Hacker News Reader

Comment by neom | Hacker News Reader

Yah, so there is a problem here in that many of these "homeland nationalistic" folks who send stuff back to "their governments" are in many instances, citizens of the country they are betraying.

Since these two[1][2] incidents, I wondered deeply for a while, what is going to happen here as it seems the CCP is getting more and more brazen. I don't believe it would be legal for "enhanced background checks" on someone in the hiring process because they are Chinese American, and it opens an awful door regardless. I don't really know the solution, but I've thought a lot about the problem. (And yes, of course there is that 5 eyes/other "Western" powers are almost certainly doing similar stuff)

[1] https://japantoday.com/category/world/2-us-navy-sailors-arre...

[2] https://www.justice.gov/opa/pr/harvard-university-professor-...

View on HN

>I don't believe it would be legal for "enhanced background checks" on someone in the hiring process because they are Chinese American,

It absolutely would be legal if AI was recategorized as ITAR. In 2022, dual citizenship was cracked down on in ITAR regulations and dual(or more) citizen individuals must now be deeply vetted for contact with ITAR Section 126.1 countries which include China.

I am by no means an expert at all, however my very basic understanding is that ITAR is primarily concerned with important/export controls? Even if AI was added but the specific staffing was to a project was totally unrelated to something that would be imported or exported (this guy was an infra guy it looks like?), would that apply?

I'm no expert either, but from living through the painful days of ITAR and encryption, it only matters that some technology/information can be used for military purpose for it to be restricted. For years we couldn't distribute the code for AES above 128 due to ITAR, even though it was widely and easily available all over the internet.

If you'd care to opine, would you then there agree with delfinom that ITAR in AI could be a solution to this? From what you just described, it made me wonder if it would impose undo burden on, and so hamper, the emergence of the next generation of NN/ML tools ("AI").

There's considerable nuance involved in the question, which I haven't done due diligence on, but with that as a disclaimer...

Thinking out loud, it does seem like it could help, although it does place a significant (at times) burden on people affected, and in the case that I worked with it made no practical difference whatsoever at preventing the use of the technology by US adversaries. So yes I think it might could help, but my inclination is away from it as I think it would ultimately be ineffective yet quite burdensome on US companies. Possibly to the point of backfiring, leading to a country like China that is (so far at least) much more "open" in regards to AI/ML.

Do you have any thoughts?

I have such limited knowledge of this stuff it's all just gut, but I'd be aligned with you when pushed to comment. The best solution is probably no solution, it's going to happen and that's that.

At the end of the day, do we care about Google trade secrets or US national security?

Thoughts on subjecting Jewish-Americans to this?

We've lost reems of technology to I*aeli espionage.

Reply on news.ycombinator.com