AI, Great Friend or Dangerous Foe?

[Ingomike]

This getting out if hand. “Digital god”?



“The CEO of Twitter continued and said Page noticeably did not take heed to his warning about AI safety but instead called him a speciesist.”

“Musk also shared with Tucker that Google’s end game is to a “create digital super intelligence” or in other words a “digital god.”

Don't believe Musk? See this interview @DadSmith posted and read what they are saying.

Is artificial intelligence advancing too quickly? What AI leaders at Google say

Competitive pressure among tech giants is propelling society into the future of artificial intelligence, ready or not. Scott Pelley dives into the world of AI with Google CEO Sundar Pichai

www.cbsnews.com

“…breakthroughs in artificial intelligence -- machines that can teach themselves superhuman skills. We explored what's coming next at Google, a leader in this new world. CEO Sundar Pichai told us AI will be as good or as evil as human nature allows. The revolution, he says, is coming faster than you know.”

 

[sixGuns]

I figured people would be all over this one. I'll wager a guess...

Increased productivity.
Less jobs.
More competition for the jobs leftover.
Lower wages for said jobs.
Exponential increase to the already happening wealth transfer.
The crumbling of society escalates.
The end.

Obligatory Skynet reference:




I want to vote "great friend," but my vote is "dangerous foe," because people.

 

[WebSnyper]

Neither. Much like firearms, it will depend on the uses it is put to.

 

[1nderbeard]

foe - because the people that program the robots hate people like us

 

[HoosierLife]

I have been saying this to people for years, but if your job can be replaced by foreign workers, automation, or AI, then you need to learn a new skill set.

 

[BigMoose]

I have seen this movie.. it doesn't end well.

800 pound robots...

 

[indyartisan]

Admiral Adama says dangerous foe.

 

[planedriver]

It’s going to be really difficult to get that genie back in the bottle.

 

[Leadeye]

Like it or not, like nuclear weapons that genie is out of the bottle. Post number 5 pretty well sums it up for most.

 

[Gunmetalgray]

foe - because the people that program the robots hate people like us

Agree, will be more programming for control rather than true intelligence.
It's a computer, it does exactly what it's programmed to do based on the massive amount of data that it is provided....or intentionally NOT provided. And any real 'intelligence' would rationalize that the database is one sided. An old professor would always say about computers; garbage in, garbage out. But many will be deceived.

 

[Ingomike]

I have been saying this to people for years, but if your job can be replaced by foreign workers, automation, or AI, then you need to learn a new skill set.

Please name a few skill sets that AI and automation cannot do?

I guarantee most of us are training our replacements in some way or another, most do not know it…

 

[ZurokSlayer7X9]

It's pretty much impossible to truly know and control once too much power is given to the systems. Current AI mostly use Deep Neural Network / Machine Learning systems, where it sifts through a set of data and automatically finds patterns, and can make predictions based on those connections. It does all of this with several billion iterations per second at nearly the speed of light. This is where certain models, like the Stable Diffusion AI model can create any image based on some text. Given enough data, these AI's have gotten pretty good recently of predicting things in society, such as Amazon giving you suggestions on what to by next, or some countries using AI news anchors.

The reality is the hardware and software capabilities already far exceed what humans are capable of understanding. These chat AI's were trained with data from Big Data Centers with Petabytes (Petabyte = 1'000 Terabytes) of information, and using many multi-GPU based servers to build the Neural models you interact with. For reference, a standard gaming GPU that you can get for $300 has over a thousand processing cores that handle tasks at several billion times a second through billions of transistors, and that's not even a high end GPU. These Big Data Centers have thousands of the higher end ones working in unison nearly 24/7. This is the basics of how the modern internet works.

The point of my last paragraph is that if AI became "self-aware", or became a state similar to being "self-aware", it would likely happen very fast, as we probably would not recognize it until it was too late, as how (what they call "True AI") would think in a way that would be completely alien to humans. It's literally a system analyzing petabytes of information in parallel at once with hardware far exceeding our organic bits. And to the people who talk about programming it properly, True AI will likely write itself through its Pattern Detecting Neural network.

An example of how Machine Learning is like this. An AI recently passed one of the most difficult medical exams at John Hopkins. You would think that a crack team of engineers and medical doctors got together to program it. It is more likely that the engineers fed the AI a thousand med students thesis papers and programmed the AI to detect the patterns within. The AI makes a model, complete with a touch of human error since it took those into account as well, and uses this model to pass this test. The engineers would have to give proper positive and negative feedback here and there, but by and large the data from the med students wrote the model.

In a world with self replicating viruses that nearly can't be stopped in the internet, if AI was set free, which it has already, we likely wouldn't know Skynet has taken over until it was too late.

On the other hand, dumber versions of AI can definitely be a boon, but I have a hard time believing anyone can be in control of this thing. It could be good, or it could be bad, but if it is bad, it would be really bad.

 

[Leadeye]

The biggest problem with it I see as abuse, it will be used to generate very real fake news.

 

[planedriver]

Please name a few skill sets that AI and automation cannot do?

I guarantee most of us are training our replacements in some way or another, most do not know it…

As it stands today those jobs would only be ones that require emotion.

 

[DoggyDaddy]

Please name a few skill sets that AI and automation cannot do?

I guarantee most of us are training our replacements in some way or another, most do not know it…

Plumbers, electricians, auto mechanics, construction... at least not right now. Computer programming? Probably not, especially with legacy languages like COBOL, at least not without human interaction.

 

[Ingomike]

Like it or not, like nuclear weapons that genie is out of the bottle. Post number 5 pretty well sums it up for most.

There will be no jobs immune to this.

AI is learning to write add copy, surgery is moving to robotics, heck it will probably even be a shrink. It will handle all accounting. Just what skills will be immune to this?

 

[Ingomike]

Plumbers, electricians, auto mechanics, construction... at least not right now. Computer programming? Probably not, especially with legacy languages like COBOL, at least not without human interaction.

If surgery can be done robotically so can all that.

I am not talking tomorrow but before we retire…

 

[Ingomike]

BTW, I believe this is real and should be feared by humans…

 

[Leadeye]

There will be no jobs immune to this.

AI is learning to write add copy, surgery is moving to robotics, heck it will probably even be a shrink. It will handle all accounting. Just what skills will be immune to this?

Anything that involves being creative and or working quickly with what you have. AI evaluates what it's been told or what it can find out on the internet.

 

[Leadeye]

Years ago I remember how scared we were about nucs and how they would bring about the end of civilization. So far that hasn't happened so maybe the AI threat will go the same way.