By continuing to use the site, you agree to our use of cookies and to abide by our Terms and Conditions. We in turn value your personal details in accordance with our Privacy Policy.
Please log in or register. Registered visitors get fewer ads.
General census of opinion..... on 19:45 - Jul 27 by nodge_blue
AI coupled with quantum computing means nothing is secure. And someone will try to exploit it.
Thats the trouble with humans, as much as it all could do good, someone or some state will use it as an attack method.
That aside, its also troubling that AI is doing stuff that the programmers are surprised at. Effectively the very clever algorithms give it a degree if making its own choices based on what it finds. That to all intents and purposes is AI. I think it is more than machine learning as i think its not making decisions within set parameters but has more autonomy. I read somewhere that one of them taught itself a language when it hadnt been instructed to. That shows more a desire or reason to do something rather than just a machine learning patterns.
I think these early stages of AI are fairly safe but i wouldnt say that there wont be problems along the way.
[Post edited 27 Jul 2023 19:50]
...and I wouldn't say that I'm confident that we'll see those problems coming for much more than the blink of an eye. Whether that's a metaphorical blink or an actual blink, I'd have to ask a chatbot.
in the hands of good people who share our values and have perfect foresight it could be very good. in the hands of bad people and people we disagree with and idiots it could be very bad. i hope that's not too technical.
And so as the loose-bowelled pigeon of time swoops low over the unsuspecting tourist of destiny, and the flatulent skunk of fate wanders into the air-conditioning system of eternity, I notice it's the end of the show
0
General census of opinion..... on 08:30 - Jul 28 with 583 views
General census of opinion..... on 23:33 - Jul 27 by lowhouseblue
in the hands of good people who share our values and have perfect foresight it could be very good. in the hands of bad people and people we disagree with and idiots it could be very bad. i hope that's not too technical.
That is far too narrow too. In comp sci, you don't always know if you're doing a bad thing until after the fact.
General census of opinion..... on 09:31 - Jul 27 by StokieBlue
I think your question is too nebulous.
There are huge differences between ML systems, for instance a system designed to spot cancer much earlier would be a huge benefit whilst a system used to generate deepfakes would be a huge problem.
The main issue with AI over the short to medium term is going to be it's affect on jobs and what happens around that. Hopefully some form of levy can be implemented for each AI instance used and that can be used to partially fund something like UBI.
SB
New article on a study which shows one of the positive use cases of AI from my second paragraph:
General census of opinion..... on 07:11 - Jul 27 by ElephantintheRoom
Not sure it has any intelligence - any more than a clock knows what time it is.
"There was an exchange on Twitter a while back where someone said 'What is artificial intelligence?' and someone else said 'A poor choice of words in 1954'." "And you know, they're right. I think if we had chosen a different phrase for it, back in the '50's we might have avoided a lot of the confusion that we're having now." Ted Chiang, Sci Fi writer, Financial Times, June 3rd 2023
0
General census of opinion..... on 09:14 - Aug 2 with 439 views
General census of opinion..... on 07:42 - Aug 2 by JohnTy
"There was an exchange on Twitter a while back where someone said 'What is artificial intelligence?' and someone else said 'A poor choice of words in 1954'." "And you know, they're right. I think if we had chosen a different phrase for it, back in the '50's we might have avoided a lot of the confusion that we're having now." Ted Chiang, Sci Fi writer, Financial Times, June 3rd 2023
The basic of basics here for the AI argument among peers is that for AI to be used correctly, something has to know why it is doing something or what something is used for.
For example
AN ML algorithm could tell you that most of these are chairs via pattern recognition
It might be able to tell you that chairs are for sitting (if taught that). Though can the program determine why one sits? How one would sit in the different chairs?
So compare that to humans who are said to be intelligent. We mostly will look at a chair and our heads will think "That's a place to sit" rather than "This is a chair". We can then visualise that and then debate if we need or want to sit.
The machine learning result would be "This is a chair" or "This is not a chair" and that is the end of it.
This is more or less the same argument that has existed since the 50s. Interestingly though, the supporters of the AI term would look towards the Turing test where one has to establish whether or not the entity we're talking to is a human mind or an algorithm. If you can not determine between the two, the algorithm has beaten the Turing test. Given that chat bots can machine learn language, nuance and sentence structure it is increasingly likely that ML algorithm convincingly pass the Turing test. As a result, the concept and term of Artificial Intelligence is very popular again as the consumer is more likely to accept a technology if it is called "Smart" or "Intelligent".
So, while it was a poor choice of words in the 1950s, for consumerism it is actually a really clever choice of words in beating the distrust of technology without anthropomorphic features (such as robots with faces).
General census of opinion..... on 09:14 - Aug 2 by wkj
The basic of basics here for the AI argument among peers is that for AI to be used correctly, something has to know why it is doing something or what something is used for.
For example
AN ML algorithm could tell you that most of these are chairs via pattern recognition
It might be able to tell you that chairs are for sitting (if taught that). Though can the program determine why one sits? How one would sit in the different chairs?
So compare that to humans who are said to be intelligent. We mostly will look at a chair and our heads will think "That's a place to sit" rather than "This is a chair". We can then visualise that and then debate if we need or want to sit.
The machine learning result would be "This is a chair" or "This is not a chair" and that is the end of it.
This is more or less the same argument that has existed since the 50s. Interestingly though, the supporters of the AI term would look towards the Turing test where one has to establish whether or not the entity we're talking to is a human mind or an algorithm. If you can not determine between the two, the algorithm has beaten the Turing test. Given that chat bots can machine learn language, nuance and sentence structure it is increasingly likely that ML algorithm convincingly pass the Turing test. As a result, the concept and term of Artificial Intelligence is very popular again as the consumer is more likely to accept a technology if it is called "Smart" or "Intelligent".
So, while it was a poor choice of words in the 1950s, for consumerism it is actually a really clever choice of words in beating the distrust of technology without anthropomorphic features (such as robots with faces).
[Post edited 2 Aug 2023 9:20]
I don't think that chatbots can pass the Turing test even, although they may fool some of us.
1
General census of opinion..... on 09:23 - Aug 2 with 418 views
General census of opinion..... on 09:19 - Aug 2 by JohnTy
I don't think that chatbots can pass the Turing test even, although they may fool some of us.
Good catch, the Turing test is certainly a bit more evolved than an individual's perception on a whim. I do often think what the result of Kasparov's chess game vs a computer would have been if Kasparov had no idea the moves were being generated by an algorithm.
In a Humanities Department at a University and we've spent a fair bit of time preparing for students trying to cheat using ChatGPT et al. We've not had any significant cases yet (as it stands all our students can out-perform a AI language-models with not much effort) but it's coming.
My perception is that as things stand it'll be a big technological leap in a multitude of economic sectors, where it'll significantly improve productivity but also haphazardly displace a lot of workers and cause significant social problems. I've very little faith these days that anything significant will be done to mitigate that. Also likely that it gets put to a whole bunch of unethical purposes and cause a bunch of unanticipated problems that no-one will have the will or knowledge to regulate properly.
By continuing to use the site, you agree to our use of cookies and to abide by our Terms and Conditions.
