay be exaggerating things here. I've never even heard of any controversy over the word "motherboard". I just did some internal searches. It seems to be a well-used word, both in code and in documentation. (I did a code search for both "motherboard" and "mainboard" and t…

"Allowlist" and "denylist" are sonically awkward. Why not "inlist" and "outlist"?

Directional terms like that would be confusing in a lot of infrastructure discussions. An outbound firewall rule's "outlist" is the list of things that don't go out.

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

#342

Post by cj »

nostromo wrote: post #31

Who was asking for Woke Clippy? Is Google just bored and completely out of ideas? They seem increasingly disconnected with what their users actually want.

You could look at Grammarly ($10+ billion dollar valuation) as evidence that there is a market for a tool/plugin that improves a user’s prose.

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

#343

Post by umvi »

mgraczyk wrote: post #9

In the article they use the example of "Motherboard" and "Landlord" and seem to suggest that these are areas where the Google AI is making mistakes or being overly strict. As a Google employee expressing my own opinion and observations of company culture, I can say that these are 100% not mistakes. Many Google employees are just so out of touch with the real world that they believe it is the duty of Google Docs to ch…

> This may seem unbelievable, but the word "motherboard" is literally banned within Google and you are required to use "mainboard" instead. You are not allowed to use this word in documentation or code, and you're also not allowed to say it privately in chats or emails.

Is this really true? I always thought I might eventually apply to work at Google someday, but I hadn't heard there was such aggressive internal thought policing.

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

#344

Post by shp0ngle »

vimy wrote: post #204
mgraczyk wrote: post #9

In the article they use the example of "Motherboard" and "Landlord" and seem to suggest that these are areas where the Google AI is making mistakes or being overly strict. As a Google employee expressing my own opinion and observations of company culture, I can say that these are 100% not mistakes. Many Google employees are just so out of touch with the real world that they believe it is the duty of Google Docs to ch…

This is why I’m pessimistic about the West winning the second cold war. While Chinese engineers are working hard on world dominating AI models and other hard tech, Western engineers are wasting man hours on crap like this.

Oh come on now.

Chinese have far bigger and far more strict word list bans.

In US, you need to say “people who menstruate” and “land owning person”; in China you will just just disappear if you say a wrong thing against the regime.

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

#345

Post by grishka »

duskwuff wrote: post #329
grishka wrote: post #269

Earlier quoted context omitted.

> somebody had to explain to them there are people in the world who are discriminated against, but aren't black. As a Russian, this is amusing to read. For me and most people I know, when you meet a black person, it's totally normal to ask them "where are you from" because they can't possibly be local. Our society just doesn't have the concept of racism it seems because of the exceeding rarity of people who aren't Eu…

> Our society just doesn't have the concept of racism it seems because of the exceeding rarity of people who aren't European or Asian. "Slave is an Ephebian word. In Om we have no word for slave," said Vorbis. "So I understand," said the Tyrant. "I imagine that fish have no word for water." -- Terry Pratchett, Small Gods

Either way, people here aren't discriminated based on their race. Never were. This particular problem seems to be uniquely American because of their history.

We do have a word for racism by the way. It's, unsurprisingly, "расизм".

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

#346

Post by AlexDragusin »

What I can see is that on Google own websites, the term motherboard is used aplenty so seems to me they fail their own standards, if what is described in the article is accurate. Wondering when humanity will become aware of the ridiculous path this whole thing is.

Examples: https://cloud.google.com/blog/topics/systems/the-past-presen... https://sustainability.google/progress/projects/circular-eco...

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

#347

Post by enriquto »

madamelic wrote: post #164
augustuspolius wrote: post #72

Earlier quoted context omitted.

Encountering this aspect of the culture war always makes me wonder why are tech companies so focused on these gender minorities and not, say, on the disabled groups. There are more blind, deaf, mute (etc, etc) people in the US than transgender people who will bear a child. It would be equally incoherent to attempt to replace all usages of “see”, “hear” with “perceive”.

Trans people are frequently in tech. Go to any college and you'll see the CompSci department is the one with the most amount of LGBT individuals. The same can't be said for people who are blind or deaf.

> Go to any college and you'll see the CompSci department is the one with the most amount of LGBT individuals.

Do you have any data on that? It does not coincide with my anecdotal experience. I'm in a Math/CompSci department and there's zero trans people here (that I know of), while there are some in other departments. Anyhow, trans people are a tiny percentage of the total population, so it would be hard to have somewhat solid statistics on them.

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

#348

Post by trollied »

bombcar wrote: post #46
gdulli wrote: post #29

I spent many years in the bluest corner of a very blue state. But still I can't imagine any objection to the word "landlord" outside of (1) a right wing attempt to satirize the left, (2) a Google product brainstorming meeting, or (3) a rare sincere outlier or concern troll. In any of those cases I don't think this is necessary. That said, if the word "landlord" organically fell out of favor and out of usage to be rep…

Landlord gets hit from multiple sides; those who don't like that it contains "lord" which is a male term, those who don't like that landlords exist, and those who don't like that the concept of owning land exists.

It’s just a modern word. “Landlord”. People know what it means. Stop anybody on the street and ask them what a landlord is? They will tell you. It’s an established word. What’s the point in changing it because a few people actively want to be offended?

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

#350

Post by exyi »

riffraff wrote: post #76

The most hilarious thing about this is that Google has used "AI" to create the most un-inclusive tool of all, Google Translate. I present you the beautiful translation of hungarian gender-less sentences to english: ő csinos. ő okos. ő csúnya. ő jó. becomes she is pretty. he is clever. she is ugly. he is good. I can't wait for inclusive warnings to come to non-english languages.

While it's hilarious, it's understandable IMHO - just basic statistics, she is more often used in front of pretty/ugly. It's showing a glitch in our society more than a glitch in Google AI.

This on the other hand looks like a hand-crafted blacklist of words that they want to remove from the language, I have no idea how would I train an AI which would classify "motherboard" as inappropriate.

Post reply on HN