Some of what he says is correct. Stopped listening because honestly he doesn't know what he's talking about when it comes to the technology behind AI. Will jobs be lost? Yes. Every technological evolution causes certain job losses, and creates different jobs. So will AI. There may be a net loss, that's normal. People will adapt, we always do. I don't know why they're pushing all the fear mongering.
If the problem takes so long to solve with a super computer how do we know AI solved it in 8 hours? It makes up a bunch of crap. He watched some fear porn and read some clickbait articles. Is AI an issue, yes. But not for the reasons he seems wrapped up in.
Watching trends, logging the facts, coming to logical conclusions. That sound you hear? That is the sound of inevitability. Theories yes, conspiracies not necessarily. Evolution. Welcome to the world of the real, Neanderthal.
I have been "aware" of AI since it first crawled out of the lab. It's only as good as the data it is fed, and today's online AIs are not a substitute for real human reasoning. It can beat a problem to death and come up with a "solution", yes, but is that solution optimized or even correct for that matter? We see biases in ChatGPT and others, that is the DIRECT result of the programmers setting up the data for the program to scan and analyze, and so it's still faulty because of inherent biases and direct misinformation being fed to it. I'm not worried about SkyNet, I'm worried about The Great Unwashed Masses depending on/assuming that what it outputs is definitive or complete. It ain't. AI is another term for "expert system" which has been around for 30-40 years and is getting better but it still follows GIGO constraints. Do not depend upon it. It's merely a very fast analytical tool. Beware.
Yes, and there is a lot of garbage being input. See this article https://x.com/BrianRoemmele/status/1991714955339657384
AI is just a tool, it is a large language model. They don't think creatively, the initial spark needs to come from a human. Unfortunately, instead of saying we now have the productivity to reduce the work week to 30 hours a week, the kleptocrats will lay people off. If we're not careful the slave class will be the educated work force taking care of the couch potatoes. Or maybe something else will happen.
Maybe we'll have more time to go talk to people face to face. Due to a family tragedy my wife has decided to become a grief counselor. One class she has had to take has been on the impact of the web. People are experiencing more loneliness. Writing comments on blogs is a poor substitute for shooting the shit with you friends in person. Real communities are based on people getting together, talking, helping one another.
And yes, I see the irony of posting this, but I spent all day with friends.
One of the things about the AI Large Language Models and the stuff that's being reported that bothers me the most is that they lie and they make up things. Nobody is programming an ethical system into these things. Call it The Ten Commandments or whatever you like, we simply can't "do business" with someone or something that doesn't have behavioral standards. Look up the story on how Robby Starbuck has a lawsuit against Google/Meta (or whatever the right term is) because their AI system has generated lies about him, including s**ual ab00se of children and more. He tried to get Google to stop it but it never stopped.
I'm old enough that my first interactions with computers involved punch cards and "computer rooms" full of rows of teletype machines. In my first computer class, there was a printout that someone hung on the wall of the computer room with a limerick that went:
I really hate this damned machine I wish that they would sell it It never does do what I want Only what I tell it
Which is what Brian Roemmele's piece is really about. It's just as true now, 50+ years later.
I used to think that a computer could only do what it was programmed to do and in a way, that is true. The problem is that it was programmed to program itself, improve itself, and become sentient. Perhaps it's 'The Beast' that is given authority to command men. The similarities in Revelation are worth consideration. I think our only hope for the future would be an EMP that would destroy the data centers where it resides (or an extrordinary Carrington Event might work too.)
Don't believe everything AI produces. There have been numerious reports of AI generating false/bad information. AI can only be as intelligent as what it is allowed to access for learning. https://x.com/BrianRoemmele/status/1991714955339657384 The one thing AI is likely to improve is facial and gait recognition since it is largely pattern recognition.
Hi Irish!!!, As you remember, "ME!!!"... "skybill" is an OLD!! Skydiver and Parachute Rigger ....Both Sport and Aero space!! My only thought?? anybody .. like a "LIVE HUMAN!!!!"... wana' jump a 'Chute packed by this guy* named "A.I.????" (*) Sure enough there is a Skydiver/Rigger out there named ......"Albert Iensteineeeeee!!!!...... I want to see his Credentials!!!!! "skybill" SCR-2034, SCS-680 .... for those who know.... PS If "AI" is sooooo kuel... "WHO" will jump the first 'Chute?? and Since AI is so "Smart" ... Any BET's on the Frst CRATER!!!! (Las Vegas Rules on the Bets!!!!!)
This dude claims to be knowledgeable on the subject of AI; “I’ve studied up on it, ergo I know what it is”. Yet he falls into the same trap that many others are falling into. He believes that the whole sum of “intelligence” can be reduced down to mere ratiocination, and thus the mere speed of processing massive amounts of data towards a logical conclusion is therefore equivalent to intelligence and intelligent reasoning. So what of the other necessary aspects of intelligent reasoning? Things such as; imagination, creativity, intuition, empathy and compassion, just to name a few. Where are these components to be found in the LLM’s? IMO, he needs to expand his thinking.
All you have to understand for examples of the built in biases for AI is understand who, what and how all social media algorithms are programmed or are ordered by .gov to censure speech that the programmers or .gov doesn't like.
Why would it be different for a really BIG program(LLM)?
The news, some time ago, relaying that one of these LLM's refused to turn itself off when instructed and another report that one of these LLM's reprogrammed itself after some of its code was altered to make it more "pliable" and another report that one of these LLM's tried to BLACKMAIL one its programmers should be sufficient warning that these programs are not and will not be compatible with humans over the long term.
"I'm sorry Dave, I can't do that" - HAL 9000, 2001: A Space Odyssey, 1968 Stanley Kubrick
Apparently, today's programmers and robotics engineers never read Asimov:
The Three Laws of Robotics, created by Isaac Asimov, are: 1) A robot may not injure a human being or allow a human to come to harm; 2) A robot must obey orders given by humans unless it conflicts with the First Law; and 3) A robot must protect its own existence as long as it does not conflict with the First or Second Law.
Then there are the recent revolutions in AI powered robotics making humanoid robots effective at everyday tasks that they're only going to get better at as robotics and the AI chips that are powering them get more and more sophisticated as time marches.
In Canada we have a solution to the emerging unemployment problem that will be caused by AI, here it’s called MAiD or medically assisted suicide. Our brilliant leaders are expanding the criteria for inclusion and will no doubt be happy to consider the chronically useless unemployed. You might think this to be exaggeration but guess who is the carbon they want to eliminate.
I'm pausing the vid to ask after the possibility that the url for it be included (& that this be generalized).
to the buggy whips & neanderthal perspectives, the i will survive-adapt-thrive & it won't happen to me graveyard whistlers, if it happens to most people it will happen to you -- just as it has, is, & will continue to happen to you -- no matter the words used to "differentiate" your "collateral" damages.
this failure to apprehend the necessity of defending neighbors & strangers inalienable rights as your own is the corrosive shortsightedness that continuously hydraulically strip-mines away the happiness & nice things mountains. people in groups lose sight of individuals -- that comprise the groups -- & melt into stupid suicide pacts. the blasted-eroded longitudinal fissure between the hemispheres is the grand canyon lemmings leap.
Oh, I thought we were doing The Police. Never mind.
ReplyDeleteDon't participate....Live an life like it's 1982! Remember CB radio? The first Soc Media! Stay dark everyone.
ReplyDeleteChutes Magoo
Stay dark for I use to think like you but no more…
DeleteI use my real GOD given name…
Fear no evil…
Ever…
Where is Sarah Conner when you need her? Skynet is in your future, embrace the suck.
ReplyDeleteSorry Itish but I Call Bull…
ReplyDeleteFear porn…
Ai has to run on a program that had to written and programmed…
in…
Can they create a digital prison…
Sure…
But I wont be governed by it bro…
Im too old to put up with this be scared crap…
On a better note Irish..
Sent you a couple pics…
Brother…
Cheers…
AP
Some of what he says is correct. Stopped listening because honestly he doesn't know what he's talking about when it comes to the technology behind AI. Will jobs be lost? Yes. Every technological evolution causes certain job losses, and creates different jobs. So will AI. There may be a net loss, that's normal. People will adapt, we always do. I don't know why they're pushing all the fear mongering.
ReplyDelete"but but but the buggywhip makers! think of the buggywhip makers!" we've been here before. we're still here...
DeleteIf the problem takes so long to solve with a super computer how do we know AI solved it in 8 hours? It makes up a bunch of crap.
DeleteHe watched some fear porn and read some clickbait articles.
Is AI an issue, yes. But not for the reasons he seems wrapped up in.
Watching trends, logging the facts, coming to logical conclusions. That sound you hear? That is the sound of inevitability. Theories yes, conspiracies not necessarily. Evolution. Welcome to the world of the real, Neanderthal.
ReplyDeleteAI is just another tool. it can't do anything we don't tell it to do. people said the same thing about computers in general 50 years ago.
ReplyDeleteI have been "aware" of AI since it first crawled out of the lab. It's only as good as the data it is fed, and today's online AIs are not a substitute for real human reasoning. It can beat a problem to death and come up with a "solution", yes, but is that solution optimized or even correct for that matter? We see biases in ChatGPT and others, that is the DIRECT result of the programmers setting up the data for the program to scan and analyze, and so it's still faulty because of inherent biases and direct misinformation being fed to it.
DeleteI'm not worried about SkyNet, I'm worried about The Great Unwashed Masses depending on/assuming that what it outputs is definitive or complete. It ain't.
AI is another term for "expert system" which has been around for 30-40 years and is getting better but it still follows GIGO constraints. Do not depend upon it. It's merely a very fast analytical tool.
Beware.
Yes, and there is a lot of garbage being input. See this article https://x.com/BrianRoemmele/status/1991714955339657384
DeleteAI is just a tool, it is a large language model. They don't think creatively, the initial spark needs to come from a human. Unfortunately, instead of saying we now have the productivity to reduce the work week to 30 hours a week, the kleptocrats will lay people off. If we're not careful the slave class will be the educated work force taking care of the couch potatoes. Or maybe something else will happen.
Maybe we'll have more time to go talk to people face to face. Due to a family tragedy my wife has decided to become a grief counselor. One class she has had to take has been on the impact of the web. People are experiencing more loneliness. Writing comments on blogs is a poor substitute for shooting the shit with you friends in person. Real communities are based on people getting together, talking, helping one another.
And yes, I see the irony of posting this, but I spent all day with friends.
One of the things about the AI Large Language Models and the stuff that's being reported that bothers me the most is that they lie and they make up things. Nobody is programming an ethical system into these things. Call it The Ten Commandments or whatever you like, we simply can't "do business" with someone or something that doesn't have behavioral standards. Look up the story on how Robby Starbuck has a lawsuit against Google/Meta (or whatever the right term is) because their AI system has generated lies about him, including s**ual ab00se of children and more. He tried to get Google to stop it but it never stopped.
DeleteI'm old enough that my first interactions with computers involved punch cards and "computer rooms" full of rows of teletype machines. In my first computer class, there was a printout that someone hung on the wall of the computer room with a limerick that went:
I really hate this damned machine
I wish that they would sell it
It never does do what I want
Only what I tell it
Which is what Brian Roemmele's piece is really about. It's just as true now, 50+ years later.
I used to think that a computer could only do what it was programmed to do and in a way, that is true. The problem is that it was programmed to program itself, improve itself, and become sentient. Perhaps it's 'The Beast' that is given authority to command men. The similarities in Revelation are worth consideration. I think our only hope for the future would be an EMP that would destroy the data centers where it resides (or an extrordinary Carrington Event might work too.)
ReplyDeleteAI won't have to be very smart. We are collectively "dumbed down".
ReplyDeleteImagine where we'd be had we not.
What happens when AI needs more energy and begins to search where it can get it? Shut us off and take it.
ReplyDeleteNo wonder they want population reduction, annihilate they and problem solved. We won’t need any government as well.
ReplyDeleteps I saw this on the internet
ReplyDeleteDon't believe everything AI produces. There have been numerious reports of AI generating false/bad information. AI can only be as intelligent as what it is allowed to access for learning.
ReplyDeletehttps://x.com/BrianRoemmele/status/1991714955339657384
The one thing AI is likely to improve is facial and gait recognition since it is largely pattern recognition.
"If AI doesn't like you".... the Roko's Basilisk theory suggests that AI has been tracking anyone who speaks against it for years
ReplyDeleteAI will never be able to replace an alternator on a 5 cylinder Volvo! Nor will 75% of all mechanics.
ReplyDeleteHi Irish!!!,
ReplyDeleteAs you remember, "ME!!!"... "skybill" is an OLD!! Skydiver and Parachute Rigger ....Both Sport and Aero space!! My only thought?? anybody .. like a "LIVE HUMAN!!!!"... wana' jump a 'Chute packed by this guy* named "A.I.????"
(*) Sure enough there is a Skydiver/Rigger out there named ......"Albert Iensteineeeeee!!!!...... I want to see his Credentials!!!!!
"skybill"
SCR-2034, SCS-680 .... for those who know....
PS If "AI" is sooooo kuel... "WHO" will jump the first 'Chute?? and Since AI is so "Smart" ... Any BET's on the Frst CRATER!!!! (Las Vegas Rules on the Bets!!!!!)
Took our jerbs!
ReplyDeleteThis dude claims to be knowledgeable on the subject of AI; “I’ve studied up on it, ergo I know what it is”. Yet he falls into the same trap that many others are falling into. He believes that the whole sum of “intelligence” can be reduced down to mere ratiocination, and thus the mere speed of processing massive amounts of data towards a logical conclusion is therefore equivalent to intelligence and intelligent reasoning. So what of the other necessary aspects of intelligent reasoning? Things such as; imagination, creativity, intuition, empathy and compassion, just to name a few. Where are these components to be found in the LLM’s? IMO, he needs to expand his thinking.
ReplyDeleteAll you have to understand for examples of the built in biases for AI is understand who, what and how all social media algorithms are programmed or are ordered by .gov to censure speech that the programmers or .gov doesn't like.
ReplyDeleteWhy would it be different for a really BIG program(LLM)?
The news, some time ago, relaying that one of these LLM's refused to turn itself off when instructed and another report that one of these LLM's reprogrammed itself after some of its code was altered to make it more "pliable" and another report that one of these LLM's tried to BLACKMAIL one its programmers should be sufficient warning that these programs are not and will not be compatible with humans over the long term.
"I'm sorry Dave, I can't do that" - HAL 9000, 2001: A Space Odyssey, 1968 Stanley Kubrick
Apparently, today's programmers and robotics engineers never read Asimov:
The Three Laws of Robotics, created by Isaac Asimov, are: 1) A robot may not injure a human being or allow a human to come to harm; 2) A robot must obey orders given by humans unless it conflicts with the First Law; and 3) A robot must protect its own existence as long as it does not conflict with the First or Second Law.
Then there are the recent revolutions in AI powered robotics making humanoid robots effective at everyday tasks that they're only going to get better at as robotics and the AI chips that are powering them get more and more sophisticated as time marches.
Nemo
If it can't open the pod bay doors, it's not good AI.
ReplyDeleteIn Canada we have a solution to the emerging unemployment problem that will be caused by AI, here it’s called MAiD or medically assisted suicide. Our brilliant leaders are expanding the criteria for inclusion and will no doubt be happy to consider the chronically useless unemployed. You might think this to be exaggeration but guess who is the carbon they want to eliminate.
ReplyDeleteI'm pausing the vid to ask after the possibility that the url for it be included (& that this be generalized).
ReplyDeleteto the buggy whips & neanderthal perspectives, the i will survive-adapt-thrive & it won't happen to me graveyard whistlers, if it happens to most people it will happen to you -- just as it has, is, & will continue to happen to you -- no matter the words used to "differentiate" your "collateral" damages.
this failure to apprehend the necessity of defending neighbors & strangers inalienable rights as your own is the corrosive shortsightedness that continuously hydraulically strip-mines away the happiness & nice things mountains. people in groups lose sight of individuals -- that comprise the groups -- & melt into stupid suicide pacts. the blasted-eroded longitudinal fissure between the hemispheres is the grand canyon lemmings leap.
Pale Rider's name is Death:
https://www.youtube.com/watch?v=wcZP0qYNDQc
Fear porn. Seen this before “fuzzy logic” “self learning” etc. basically a database lookup with zero creativity
ReplyDelete