3 ms·

Reflects is a good word. Generative AI only works for the most popular languages on the internet, and due to its use cases it reflects those languages back into

Reflects is a good word. Generative AI only works for the most popular languages on the internet, and due to its use cases it reflects those languages back into the internet. This creates a feedback loop that amplifies the popularity of the most popular languages at the expense of the ones that are already more marginal. Yes, it's continuation of the same pattern we've been seeing for decades, but I think it's fair to say that generative AI will escalate that pattern dramatically unless specifically addressed. > How is that a problem? Setting aside my aesthetic enjoyment of the wide variety of languages that exist in the world, there's a strictly utilitarian argument to be made that placing the majority of the world's population in the status of second-class citizen (having to perpetually conduct all business in a language other than their native tongue) is not the best way to serve humanity as a whole.

Galatians4_16 3y ago
So band together, make a list, and use those languages? Seems solvable enough to me.
- lolinder 3y ago
  Yes, it's solvable! Step one is to identify the problem, though, which is what this article is doing.
roenxi 3y ago
> ... my aesthetic enjoyment ... Without asserting that English would be a good choice of global language; this seems like a terrible reason to silo huge chunks of humanity from each other. There is an extremely good argument that we should all standardise on Mandarin Chinese with a better alphabet to promote good international communication. We're in an era where nuclear-armed powers are fighting each other again. Unclear international communication could literally get us all killed. The aesthetics of everyone having their own lingo is not worth the risk.
- lolinder 3y ago
  > this seems like a terrible reason to silo huge chunks of humanity from each other That's why I set that argument aside.
- Turing_Machine 3y ago
  There is a difference between "language with the most native speakers" and "language that is the most widely spoken". Mandarin may win the first one, but fails the second one dramatically. Few people outside China speak Mandarin as their first language. Mandarin has L1: 940 million, L2: 200 million -> 1.14 billion English has L1: 380 million L2: 1.077 billion -> 1.46 billion Also, most of the L2 Mandarin speakers live in China, with another Chinese dialect as their native language.
dumbfounder 3y ago
I appreciate your point about generative AI exacerbating the dominance of major languages. It's a nuanced issue, indeed. While it's true that AI reflects the linguistic data it's fed, this doesn't necessarily mean it only perpetuates these languages without offering potential benefits to less-represented ones. Consider the dual nature of technology: it can indeed centralize or decentralize. For less-spoken languages, AI and machine learning can be leveraged to preserve and revitalize them. Projects like Google's Woolaroo (https://artsandculture.google.com/project/woolaroo https://artsandculture.google.com/project/woolaroo), using machine learning to help preserve endangered languages by translating them into more widely spoken ones, hint at potential pathways forward. While generative AI may currently favor dominant languages due to the sheer volume of data available, there’s growing awareness and effort to make these technologies more inclusive. This includes not only better representation in training data but also the development of tools specifically aimed at language preservation.
- lolinder 3y ago
  Agreed! And I hope we continue to see such efforts to address language representation. I don't think hope is lost, far from it, I'm just trying to answer OP's question about why reflecting the current state of languages can be a problem.
jstummbillig 3y ago
> Yes, it's continuation of the same pattern we've been seeing for decades, but I think it's fair to say that generative AI will escalate that pattern dramatically unless specifically addressed. It's cool that it can be addressed though, right? AI can be both, the fast track to escalating the issue and the fast track to deescalating the issue. Interesting.
jhbadger 3y ago
>Generative AI only works for the most popular languages on the internet I'm not so sure about that. I was amused when playing with ChatGPT shortly after it was public that it could answer questions in Esperanto and even write the silly sorts of stories in it than it can in English.
BlueTemplar 3y ago
Not decades, centuries (at least). A synonym for "AI" is machine : consider what effect the printing press (a written word-replication machine) had on less popular languages ? (It would be interesting to compare Europe and Asia...) This article is just desperately lacking in awareness like this - see also how, frustratingly, it barely mentions how a language is spoken first, and only maybe written second, but doesn't spend much time on that aspect.

Galatians4_16 3y ago

So band together, make a list, and use those languages? Seems solvable enough to me.

lolinder 3y ago
Yes, it's solvable! Step one is to identify the problem, though, which is what this article is doing.

roenxi 3y ago

> ... my aesthetic enjoyment ... Without asserting that English would be a good choice of global language; this seems like a terrible reason to silo huge chunks of humanity from each other. There is an extremely good argument that we should all standardise on Mandarin Chinese with a better alphabet to promote good international communication. We're in an era where nuclear-armed powers are fighting each other again. Unclear international communication could literally get us all killed. The aesthetics of everyone having their own lingo is not worth the risk.

lolinder 3y ago
> this seems like a terrible reason to silo huge chunks of humanity from each other That's why I set that argument aside.
Turing_Machine 3y ago
There is a difference between "language with the most native speakers" and "language that is the most widely spoken". Mandarin may win the first one, but fails the second one dramatically. Few people outside China speak Mandarin as their first language. Mandarin has L1: 940 million, L2: 200 million -> 1.14 billion English has L1: 380 million L2: 1.077 billion -> 1.46 billion Also, most of the L2 Mandarin speakers live in China, with another Chinese dialect as their native language.

dumbfounder 3y ago

I appreciate your point about generative AI exacerbating the dominance of major languages. It's a nuanced issue, indeed. While it's true that AI reflects the linguistic data it's fed, this doesn't necessarily mean it only perpetuates these languages without offering potential benefits to less-represented ones. Consider the dual nature of technology: it can indeed centralize or decentralize. For less-spoken languages, AI and machine learning can be leveraged to preserve and revitalize them. Projects like Google's Woolaroo (https://artsandculture.google.com/project/woolaroo https://artsandculture.google.com/project/woolaroo), using machine learning to help preserve endangered languages by translating them into more widely spoken ones, hint at potential pathways forward. While generative AI may currently favor dominant languages due to the sheer volume of data available, there’s growing awareness and effort to make these technologies more inclusive. This includes not only better representation in training data but also the development of tools specifically aimed at language preservation.

lolinder 3y ago
Agreed! And I hope we continue to see such efforts to address language representation. I don't think hope is lost, far from it, I'm just trying to answer OP's question about why reflecting the current state of languages can be a problem.

jstummbillig 3y ago

> Yes, it's continuation of the same pattern we've been seeing for decades, but I think it's fair to say that generative AI will escalate that pattern dramatically unless specifically addressed. It's cool that it can be addressed though, right? AI can be both, the fast track to escalating the issue and the fast track to deescalating the issue. Interesting.

jhbadger 3y ago

>Generative AI only works for the most popular languages on the internet I'm not so sure about that. I was amused when playing with ChatGPT shortly after it was public that it could answer questions in Esperanto and even write the silly sorts of stories in it than it can in English.

BlueTemplar 3y ago

Not decades, centuries (at least). A synonym for "AI" is machine : consider what effect the printing press (a written word-replication machine) had on less popular languages ? (It would be interesting to compare Europe and Asia...) This article is just desperately lacking in awareness like this - see also how, frustratingly, it barely mentions how a language is spoken first, and only maybe written second, but doesn't spend much time on that aspect.