Let's Riff on the next Tech/Social Revolution

[DrYouth]

Not sure where you're going with this either...

I believe technology has gotten to the point where the ruling class no longer has to fear the proles....in other words, there won't be another French Revolution - no matter how much life sucks for the proles. So when prole jobs get replaced via automation, the rich won't have to give a fuck as to how to handle them....

Except, except... the proles still have to keep the economy running (by buying stuff)... unless you are selling to foreign markets I suppose.

[DBTrek]

Except, except... the proles still have to keep the economy running (by buying stuff)... unless you are selling to foreign markets I suppose.

Stalin and his five year plans say "no".

[Smitty-48]

What did the French Revolution achieve for the proles? France went into a hundred year fiscal and economic decline while the British made out like bandits, dumbest thing the French ever did, blew themselves up for the British, nary a shot fired. In the end, the British had to bail the French Republic out, and the British were so rich by this point, only they could do it.

The French Revolution, was the final war of French Hegemonic Succession, which made the British the undisputed Hegemon, for the next hundred and twenty years.

If the American proles were to overthrow King Dollah, where does that get them, exactly?

[heydaralon]

Sapiens was great, but his follow up book was kind of lousy imo. It is silly to reflexively hate all new tech, but it is equally silly to don rose colored glasses like Harari does when discussing genetic engineering and the like.

I like Sapiens better... but Homo Deus was thought provoking.
I like his take on Humanism as an ideology and the implications the coming tech will likely have on it.
We kind of take Humanism for granted as we are so immersed in it.

Humanism as in believing we are unique and that morphing into some kind of robotic cyborg is inherently inhumane?

Not inhumane, but secular humanism is a shoddy replica of christianity. They borrow so much from it, yet most of them don't believe in God. This creates a strange philosophical problem. Harari is not ignorant of this, and he tackles it often in his sapiens book. Homo Deus however, takes the assetions of humanism, transhumanism, at face value. He takes it as a given that tech will create a morally superior society, and that the problems we are debating in this thread will be solved. This is a bit hubristic imo. I would encourage everyone to read Sapiens, as it is a flawed but interesting "Grand History" type book. The sequel left a lot to be desired, and felt rushed.

[DrYouth]

He takes it as a given that tech will create a morally superior society, and that the problems we are debating in this thread will be solved. This is a bit hubristic imo.

Really?

I didn't read it that way at all.

I read him as very cautionary about the potential downsides of the new technology and the impetus of humanism.

I believe he Harari sees the impetus of humanism, in the coming age as towards efforts to engineer immortality, happiness and superhuman abilities.

We already see these efforts at present but they will only increase with the advancement of AI and genetic bioscience.

That itself is very frightening at face value... because these are in all likelihood hubristic aspirations with plenty of potential for blowback. He is very aware of the hubris factor IMO.

[DrYouth]

I'm curious about the state of AI.

As you might now the AI that won jeopardy - Watson, is now available online, and apparently can be trained in all sorts of ways.
I heard it has been let loose on medical and legal databases with considerable success.

Here's Watson's website...
An IBM product, apparently.

https://www.ibm.com/watson/

Any of you have any experience with this sort of thing?

[Speaker to Animals]

We are nowhere close to general AI. Weak AI has made tremendous progress, but most people would be disappointed if they learned how it works. It's sort of like magic tricks, really.

[SuburbanFarmer]

I'm curious about the state of AI.

As you might now the AI that won jeopardy - Watson, is now available online, and apparently can be trained in all sorts of ways.
I heard it has been let loose on medical and legal databases with considerable success.

Here's Watson's website...
An IBM product, apparently.

https://www.ibm.com/watson/

Any of you have any experience with this sort of thing?

It's not really "AI" in the sense of a self-aware machine, or anything. It's a "learning" algorithm that searches the internet for context to a question, and puts together the most likely answer. It's extremely advanced tech, and could eventually evolve into something more, but right now... Uber-Librarian.

[Speaker to Animals]

Most of us work on specific problem domains. Right now, I am focused mainly on artificial neural networks, which are good at pattern recognition. They are very applicable to image processing and computer vision. I have also been very interested in expert systems (like the CLIPS system used by NASA rovers), and plan to integrate the two in a hybrid framework.

That still doesn't make for self-aware, self-directed general intelligence. There are a few things we really don't understand.

Personally, I think some of the largest impediments have to do with embodiment and self-reflection. The embodiment problem can be solved, though I don't have any idea how to make a self-aware AI artifact. Embodiment just means that your understanding of the world around always goes back in a chain of inferences to your body. When you are born, it's all you have. You don't even know individual things from one another. You learn about the world first by experiencing you have a body, with senses and organs that allow you to manipulate the world or cause reactions in other people, and you learn through inference.

An argument can be made that we cannot truly understand anything that does not somehow relate through some chain of inferences to our bodies. I am not sure that is correct, but you can see how this works even in our scientific investigations. The early models of atoms, for example, were based on our observations of our solar system in a vague fashion, even though we now know it doesn't look anything like that silly model we were taught as recently as the 1970s and 1980s. We look for similarities to some ontological construct we know and then try to extrapolate inferences. It might be a shitty model at first, but we can refine it later.

So, that said, I personally feel that when we do have true general AI, it won't be some program running on a computer like HAL 9000. An AI will be the spaceship (the computer is the brain, the ship is the body). It will need to have a body, and it's understanding of the world will derive from that embodiment. If you want to construct an AI that is at all relatable to humanity, you will need to build an artificial human body for it.

[TheReal_ND]

I'm curious about the state of AI.

As you might now the AI that won jeopardy - Watson, is now available online, and apparently can be trained in all sorts of ways.
I heard it has been let loose on medical and legal databases with considerable success.

Here's Watson's website...
An IBM product, apparently.

https://www.ibm.com/watson/

Any of you have any experience with this sort of thing?

we need new forum members...