SoylentNews is people
SoylentNews
A sobering message about the future at AI's biggest party
Blaise Aguera y Arcas praised the revolutionary technique known as deep learning that has seen teams like his get phones to recognize faces and voices. He also lamented the limitations of that technology, which involves designing software called artificial neural networks that can get better at a specific task by experience or seeing labeled examples of correct answers.
"We're kind of like the dog who caught the car," Aguera y Arcas said. Deep learning has rapidly knocked down some longstanding challenges in AI—but doesn't immediately seem well suited to many that remain. Problems that involve reasoning or social intelligence, such as weighing up a potential hire in the way a human would, are still out of reach, he said. "All of the models that we have learned how to train are about passing a test or winning a game with a score [but] so many things that intelligences do aren't covered by that rubric at all," he said.
(Score: 2) by JoeMerchant on Sunday December 15 2019, @04:03PM (4 children)
You must not see the same candidate pools I see. While it's easy to criticize bad hires after the fact, it's much harder to screen for those "latent defects" in the interview process. However, the interview process is still valuable- even with its high rate of false positives and false negatives, at least for out hires I think we've got better than 50 percent true negative screening.
🌻🌻 [google.com]
(Score: 1, Interesting) by Anonymous Coward on Sunday December 15 2019, @09:44PM (1 child)
Having hired and interviewed countless people in tech, I'm actually a big believer of psychometric tests over tech tests and/or face-to-face interview. It tells me so much more about how the propeller head is wired inside which is a lot more valuable then any tech exam results or interview. Even face-to-face soft question interviews don't cut it as a smart enough cookie can easily say what you want to hear.
A psychometric exam on the other hand asks the same questions 10 different ways and deducts the person's psychological makeup. I've ignored psych results in the pass to only learn the hard way that I shouldn't.
(Score: 2) by JoeMerchant on Monday December 16 2019, @02:01AM
Actually, it tells you about how the propeller head fills out a psychometric test. By age 12, any decent propeller head is intelligent and insightful enough to ask (and answer) "what are they looking for with this?"
🌻🌻 [google.com]
(Score: 2) by bzipitidoo on Sunday December 15 2019, @11:24PM (1 child)
So, what's a typical situation, 100 candidates per opening, but over half of them can be dismissed as resume spammers who are painfully obviously totally unqualified? Like the person applying for a programming job despite never having written a line of code in anything, ever, not even a formula in a spreadsheet?
So, let's say weeding out the obvious stuff knocks it down to 20 candidates. That's the point at which I was thinking that choosing at random would do better than applying messed up and widely disproven criteria that some crazy employers still hold dear in defiance of all evidence and sanity.
(Score: 2) by JoeMerchant on Monday December 16 2019, @02:11AM
HR does terrible things on the front end, not only matching (mostly meaningless) resume items to overly restrictive job descriptions, but also juicing the field to skew "irrelevant" parameters like race and sex toward more desirable ratios. What this ends up doing is freezing out lots of qualified candidates while we are required to interview more people of desirable counter-discrimination profiles.
For positions of any importance, we will typically put at least 4 candidates through the process with 5-6 face to face interviews each, then get together and compare notes. Out of the pool of 4, there are almost always at least two who get the universal head-shake no for various reasons. About half the time, we're not happy with any of them and go back to HR for more.
What's sad is when the weak ones are let through for various reasons. We hired a very personable engineer, enthusiastic, active in the community, fun to be around, but even in the interview they were clearly weak in technical execution abilities and if they haven't gotten those chops by the end of a Master's degree, they are unlikely to learn on the job. Anyway... fast forward two years and we've transferred them off to another division where they can hopefully contribute as a smaller player in a bigger team, we're just not big enough to take up that kind of slack.
🌻🌻 [google.com]