Clif High on Real AI
(www.bitchute.com)
Comments (8)
sorted by:
Presentations like this are the reason I don't like Clif. He presents his opinion as if it were fact. He misses the real issue with AI.
That people will prefer to interact with AI instead of real people, because AI will cater to whatever crazy ideas you happen to have. Human ingenuity advances through dissenssus and interaction with those who disagree with you. The introduction of AI promises to not just limit social interaction, but completely supplant it. And people will gravitate to it for the simple reason that it strokes their ego.
That these jobs being taken over are going to creep up the social hierarchy. It is true that ChatGPT-4 is not very creative. But given a complex enough model, it can simulate creativity quite well. It can just explore parameter space and try 1000 different random things in the time a human might be trying to build a mental model of what to create. And that can look very much like creativity.
That as AI improves, we are all going to be displaced. All of us. And since our entire social structure is built around the myth that I deserve what I have and you don't because you don't generate as much value as I do, this is going to create a crisis. In the world of advanced artificial general intelligence, none of us will be able to out compete and AI. None of us will deserve anything. That doesn't mean humans have nothing to offer, but the vast majority of people are going to have their sense of self worth entirely destroyed. It will require a complete re-engineering of society, with radically different models of how we value ourselves and other humans. Essentially, a wholesale return to spirituality.
I could write a 20 page essay on all the genuine problems with AI. (I do agree it definitely isn't what he called item #1). We have the ability to stop it of course, but right now, every business is running frantically as fast as they can to get AI into every product they can, without even considering the larger ramifications. This has to stop, and we have to completely remove the profit motive from AI to get everyone to take a breather.
Clif does mention that we can lose jobs - he gives an example of law clerks, insurance, finance, government, medical jobs where you handle text. Anything that requires handling paper (and words). It also includes jobs that require design like boat making, It reminds me of a kind of more computerized CAD. He says that AI will be more like an assistant for many jobs. It will perform a lot of the nuisance paperwork. We will still need IT jobs, especially for coding. But he says it does not replace creativity. I remember reading about GE trying to use a kind of AI years ago to build and maintain the many kinds of mechanical systems they sold. They found out that it all worked out well until there was a trouble shooting event. Humans were needed for solutions. They concluded that you could not exclude humans from their workforce.
I agree with you that there is a strong demand by industry for AI implementation. I read a lot of monthly magazines on the energy sector and manufacturing and they are all chirping it up. It's the next "disruptive" technology. We will probably hire a lot of H1B Visa people to do that work. That's where I see USA job loss. Probably will need a large help desk!
I attended a Zoom meeting last at a university where I work and they are pushing AI for classwork. Some profs are going to form focus groups to try it out. Yet many profs don't like it because there is no real way to evaluate plagiarism. Duh I asked my students if they use AI and most don't. Those that do, say they use it for things like cover letters, repetitive and boring writing like that. But they told me that they all like the hands-on work we do (I teach a science class). It will not be simple to get the younger generation on board.
The state I live in a blue, very metropolitan state that still uses fax instead of email for things like paperwork for pensions)!
I think Clif has given us a good presentation on many of the aspects of AI that concern ordinary people. He built his own linguistic computer system and the man is into words. He started this podcast by talking about how he asked people like store clerks and tree loggers how they felt about AI.
I listened to his entire speech. I simply disagree with his assessment. He is not the only person who understands LLMs and how they work. Yes, you can make an argument that everything he says is true right now. But that says nothing about whether they will remain true after a couple more years. Capabilities will creep slowly up the hierarchy, and he is ignoring the larger threat because of his personal biases. That is fine. We all have them. But he never admits these are his own personal opinions. He presents them as if they are established facts, which is far from the case.
Clif does present his information in a professorial way. Hey - he doesn't believe that gravity exists and of course, he had to include space aliens, But then again. he is uncle Clif :-) and I enjoy many of the topics he brings up.
The flying paintbox, lots of experiments with cavitation. has not been debunked.
I haven't followed that.
Thank you. I'll watch this with interest.
Clif is the one to explain this - it's really good. This is his kind of stuff.