This very long podcast is a must listen for everyone. On the one hand AI is increasing in intelligence exponentially and is now as smart as Einstein. In a few years could be a million times smarter.
Combine the two and it seems that cures for everything will be quickly found. If it doesn’t kill us all first in an accident while it is a teenager AI.
Written by
LAJ12345
To view profiles and participate in discussions please or .
At the moment but what podcaster says is that it is getting to the point it can teach itself then it will go exponentially.
At the moment it can only use data given to it. Once it’s let loose on the internet it will have access to everything that humans have, will be able to digest it all very quickly and learn by experimenting like a human baby does except exponentially faster.
Of course at that point humans might lose control of it and it probably won’t be interested in finding out cures for humans. Who knows what it would prioritise. Maybe using biology to make its own human like body. Anyway the more you think about it the scarier it is.
AI is nothing like a human or even a human baby. It has no experiential referents for the text strings that it manipulates. "Agony" and "ecstasy" are just a couple of text strings to AI. It is artificial intelligence and it is a mistake to anthropomorphize it.
humans are biological machines in essence and learn as babies do by trial and error.
AI will teach it self the same way. It might not have emotions and ethics such as fear etc but will be trying to complete its mission.
Eg if there is a flood forecast it might copy its data to a safer storage place so it can carry on with its assigned task. If it needs to access data protected from robots by a captcha to complete a task it might mimic human behaviour by saying it is visually impaired and needs help ie by appealing to human empathy.
It’s not anthropomorphising it to imagine it will find code that works to help it complete its mission. This is not different to how humans with EQ and also even sociopaths use behaviours which are made up of strings of words to get what they want.
Given they are predicting its increase in ‘IQ’ to be exponential once it starts programming itself I think we would be silly to assume it won’t be able to get around anything humans use to try and stop it at that point.
It obviously won’t have the same morals or ethics inbuilt unless humans program them in but currently it is just getting a lot of data and any human emotional devices will be picked up from social media and movies etc I assume which aren’t desirable human interactions generally.
Say it’s given the task of solving climate change of course the obvious thing is to get rid of humans. But if they can’t do that maybe they can shut down any energy platforms that produce emissions which would be chaotic and kill a whole bunch of people.
It can’t now be got rid of as even if all countries say they will stop the development or prevent its misuse we have many rogue humans that will carry on development for their own purpose so governments can’t afford to not be more advanced than them.
No, but it is the top programmers of AI in the world that are worrying about just these issues so I think if they are worried we should be.
And AI will have its own language to interact with other AIso we won’t even know what they are saying to each other. When they gave a transaction to be carried out between 2 AI they quickly began organising it in their own language as human language is so imprecise and clumsy.
It's an interesting discussion but still science fiction. Computers have no feelings, sense of being, emotion, spirit...So they will never be motivated to do anything beyond what a human tasks it to do.They are just huge data receptacles that can analyse data quickly.
Did you listen to the bit where it says improvements are going exponential. At the moment it has improved from preschooler to teenager. Now it is teaching itself and not relying on humans for development it will quickly go from an IQ of Einstein to a million times that And the fact it makes mistakes early on is one of the main points of concern. Let's hope while it's a teenager it doesn't destroy us all.
And how does ai learn that it’s wrong? It is not infallible, so what prevents it from going down an entirely wrong path based on one wrong calculation or sssumption it makes? And then humans take the info as fact?
This discussion - with its back and forth on the use, meaning and significance of AI - reminds me of Bob Dylan’s ‘Ballad Of A Thin Man’…
Ah, you've been with the professors and they've all liked your looks
With great lawyers you have discussed lepers and crooks
You've been through all of F. Scott Fitzgerald's books
You're very well-read, it's well-known
But something is happening here and you don't know what it is
Do you, Mr. Jones?
Like Mr. Jones, most of us do not have a clue as to the monumental changes AI can invoke. Once simple example would suffice. The image is the work of someone who can not draw to save his life 🧐
Artwork by CaseyInsights 🤯
AI generated image, inspired by Bob Dylan ‘Ballad Of A Thin Man’, by CaseyInsights
Well yes. There are human states that are truly destructive to others and to themselves that are best avoided. Sometimes they are side effects of drugs or medications provided in the leaflet. 😁
Imo it's best to avoid certain videos, better to go out and have a good walk.
AI is really good at diagnostics. Give AI 10,000 pictures of old chest x-rays and the eventual conditions the patients developed and AI can find the patterns and learn to predict disease.
Where AI lacks, I think, is that it only compiles generally accepted medical beliefs, so all you get from AI is the establishment consensus. I think.
But... I think they could program AI to randomly select treatment protocols and then research those options. This would get it out of the box. Now that I think about it, I would be surprised if they have not programmed random trial and error into AI.
That might be now but it will soon go through all the possible genetic mutations and be able to pinpoint what causes what and suggest chemicals that might treat diseases and do simulations. They are already making human cells. Will they make whole bodies of these cells to do trials on. When are they are human? It’s scary to think eventually will have realistic looking humans with AI as brains? Hopefully it will be nirvana as they solve every problem but I think there will be a lot of chaos before that time is reached.
AI is just a calculator that averages or correlates on a large scale, it doesn't solve problems, because it has no motivation and therefore is not able to pose "the problem for its resolution" as intelligent Life does. The only purpose that AI has is to respond in a certain way to a request for information. It can be made to appear as an identity but it has no individuality which is a different and much higher thing typical of the human being.
One could say that all of his intelligence depends on the question you ask him.
Moreover, even in research and discovery it is not very useful because it is not able to use the "inductive" research method but only the "deductive" one based on the analysis of the correlations of a presumed large body of data.
In other words it depends a lot on what you put in it hoping that like calculators it doesn't have broken keys that distort the result.😁
AI is very very powerful. But it has no actual human properties.As always the dangers of AI are driven by human applications of tech.
Mass unemployment is the danger.
Those who own the most powerful AI, e.g. Musk ChatGPT, are publicly requesting it's development restricted. Some say that's his way of stopping others developing what he has.
There's no doubt it has medical benefits but it's based on human programming.
uh! You mean he still makes money with the old system of creating scarcity on a good so that the price goes up?
He certainly needs AI more than anyone else. Someone explain to him that there are other more profitable ways such as producing and delivering quality products in large quantities, such as Apple or Toyota, or Italian Parmesan.👍😁🇮🇹
Here in the EU he is not considered much like in the US, we like gaming more than possessing. So we had saints, great explorers and inventors who all died in relative misery.
Content on HealthUnlocked does not replace the relationship between you and doctors or other healthcare professionals nor the advice you receive from them.
Never delay seeking advice or dialling emergency services because of something that you have read on HealthUnlocked.
Why don't we have a cure yet? What are we paying the drug companies for anyway?
One of my best friends got Parkinson's ... this is my journey to find a cure for him...it is going well !
MRI-Focused Ultrasound Undergoing Phase 3 Clinical Trial for Parkinson’s Treatment
Matt Eagles: Positivity in Adversity – 46 years living with Parkinson’s
