Hacker News Reader

>Societal risks, such as bias and discrimination

Absolutely unacceptable that pressure is coming from the government to change AI outputs to meet criteria that is so subjective, like the output not displaying "bias and discrimination", and which furthermore, absolutely doesn't meet the bar for a national security threat that justifies government intervention in software development.

View on HN

> absolutely doesn't meet the bar for a national security threat that justifies government intervention in software development.

The idea that the only basis for government action is “national security threat” is... very far from a consensus view.

Government prohibitions, and the threat of them, being applied to software development, requires that criteria being met

There are decades of well defined laws (and case law) regulating discrimination in hiring, housing, commerce, etc. Why should software be treated any differently?

Because it's speech.

Reply on news.ycombinator.com