I've made a good amount of "enemies" debunking common bullshit believed about technology at table discussions. People just love believing we already have SkyNet but our governments have it by the balls, and the utopian future is juuuuust around the corner.
They believed that for _decades_ now. Wish I was joking.
But the modern AI / ML folk is very much to blame as well. Every 0.01% improvement in cat recognition NNs is hailed as the solution to all our problems.
(Not to mention peer review rings where AI "papers" with zero reproducible experiments get applauded and published with almost zero scrutiny. There were a few articles about it on HN as well.)
All of that certainly doesn't help for the general public to have the right info.
The public discourse about the capabilities of technology really needs a bit more conservative revisit that must temper expectations. Especially having in mind that a lot of technology actually can achieve much more than it does right now... but is stymied by politico-economical and social factors, like 90% of the time.
I am sure that wishful thinking / finding what you're looking for is a an occupational hazard for people in the field. I would guess that for everyone, AI researchers or not, there's an element of pareidolia[1], humans are always on the lookout for someone to talk to...
I took a look at your comment history because I became curious if you yourself were perhaps an AI being tested on Hacker News. I noticed you said in another thread that you might have had enough of life. I doubt a non-sentient AI would say that, although Douglas Adams made famous a character he created called Marvin the depressed robot.
I think depression in a person is a good test of sentience. You might be interested in Kurt Vonnegut's "Breakfast of Champions" (not the movie with Bruce Willis, which is an abomination) in which he feels everyone but himself is a robot and considers killing himself until he meets an artist who restores his faith in the creative spark and soul of humanity.
Sentience is a scale anyway and apparently Elon Musk gets a low score, meaning he appears to be a bot - to Twitter at least
> I noticed you said in another thread that you might have had enough of life. I doubt a non-sentient AI would say that
A non-sentient AI trained on a corpus of text conversations between depressed people would almost certainly say that.
Remember a few years ago when some company trained their natural language AI using conversations collected from the internet? And the non-sentient AI turned out to be racist, sexist, and generally rude?
Depression in my experience encompasses an entire world view. P.K. Dick says in Exegesis that readers have pointed out he keeps creating the same world over and over in his books, so he concludes that world must be him. That's who he is. Some philosophers create their philosophy out of their particular world view, which is why it doesn't resonate with me.
tl;dr - A depressed AI would have to create an entire world view of thought to be convincing, not just mimic what it read in comments
Because they _want_ it to be true.
I've made a good amount of "enemies" debunking common bullshit believed about technology at table discussions. People just love believing we already have SkyNet but our governments have it by the balls, and the utopian future is juuuuust around the corner.
They believed that for _decades_ now. Wish I was joking.
But the modern AI / ML folk is very much to blame as well. Every 0.01% improvement in cat recognition NNs is hailed as the solution to all our problems.
(Not to mention peer review rings where AI "papers" with zero reproducible experiments get applauded and published with almost zero scrutiny. There were a few articles about it on HN as well.)
All of that certainly doesn't help for the general public to have the right info.
The public discourse about the capabilities of technology really needs a bit more conservative revisit that must temper expectations. Especially having in mind that a lot of technology actually can achieve much more than it does right now... but is stymied by politico-economical and social factors, like 90% of the time.
I am sure that wishful thinking / finding what you're looking for is a an occupational hazard for people in the field. I would guess that for everyone, AI researchers or not, there's an element of pareidolia[1], humans are always on the lookout for someone to talk to...
1: https://en.wikipedia.org/wiki/Pareidolia
It's IMO a combination of many factors, yours and mine included.
We always kind of live in the future. It's both stupid/naive and kind of charming. :)
I've found no evidence that other people are sentient, I have severe doubt that I myself am sentient.
That said, current AI lacking sentience does not guarantee that it impossible for future AI to become sentient.
I took a look at your comment history because I became curious if you yourself were perhaps an AI being tested on Hacker News. I noticed you said in another thread that you might have had enough of life. I doubt a non-sentient AI would say that, although Douglas Adams made famous a character he created called Marvin the depressed robot.
I think depression in a person is a good test of sentience. You might be interested in Kurt Vonnegut's "Breakfast of Champions" (not the movie with Bruce Willis, which is an abomination) in which he feels everyone but himself is a robot and considers killing himself until he meets an artist who restores his faith in the creative spark and soul of humanity.
Sentience is a scale anyway and apparently Elon Musk gets a low score, meaning he appears to be a bot - to Twitter at least
https://arstechnica.com/tech-policy/2022/08/twitter-says-mus...
> I noticed you said in another thread that you might have had enough of life. I doubt a non-sentient AI would say that
A non-sentient AI trained on a corpus of text conversations between depressed people would almost certainly say that.
Remember a few years ago when some company trained their natural language AI using conversations collected from the internet? And the non-sentient AI turned out to be racist, sexist, and generally rude?
Depression in my experience encompasses an entire world view. P.K. Dick says in Exegesis that readers have pointed out he keeps creating the same world over and over in his books, so he concludes that world must be him. That's who he is. Some philosophers create their philosophy out of their particular world view, which is why it doesn't resonate with me.
tl;dr - A depressed AI would have to create an entire world view of thought to be convincing, not just mimic what it read in comments
No one thought Tay was a reincarnated Nazi
https://en.wikipedia.org/wiki/Tay_(bot)
Because they don't know what AI is.
https://archive.ph/H5Ibt