Federal Extreme Vetting Plan Castigated By Tech Experts (apnews.com)

An anonymous reader shares an Associated Press report: Leading researchers castigated a federal plan that would use artificial intelligence methods to scrutinize immigrants and visa applicants, saying it is unworkable as written and likely to be "inaccurate and biased" if deployed. The experts, a group of more than 50 computer and data scientists, mathematicians and other specialists in automated decision-making, urged the Department of Homeland Security to abandon the project, dubbed the "Extreme Vetting Initiative." That plan has its roots in President Donald Trump's repeated pledge during the 2016 campaign to subject immigrants seeking admission to the United States to more intense ideological scrutiny -- or, as he put it, "extreme vetting." Over the summer, DHS published a "statement of objectives" for a system that would use computer algorithms to scan social media and other material in order to automatically flag undesirable entrants -- and to continuously scan the activities of those allowed into the U.S.

Comments

The perfect is the enemy of the good

[Anonymous Coward]

"It won't be perfect, so lets scrap the whole thing!"

Re:The perfect is the enemy of the good

[rtb61]

Yes, because the greater the imperfection, the more likely you will pursue empty investigations whilst the real criminals further down the list, carry out their crimes. Have you failed to notice how often they have information about terrorists and failed to act. That is because they wasted resources upon empty investigations and did not get to the necessary ones. Add to that performance based investigatory demands. You must get results, you must profit, so they take the easy way out, the profitable way out, find some nutter and spend months stitching them up for a guaranteed prosecution, it is the corporate for profit way (that the nutter would have done nothing without intervention meaningless in a for profit world, it was a cheap investigation with a guarantee of success much easier than all those other hard investigations that you get blamed for when you don't deliver results). This is the reason why US security so routinely fails, why it's successes upon deep scrutiny are not successes at all, why performance based is a stupid failure, why an attitude of maximise profits not maximise results kills. You know what will happen, the contractors will pay lobbyists who will corrupt force through the failed programs to generate profits and basically fuck the results, they don't care, it was profitable.

Re:The gist of their argument

[sycodon]

As far as we know, no one has ever defined or quantified these characteristics

1. Can Read and write.
2. Speaks English and do #1 in English.
3. A degree of some kind?
4. Lack of association with any radical groups.
5. A US citizen who will sponsor them.