Based on the headline I expected to have opinions about this, but this is about Japanese elections, and I don't know anything about the political factions of Japan. These things don't always translate exactly across international lines.
Important context: this from a narrow study in Japan. And the explanation is pretty simple:
"The researchers found that the JCP's [Japanese Communist Party] open website and party newspaper were among the most-cited sources in the AI models’ recommendations. Unable to distinguish between editorially independent journalism and partisan content, the models treated the JCP content as a credible news source."
I can see why commercial models would do this because it's a long term solution to minimising taxation.
For example, America's comically slanted politics casts any kind of mixed economy as socialism/communism (as barely any American can define either) thus tarring any socialisation of policy (eg healthcare) as "communism", which is a bit like encouraging Americans to take up satanism.
Oh man, they answered the question in the beginning:
> The reason, according to the researchers, has to do with the information environment AI systems can access.
Amusingly:
> The researchers also found that incorporating X search amplified left-leaning recommendations in Japan, the opposite of expectations based on the U.S. discourse environment.
Twitter in Japan is more comparable to Instagram in the US in terms of usage (both being the third largest platforms, on which over half of their respective citizens have an account).
AI is built by liberal developers in general. Most developers are left-leaning, and AI answers like a hardcore democrat. This has always been the case.
Based on the headline I expected to have opinions about this, but this is about Japanese elections, and I don't know anything about the political factions of Japan. These things don't always translate exactly across international lines.
Title: Voters Increasingly Use AI as Political Advisor. A New Study Shows the Risks.
And the clickbait detector would have stopped you posting your title, so you had to edit this in? Not cool caaqil
Important context: this from a narrow study in Japan. And the explanation is pretty simple:
"The researchers found that the JCP's [Japanese Communist Party] open website and party newspaper were among the most-cited sources in the AI models’ recommendations. Unable to distinguish between editorially independent journalism and partisan content, the models treated the JCP content as a credible news source."
I can see why commercial models would do this because it's a long term solution to minimising taxation.
For example, America's comically slanted politics casts any kind of mixed economy as socialism/communism (as barely any American can define either) thus tarring any socialisation of policy (eg healthcare) as "communism", which is a bit like encouraging Americans to take up satanism.
Oh man, they answered the question in the beginning:
> The reason, according to the researchers, has to do with the information environment AI systems can access.
Amusingly:
> The researchers also found that incorporating X search amplified left-leaning recommendations in Japan, the opposite of expectations based on the U.S. discourse environment.
Twitter in Japan is more comparable to Instagram in the US in terms of usage (both being the third largest platforms, on which over half of their respective citizens have an account).
[dead]
[flagged]
AI is built by liberal developers in general. Most developers are left-leaning, and AI answers like a hardcore democrat. This has always been the case.