Shortly after going to the doctor about wax in my ears I was ad-bombed with online hearing remedies - is that not a Hippocratic oath that's being wrecked?
Reminds me of a discussion group mad at a story-writing ai because had issues trying to write lgbtq+ romance stories. Forgetting that there was almost no lgbtq+ story data in the training data. Or another group that found a story about a vampire that assaults children in ai dungeon, Of course, they found it during the company having a fallout over strict policies and a huge database leak.
I think certain AI applications will always be co-pilot with humans, until we have superhuman AGI anyway. The black box is still an issue, but as you pointed out, we've got methods to shine a light in the box, and those methods will only get better. Even with a black box, if the AI reveals some fundamental fact without showing its work, knowing the fact is true will allow us to get there on our own eventually... assuming we even care to do so.
I think there is a lot to be gained by using AI. I have published a paper where we used AI to improve some analysis that we were doing. I think the concern is the speed of adoption might lead to problems that are hard to undo.
AI (machine learning) is turning out to be a kind of Pandora's box. Especially considering the integration of AI and the human brain. This combination is a black box within a black box. There is proud implications for good as well as harm.
What is needed is to move from a MONETARY based civilization to a RESOURCES based one (the venus project), that's the only way to avoid AI to be corrupted by the enterprises interests, but also the only way to avoid the same on humans.
Well, just because we invented typewriters doesn't mean there aren't people with excellent handwriting. Its still a skill you can master, calligraphy and lettering...
Why don't we prepare AI for us, it should remain a tool to us and nothing more, if it is racist, it simply can't be implemented until proven otherwise. We cant just give ourselves away to AI just because its 'soo good' at what it does, it clearly is not that 'good' if it has such bad flaws.
With the hiring example we can conclude AI is stupid :) It just picks up every word and give weights. Like you give it CVs and hiring statistics, and it says "oh, so you guys don't want women to be hired, I solved it". But it's just a wrong conclusion, the AI is just silly :)
I am not sure that it will replace teacher completely. I think we might see a shift in the role of teachers as we shift to use AI for explanations. Perhaps teachers will be there for social teaching and caring.
Shortly after going to the doctor about wax in my ears I was ad-bombed with online hearing remedies - is that not a Hippocratic oath that's being wrecked?
Reminds me of a discussion group mad at a story-writing ai because had issues trying to write lgbtq+ romance stories. Forgetting that there was almost no lgbtq+ story data in the training data. Or another group that found a story about a vampire that assaults children in ai dungeon, Of course, they found it during the company having a fallout over strict policies and a huge database leak.
Immortality would be terrifying. Well, on the bright side that means life could continue on forever in multiple multiple universes.
I think certain AI applications will always be co-pilot with humans, until we have superhuman AGI anyway. The black box is still an issue, but as you pointed out, we've got methods to shine a light in the box, and those methods will only get better.
Even with a black box, if the AI reveals some fundamental fact without showing its work, knowing the fact is true will allow us to get there on our own eventually... assuming we even care to do so.
I think there is a lot to be gained by using AI. I have published a paper where we used AI to improve some analysis that we were doing. I think the concern is the speed of adoption might lead to problems that are hard to undo.
AI (machine learning) is turning out to be a kind of Pandora's box. Especially considering the integration of AI and the human brain. This combination is a black box within a black box. There is proud implications for good as well as harm.
What is needed is to move from a MONETARY based civilization to a RESOURCES based one (the venus project), that's the only way to avoid AI to be corrupted by the enterprises interests, but also the only way to avoid the same on humans.
A very hard transition for us to make.
@@ScienceDiscussed the question is, we will do it the good way or the hard way? :)
haha probably the hard way.
Well, just because we invented typewriters doesn't mean there aren't people with excellent handwriting. Its still a skill you can master, calligraphy and lettering...
Why don't we prepare AI for us, it should remain a tool to us and nothing more, if it is racist, it simply can't be implemented until proven otherwise. We cant just give ourselves away to AI just because its 'soo good' at what it does, it clearly is not that 'good' if it has such bad flaws.
Very important video
Glad you think so!
With the hiring example we can conclude AI is stupid :) It just picks up every word and give weights. Like you give it CVs and hiring statistics, and it says "oh, so you guys don't want women to be hired, I solved it". But it's just a wrong conclusion, the AI is just silly :)
yeah exactly. Bad data = bad AI
I keep A.I. away, by not creating an account. 😂
Ai will replace teachers in school
I am not sure that it will replace teacher completely. I think we might see a shift in the role of teachers as we shift to use AI for explanations. Perhaps teachers will be there for social teaching and caring.
Garbage in, garbage out. There, I saved you 12 minutes of watching this video.
Garbage in garbage, garbage out. The 'human' programmer is the source. No difference between the analog, digital, or the quantum.