foreveryoung Posted February 17, 2023 Share Posted February 17, 2023 It's been said AI is going to be a bigger threat in the future than nuclear war. Link to comment Share on other sites More sharing options...
Amsterdam_Neil_D Posted February 17, 2023 Share Posted February 17, 2023 2 hours ago, CVByrne said: Just think of how useful AI would be in medicine? The trying to get a GP appointment. It'll improve the productivity and efficiency There is no reason why AI can't call for an appointment for you and drop it into your calendar when arranged. I suspect it would get through the choose 1 for X or choose 2 for Y as they are pretty easy questions. No more being on hold. Link to comment Share on other sites More sharing options...
PussEKatt Posted February 17, 2023 Share Posted February 17, 2023 Dont forget AI beat the current Chess champion back in the 1980s I think. On a more sombre note,I remember that on the ZX Spectrum there was some sort of bug in the circutry/programming that would give you strange results every now and then. 1 Link to comment Share on other sites More sharing options...
Wainy316 Posted February 17, 2023 Share Posted February 17, 2023 47 minutes ago, foreveryoung said: It's been said AI is going to be a bigger threat in the future than nuclear war. Or the AI cause the nuclear war? Come on, you've seen Terminator. Link to comment Share on other sites More sharing options...
Straggler Posted February 17, 2023 Share Posted February 17, 2023 I have a friend that I met through my work in IT that is probably the most intelligent person I have ever met. We were working at an IT security company and the joke was that we didn't really have any software, we just plugged everything into his brain. He also was involved in research into AI, flying between the UK and the USA working at the sorts of universities that you will have heard of. He told me the story of his scariest day at work. They were working on AI the sort of which you see in the movies, trying to get something that could learn broadly on its own. One day they are at the lab experimenting away when the AI mounts and escape attempt. It hacked its way out of the lab network into the building control system. It was a high tech building with all sorts of security procedures and networked features. Unfortunately for the team in the lab as they were trying to contain the AI one of the feature were remote locking of doors which a security team used to allow folks in and out of secure areas. The AI having got loose locked down the building, securing all the flesh bags into their labs. My friend tells me that it was a really freaky moment knowing that the AI had done that (even though it is not a possibility that it was a deliberate action against humans) Fortunately the lab was completely air gapped from the internet so the AI was contained in the building which they eventually regained control of by completely cutting power. My friend looked suitably concerned when I asked him what might have happened had it got out and copied itself across the www. 1 2 Link to comment Share on other sites More sharing options...
hogso Posted February 17, 2023 Share Posted February 17, 2023 6 hours ago, sne said: On that note, maybe this is the way for those who have lost faith in G.R.R Martin ever finishing GOT. And as I write that they bring it up in the LTT video above Someone got ChatGPT to write the ending for Berserk - and it's pretty good. So yeah, you should probably do that and post your findings. 1 Link to comment Share on other sites More sharing options...
Designer1 Posted February 17, 2023 VT Supporter Share Posted February 17, 2023 6 hours ago, sne said: On that note, maybe this is the way for those who have lost faith in G.R.R Martin ever finishing GOT. Isn't that what Benioff and Weiss did from s5 onwards anyway 2 1 Link to comment Share on other sites More sharing options...
OutByEaster? Posted February 17, 2023 Moderator Share Posted February 17, 2023 6 hours ago, CVByrne said: Just think of how useful AI would be in medicine? The trying to get a GP appointment. It'll improve the productivity and efficiency There are areas and studies where AI has been proven to be more efficient than the average Doctor at diagnosing diseases - I'm don't know how widely AI diagnosis is being used in hospitals and health services already, but the theory and the testing is out there. Disease diagnosis is based largely on pattern recognition and AI is obviously good at that. Link to comment Share on other sites More sharing options...
OutByEaster? Posted February 17, 2023 Moderator Share Posted February 17, 2023 There's a take here on how difficult it is for people to accept AI in relation to their own health, even when it's proven to outperform Doctors. https://hbr.org/2019/10/ai-can-outperform-doctors-so-why-dont-patients-trust-it Quote Our recent research indicates that patients are reluctant to use health care provided by medical artificial intelligence even when it outperforms human doctors. Why? Because patients believe that their medical needs are unique and cannot be adequately addressed by algorithms. To realize the many advantages and cost savings that medical AI promises, care providers must find ways to overcome these misgivings. Link to comment Share on other sites More sharing options...
tinker Posted February 17, 2023 VT Supporter Share Posted February 17, 2023 AI's real danger is how it learns to influence humans to believe they are forming their own opinions when really they are being influenced by AI. It's quite possible to influence whole electorates to vote certain ways. I have read my own results from a physiological profiling test and was shocked out how accurate it was in reading my personality, my likes and dislikes , weakness and strengths. The questionnaire seemed pretty harmless roughly 50 multiple choice type questions. The result was about 500 words profiling my personality, friends I had for 20 years wouldn't have been that accurate with my personality. As per usual our politicians will follow the developments and quite possibly close the door once the horse has bolted. 1 Link to comment Share on other sites More sharing options...
villakram Posted February 17, 2023 Share Posted February 17, 2023 23 minutes ago, OutByEaster? said: There's a take here on how difficult it is for people to accept AI in relation to their own health, even when it's proven to outperform Doctors. https://hbr.org/2019/10/ai-can-outperform-doctors-so-why-dont-patients-trust-it The issue is mostly legal. An algorithm produced by AI corp promising 90% efficiency at detecting stress fractures in MRI imaging (for example), is going to need some time to accumulate real world data to have a real world 90% number to throw at the 10% and then it better not include anyone in the 10% who a human would have spotted (very very hard). As an aid, these are being deployed now. We are some time from these pattern recognition algo's being general. Link to comment Share on other sites More sharing options...
villakram Posted February 17, 2023 Share Posted February 17, 2023 1 minute ago, tinker said: AI's real danger is how it learns to influence humans to believe they are forming their own opinions when really they are being influenced by AI. It's quite possible to influence whole electorates to vote certain ways. I have read my own results from a physiological profiling test and was shocked out how accurate it was in reading my personality, my likes and dislikes , weakness and strengths. The questionnaire seemed pretty harmless roughly 50 multiple choice type questions. The result was about 500 words profiling my personality, friends I had for 20 years wouldn't have been that accurate with my personality. As per usual our politicians will follow the developments and quite possibly close the door once the horse has bolted. Now think how many of those questions you answer inadvertently online via your navigation patterns, click through time as a function of time of day, day of week, big life events etc., etc. Combine this with publicly for sale info. about addresses, credit score, education etc. Bet most of us are close to answering those 50 questions. Hi digital me! 3 Link to comment Share on other sites More sharing options...
KentVillan Posted February 17, 2023 Share Posted February 17, 2023 3 hours ago, Straggler said: I have a friend that I met through my work in IT that is probably the most intelligent person I have ever met. We were working at an IT security company and the joke was that we didn't really have any software, we just plugged everything into his brain. He also was involved in research into AI, flying between the UK and the USA working at the sorts of universities that you will have heard of. He told me the story of his scariest day at work. They were working on AI the sort of which you see in the movies, trying to get something that could learn broadly on its own. One day they are at the lab experimenting away when the AI mounts and escape attempt. It hacked its way out of the lab network into the building control system. It was a high tech building with all sorts of security procedures and networked features. Unfortunately for the team in the lab as they were trying to contain the AI one of the feature were remote locking of doors which a security team used to allow folks in and out of secure areas. The AI having got loose locked down the building, securing all the flesh bags into their labs. My friend tells me that it was a really freaky moment knowing that the AI had done that (even though it is not a possibility that it was a deliberate action against humans) Fortunately the lab was completely air gapped from the internet so the AI was contained in the building which they eventually regained control of by completely cutting power. My friend looked suitably concerned when I asked him what might have happened had it got out and copied itself across the www. And then everyone clapped? 2 Link to comment Share on other sites More sharing options...
mjmooney Posted February 17, 2023 VT Supporter Share Posted February 17, 2023 In my experience, the biggest failure of medical professionals is their inability to treat patients holistically. They tend not to treat the person, but the symptoms. It's not their fault, it's an inevitable consequence of how they have to learn medicine; either they become GPs, who - lacking in-depth knowledge of the huge range of problems that flesh is heir to - fall back on acting as a referral service for hospital specialists... or they become one of those specialists, with a great deal of expertise and experience in a limited number of body parts and conditions. But if you're unfortunate enough to have multiple problems, who has the skillset to understand how they may be interacting, and treat/prescribe safely and effectively? If AI can assist with those sort of issues, it could be a real boon. 2 Link to comment Share on other sites More sharing options...
KentVillan Posted February 17, 2023 Share Posted February 17, 2023 6 minutes ago, mjmooney said: In my experience, the biggest failure of medical professionals is their inability to treat patients holistically. They tend not to treat the person, but the symptoms. It's not their fault, it's an inevitable consequence of how they have to learn medicine; either they become GPs, who - lacking in-depth knowledge of the huge range of problems that flesh is heir to - fall back on acting as a referral service for hospital specialists... or they become one of those specialists, with a great deal of expertise and experience in a limited number of body parts and conditions. But if you're unfortunate enough to have multiple problems, who has the skillset to understand how they may be interacting, and treat/prescribe safely and effectively? If AI can assist with those sort of issues, it could be a real boon. The way I see it is that the evolution of jobs tends to be from people who are very good at memorising/processing information, to people who are good at mediating. Eg accountants used to be people who were efficient at maths, bookkeeping, etc. Now so much of that stuff is handled by the software, and more of the accountant’s value is about how they deal with the client and advise on specific business problems. Or taxi drivers used to memorise locations, but now freely available sat nav has made that pointless. I wonder if it will be the same with doctors. Currently it’s still a profession where a lot of rote memorisation and repetition is central to the training, and that’s why you have these senior doctors who know their field inside out but have no social skills, because they’ve been selected for “exam brain”. AI tools may help to humanise the doctor’s work, by taking on some of the heavy lifting with diagnostics, obscure diseases, that kind of thing, and freeing up more of the doctor’s brain to deal with the patient on a personal level. 1 Link to comment Share on other sites More sharing options...
sne Posted February 17, 2023 Share Posted February 17, 2023 8 hours ago, CVByrne said: Just need to do the tests in room though. Also if a free AI can do something for you while you're in school, should you need to be tested on it? When you're in workplace years later nobody is going to ask you to do a task AI can do. For regular exams sure. But for hand in assignments, papers and such it's a real issue. 1 Link to comment Share on other sites More sharing options...
CVByrne Posted February 17, 2023 Share Posted February 17, 2023 1 minute ago, sne said: For regular exams sure. But for hand in assignments, papers and such it's a real issue. Yeah, my point is hand in exams and papers are no longer needed at all as part of education. If we have AI to do it. You replace it with a presentation where you answer questions on the topic etc.. Education is the slowest oil tanker in terms of changing. The whole concept of a University bestowing some special piece of paper upon you. It ignores the fact it's mostly useless what you learnt there. Link to comment Share on other sites More sharing options...
sne Posted February 17, 2023 Share Posted February 17, 2023 2 minutes ago, CVByrne said: Yeah, my point is hand in exams and papers are no longer needed at all as part of education. If we have AI to do it. You replace it with a presentation where you answer questions on the topic etc.. Education is the slowest oil tanker in terms of changing. The whole concept of a University bestowing some special piece of paper upon you. It ignores the fact it's mostly useless what you learnt there. Yeah that change is gonna take decades. Meanwhile they'll have to figure out how to adjust Link to comment Share on other sites More sharing options...
CVByrne Posted February 17, 2023 Share Posted February 17, 2023 6 minutes ago, sne said: Yeah that change is gonna take decades. Meanwhile they'll have to figure out how to adjust Ask ChatGPT if it wrote this paper ? 1 Link to comment Share on other sites More sharing options...
leemond2008 Posted February 17, 2023 Share Posted February 17, 2023 (edited) I've been using it a fair bit at work, I deal with injury claims and every now and again I'll come across a condition or injury that I don't know or I'll need to know the average recovery period or treatment for a specific injury, chatgpt is fantastic for that sort of stuff, much much better than a basic google search Also I haven't mentioned it to anyone else from work because it makes me look really good when seconds after being asked a question I'm throwing out detailed answers like I have the slightest of clues what I'm actually talking about Edited February 17, 2023 by leemond2008 Link to comment Share on other sites More sharing options...
Recommended Posts