re: War Room guest Steve Cortes says AI is beginning to negatively affect employment in US

quote:

---

But AI is a different animal I don’t understand it, and it scares me

---

I don't understand it either and find it alarming... but, I do believe mankind can and will adapt to environment as necessary. Always have so done before, likely will again and again.

HR is worthless.

Good riddance.

We got to stop trying to exterminate ourselves first.

AI can be be beneficial no doubt but the problem is the world elites could use it for nefarious purposes.

Walmart has been cutting back on self checkouts due to theft.

In my local Walmart, someone tried to leave with $200 worth of stolen groceries.

Is this anything like what all the top climate “experts” have predicted?

quote:

---

I don’t understand it, and it scares me

---

What parts do you want to understand?

quote:

---

What parts do you want to understand?

---

The fear parts

Can it actually “take over” and do whatever it wants?
When guys like Musk have concerns it’s a little scary

The possibilities for doing good and achieving great things is exciting

quote:

---

Can it actually “take over” and do whatever it wants?

---

I guess there are two answers to that.

Current non sentient AI's have no wants. They work via mathematical reward mechanisms that we give them.
Humans are constantly tuning the data that goes into even unsupervised models to get the desired output.

AGI will be a point where computers achieve human like intelligence, and are capable of learning and replicating anything a human can.

Now there are a number of limitations with that.
Humans don't cost Billions to train.
You can only train them where there is sufficient data.
Even under Super AGI (Artificial Super Intelligence - ASI) this holds true, but the AI would be able to exceed humans in problem solving.

AGI is the goal for the foreseeable future, ASI is more of a theoretical state.
In both of these cases someone would have to choose to train the AGI or ASI to do so.
An AGI even if trained for said purpose would miserably fail.

An ASI would probably also fail due to the complexities of the real world. Aka even "smarter humans" wouldn't be able to take over and rule the planet.
I mean look at the US in Afghanistan.
The IQ difference between the US planners and the average Afghani was at least 50 points.
Even with Quantum computers we don't have an architecture to get beyond really smart problem solvers, even if they were trained to specifically take over.

Now the second part.

People seem to really want to create a sentient AI, we don't know how to do that.
But based on all other AI we have the problem comes down to 3 major parts.

1. Math: We would need to develop a family of algorithms that allow what we determine to be sentience.
2. Computer Architecture: A hardware/firmware/wetware/software structure that mimics the brain or better. I've seen papers on weird things like cloned rat cells in a petri dish hooked up to electrodes.
3. Power: Training cost could be enormous, Humans use about 2x the energy of a computer, but about 1/6th the energy used by a supercomputer, and 1/20th of a quantum computer. And they haven't even gotten to AGI yet.

Now to note, even if we had ASI super intelligence, and sentient algorithms,
they would probably not work together.
So your first sentient algos would probably be fish like intelligence.

We also currently only have 2 major paths to this, Bio-mimicry and Quantum.
Bio-mimicry has potential as it was the foundation of Neural nets.
But it goes far beyond that with experiments into wetware.
Fields of cloned human brain cells in a massive petri dish is a scary thought.
On the quantum front, super AI looks like we can get there. the multistate computations allow faster and more complex looks at reality for problem solving.

So in summary, Super Intelligent AI would not be a threat. Quantum is likely to get there in our lifetime.

There is no algorithmic path to sentience without bio-mimicry.
Wetware could get to sentience, but that doesn't mean it could get to Super Intelligence.
Our brains are wetware, sentient, and intelligent.
So the answer is it's far enough into the future where it's still in the realm of science fiction, but some people want to get there.
There is no guarantee sentience outside of biology is even possible, everything we know says it is not.

That does not stop some Anti-God people with a god complex from trying to create life and create sentience.
That's the problem in my opinion.
A lack of human vitality.
Helpful ASI and robotic aids that we rely on for everything.
We stop struggling and thriving, we stop having children because we can live a life of ease and entertainment as permanent children.

And then we just die off as we never mature and continue the species.
Life gets too easy.