Live data from Hacker News

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

vice.com

571–580 of 1001 posts

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

#571
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 the company that fired James damore and then turned around and spouted his talking points a few months later not even realising the hypocrisy.

https://mashable.com/article/google-youtube-women-in-tech-di...

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

#572

Earlier quoted context omitted.

This entire comment is a lie. Terms like "motherboard" are readily findable on Google's internal search. I could not find anything referencing a banned non-inclusive words policy in the employee policy guide. I have never heard of anything remotely resembling someone catching heat for saying "motherboard" or whatever. Source: I also work at Google.

You're welcome to ping me internally. It's in the banned words list (happy to share a link internally) and I double checked policy before posting this comment to make sure I was accurately representing the rules that are applied to me as an employee.

Okay but that’s just a list that someone created. People are allowed to create lists; I’ve never heard of anyone enforcing this list.

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

#573
post #319

Earlier quoted context omitted.

Reminds me of when GitHub stopped using "master" for the default branch because it was somehow offensive.

When I built a distributed system, I wanted to avoid the terms "master and slave" for the "coordinator and worker", so instead I chose something I thought was relatively less controversial- daimyo, honcho, and peasant. Only later did I realize I had merely recapitulated the power structure of feudal japan.

I prefer sovereign and vassal. No one has called me out on it.

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

#574
post #291

Earlier quoted context omitted.

Main is better. Master/slave has uncomfortable connotations. This one isn't a big deal.

Where is the slave in git?

There is no "slave" in Git, but the term "master", like many things in Git, is taken from BitKeeper, which did have a "slave". Or so I've read, for example here: https://english.stackexchange.com/questions/474419/does-the-...

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

#575

I can't even properly describe how angry this makes me.

Imagine the life of someone who thinks that this is a good and important change. It's probably miserable-- a slight around every corner and no higher purpose than bulling people over terminology. Instead of anger, try gratitude that you're not in their shoes.

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

#577

Earlier quoted context omitted.

Silly question: ignoring any kind of history, why fight for terms like "whitelist" and "blacklist" when terms like "allowlist" and "blocklist" are objectively clearer, especially for folks who's first language is not English?

I am fine with "blocklist" instead of "blacklist", because it sounds similar. But "allowlist" is just more difficult to pronounce. Regarding your concern about people for whom English isn't the first language, I dont know how valid it is. Because those terms exist under the same names and fall under the same usage in many other languages. I can confirm that it is the case with Russian, as "черный список" is a commonl…

> I am fine with "blocklist" instead of "blacklist"

Better hope you're not dealing with a block device or other data structure that involves units of data called blocks.

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

#578

Earlier quoted context omitted.

Reminds me of when GitHub stopped using "master" for the default branch because it was somehow offensive.

Main is better. Master/slave has uncomfortable connotations. This one isn't a big deal.

Can you please not use the word s____? My ancestors were Slavs, and when I see the word it is a painful reminder of how they were treated as property by Romans who bought and sold their "ex slava" captives.

Your use of the word is violence.

/s?

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

#579
post #342

Earlier quoted context omitted.

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.

Being told I use too many adjectives, my word choice is too heavy for my audience or that I used too many clauses is useful. Being passive-aggressively told "blacklist" is a no-no word is not. If you were really clever, you'd let the user set a "sensitivity" level for the writing from "normal human speak" to "corporate PR approved".

In this particular case, the tool being discussed is meant almost exclusively for people writing "corporate PR approved" content, so it makes sense that the sensitivity setting is cranked.

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

#580

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.

Yeah, who asked for this? One would think they'd be trying to achieve feature parity with Microsoft Office instead.
Post reply on HN