|
|
Post by Baron von Lotsov on Jun 11, 2024 17:59:37 GMT
I may well do that, but I'm not married to him. It's rather a personal question.
What I'm not getting from you Brits though is any realisation that this 93% medical diagnoses accuracy is a figure that indicates that as things progress we will no doubt hit figures more like 99% or even 99.9%. It's already in the same league as medical professionals, so soup up the AI and use the more modern custom AI chips and you are onto a winner. That is unless you are in the team of losers carping at the sidelines with nothing to show of your own medial tech and a 7m waiting list.
I have a Brit tech thread, so when the first British AI hospital comes online I'll post that too for fairness.
Well i’m not convinced about this AI bot business I don’t understand the 10,000 cases in a few days. Remember, i used to be a medical biochemist. That article claims huge gains but i guess that’s really down to many hundreds of booths at which patients are ‘seen’ by a robot. I wouldn’t trust a machine to correctly take my blood pressure and i wouldn't trust a nervous or unskilled patient to put tbe cuff on properly. I sure as hell wouldn’t let a machine draw an analytical quantity of blood from me If i had more information i may be less sceptical, I've seen some modern China robot AI motor control and believe it can be done. I also think it is a very difficult level of sophistication to reach. As a scientist, surely you would wish to see a demonstration first.
By the way, the robot I saw was trained in Minecraft. I can't find the link anymore due to the word Minecraft in the search, but the idea was to train the machine in a virtual world so you could speed up the training much faster than in the physical world. You'd probably think that wouldn't work, but the demo was amazing. It had learnt all sorts of tricks with its hands when handling objects.
Anyway, the university it is coming from in this example is like China's version of Cambridge. I've got another of their inventions in this thread in fact called SSMB or Steady-state microbunching. As for this AI hospital, I only have this one press release which is a copy of the one published in the China press. I expect we will see more in the coming months. It's a bold endeavour.
|
|
|
|
Post by Vinny on Jun 11, 2024 19:41:24 GMT
No, it's a healthy wariness of the programmers.
See, with real doctors there's oaths, morals, duties of care.
And unions.
But with a machine there's simply programming to follow.
And if maliciously programmed to kill a dissident, it's very easy.
Let's say the machine was programmed to perform open heart surgery. Very easy for it to kill and not arouse suspicion. All it has to do is slightly botch the surgery.
And as a robot, it can very precisely botch the surgery and then perform a full factory reset on itself deleting the kill instruction in the process.
|
|
|
|
Post by Baron von Lotsov on Jun 11, 2024 21:03:01 GMT
No, it's a healthy wariness of the programmers. See, with real doctors there's oaths, morals, duties of care. And unions. But with a machine there's simply programming to follow. And if maliciously programmed to kill a dissident, it's very easy. Let's say the machine was programmed to perform open heart surgery. Very easy for it to kill and not arouse suspicion. All it has to do is slightly botch the surgery. And as a robot, it can very precisely botch the surgery and then perform a full factory reset on itself deleting the kill instruction in the process. This is why you need the world's top experts to develop it. In China they also have safety regs, so it would be extensively tested. You have to think, what if it gave the patient greater chance of survival than if a human did the job. Also to make sure it is not going to kill you there are safeguards programmed into these things. They do remote surgery using a robot controlled by a surgeon and they use VR to say give graphical outlines of the organ or whatever, and the system will not allow the knife to go where it shouldn't. This is achieved using 3D tomography scans. Similarly you can scan a tumour so you know every bit that needs removing. Anyhow the AI hospitals are so far only capable of diagnoses, i.e. doctors walking from bed to bed taking measurements and talking to patients. It will give doctors more time to do the things that need their expertise to the full. I think China will have something useful at the end of this research project. These people work very fast. |
|
|
|
Post by Vinny on Jun 11, 2024 22:51:22 GMT
The idea of trusting the Communist Party of China with people's health is like trusting Gary Glitter with your kids.
It's absurd.
|
|
|
|
Post by Dan Dare on Jun 12, 2024 8:41:08 GMT
The Baron's unquestioning faith in the honesty and integrity of the CCP is really quite touching. I haven't seen anything like it since the 1970s when we were waving our Little Red Books in protest at the Vietnam War.
The truly amazing thing about it is the fact that he has never set foot in the country and as far as I'm aware he doesn't speak or read a word of the language. The Power of the Internet writ large.
|
|
|
|
Post by Baron von Lotsov on Jun 12, 2024 9:27:02 GMT
I may well do that, but I'm not married to him. It's rather a personal question.
What I'm not getting from you Brits though is any realisation that this 93% medical diagnoses accuracy is a figure that indicates that as things progress we will no doubt hit figures more like 99% or even 99.9%. It's already in the same league as medical professionals, so soup up the AI and use the more modern custom AI chips and you are onto a winner. That is unless you are in the team of losers carping at the sidelines with nothing to show of your own medial tech and a 7m waiting list.
I have a Brit tech thread, so when the first British AI hospital comes online I'll post that too for fairness.
Well i’m not convinced about this AI bot business I don’t understand the 10,000 cases in a few days. Remember, i used to be a medical biochemist. That article claims huge gains but i guess that’s really down to many hundreds of booths at which patients are ‘seen’ by a robot. I wouldn’t trust a machine to correctly take my blood pressure and i wouldn't trust a nervous or unskilled patient to put tbe cuff on properly. I sure as hell wouldn’t let a machine draw an analytical quantity of blood from me If i had more information i may be less sceptical, Fair comment then. My usual experience with Chinese is when a figure is quoted there is always a reason for it. It's not like this country where some spastic thinks of a number from his head and then lies. I would expect that to be a figure given by the researchers to Global Times and then copied into the article in Interesting Engineering. The thing is Global Times is a publication of the Chinese state in the same way as CGNT is a broadcast of the Chinese state and in the cases where it is a state broadcast, I heard from the BBC once that the editorial goes though some 6 layers of checks. They are ultra scrupulous with the facts and information they broadcast in terms of accuracy. The state in China must never be seen to make a mistake. So given that, I think you can trust that would have come from the researchers and they would likely have published it in some peer-reviewed science paper or whatever. This would mean there is some scientific basis for it, as per what was calculated from test results. In my experience also, when it is a sensitive technology the reports are intentionally vague so as to stop the Americans copying. However if they do get the tech to work then we will know about it and we will hopefully see the article was accurate all along. I don't think this is vapourware. After covid you can imagine the state thinking, damn we had better pull our boots up and have something in place for the next one. This could well be it. |
|
|
|
Post by Baron von Lotsov on Jun 12, 2024 9:41:29 GMT
The Baron's unquestioning faith in the honesty and integrity of the CCP is really quite touching. I haven't seen anything like it since the 1970s when we were waving our Little Red Books in protest at the Vietnam War. The truly amazing thing about it is the fact that he has never set foot in the country and as far as I'm aware he doesn't speak or read a word of the language. The Power of the Internet writ large. Yes well I've always had a skill of being able to figure out the most information from the least, or from incomplete data. It is the opposite to how most Brits are today where they have to repeat it to them many times before they learn one thing, as in beyond spoon feeding. I learnt this skills, not via China, but via dishonest Brits, especially those wankers who sit in our government and places of authority. I had a knack of undermining and getting one over on the bastards and this was possible due to the precise nature of analysing information available to me. One small clue can and often does make a big difference to one's fortunes. Today most Brits are intellectually lazy and as with a person I met the other day, he asks ChatGPT everything. Now that's dangerous! |
|
|
|
Post by Vinny on Jun 12, 2024 9:42:32 GMT
Vinny - we can't have every thread about China, side tracked into a debate on the current Chinese government and/or its human rights record - I've deleted the image, as it's completely off-topic and inappropriate for a discussion about Chinese technology.
If you want to discuss or talk about the Chinese government and its human rights etc, you are free to do so in another thread & you can use that image there.
Thanks.
|
|
|
|
Post by Tinculin on Jun 12, 2024 13:51:34 GMT
|
|
|
|
Post by Vinny on Jun 12, 2024 15:42:40 GMT
Tinculin, what I am talking about is whether a regime that does such things (and spies on people too) can be trusted to provide high tech items from mobile phones to medical robots.
After all, a medical robot has no ethics it simply does as its programming instructs, if it is a surgeon robot it has the means to assassinate someone without even knowing it is doing it.
With sourcecode compiled and proprietary so that nobody outside the company can look at it, who is to know what it is programmed for, except those who programmed it, and even if someone were to somehow decompile the code back to human readable, it is possible for the thing to have two sets of programs on it.
How can we trust any technology from a regime that kills people?
Ethics and human rights are quite relevant to this thread because the tech has the potential to be unethical and has the potential to be abused.
|
|
|
|
Post by Baron von Lotsov on Jun 13, 2024 11:32:28 GMT
Here is an announcement on the construction of the world's first robot university. No it is not a university that studies robots, it is a university attended by robot!
It's the quiet revolution our morons are ignoring. This is what we should be doing, and yet we lag behind. As you can see, this move fits nicely with the AI hospital plan. Anyway, as with many things, this university will be in Shenzhen. I do a lot of shopping in that city. It has the best tech and it is all at great prices.
|
|
|
|
Post by Vinny on Jun 13, 2024 12:01:38 GMT
A robot only does what it is programmed to do.
It has no thoughts of its own, no conscience.
Therefore, it has the potential to be the perfect assassin, if so programmed.
I can see such menaces becoming popular killers for the regime in North Korea.
In you go needing a hip replacement or something, out you come, in a box, dead.
Quiet concealed assassinations.
|
|
|
|
Post by Baron von Lotsov on Jun 13, 2024 14:02:28 GMT
A robot only does what it is programmed to do. It has no thoughts of its own, no conscience. Therefore, it has the potential to be the perfect assassin, if so programmed. I can see such menaces becoming popular killers for the regime in North Korea. In you go needing a hip replacement or something, out you come, in a box, dead. Quiet concealed assassinations. That sort of thing does go on in North Korea, but North Korea is a very backward nation ruled by propaganda.
What you don't see in this bigger picture with China is China does not need to assassinate anyone to gain more power. In fact it won't gain more power by doing such evil stuff, rather they would lose power as we have done.
The plan is very simple for me to understand and it goes as follows. They develop these robots themselves, and the develop them to replace jobs. China has a serious problem which you and others have mentioned in your copy and paste hit pieces from time to time that tells the world China has an ageing population. This is factually true. The one child policy is now admitted by the government that it was a mistake and they are now trying to promote more children in the government system of taxation and subsidy. However, even if they do they will still move into a time when they have less workers to old people ratio and this will affect their economy. I had always expected them t fix this problem with robots and I'm now convinced the technology is available.
However, and this is a big however. As my link explains they already have the industrial capacity to knock out millions of robots. As my thread explains, they have a lot of emerging technology in the hardware to give the computational performance to these robots. We're talking new types of AI chip here, most likely they will be photonic in the future.
Now can you see what this means? The technology may well be very expensive to develop, but it is like software. It costs as much to write software for one computer as it does ten million. We say the software has a zero marginal cost. So you can see if they could produce millions of highly intelligent robots, these would replace many of the world's jobs. They will export them to the world, and those robots who save lives will be highly appreciated. This kind of puts China as king of the castle and yet no one needed to be assassinated and all the world are very grateful to the ones who create them. We did the same in Manchester with the industrial revolution. You really don't need to kill people. You can make everyone like you if you are a big benefit to them. It gives you political support. Anyway, I wouldn't mind betting these guys will pull this off. There are some extremely clever Chinese academics, way above us lot.
|
|
|
|
Post by Vinny on Jun 13, 2024 14:28:10 GMT
We are talking ethics, not industrial capacity.
The ability to churn out thousands of robots programmed to sometimes kill, is not a good thing.
|
|
|
|
Post by Baron von Lotsov on Jun 13, 2024 14:45:47 GMT
We are talking ethics, not industrial capacity. The ability to churn out thousands of robots programmed to sometimes kill, is not a good thing. Yes of course, so why do it. Surely it is better to use the power of the robots to save life. Look around the world at how many billions of people there are and how many die through inadequate medical treatment due to the cost and the time it takes to train medics. If you end up building a hospital for robots to work in that hospital would be like medics working a 24hour day and they could be made to work at many times the speed, so the capital cost of a hospital shrinks dramatically. AI is all about training and using chips that can handle the throughput. The more robots you have the more real data you collect which you feed back into the training model. So you can see that if you are going to do that, you really need the huge scale for it to work, but China has the capacity to do that and neither the UK nor America has the industrial capacity. Chances are this was planned a long time ago. They will take over the medical market with it. We had better watch out. In fact better still would be if we joined in this development and lent our expertise so we can get there quicker and still be in business. |
|
