re: Scientists and Columbia University Create Self Aware Robot

quote:

---

What the purpose of this? Stupid idiotic scientists

---

Tuning controllers can be a tedious process when tasks get more complicated. Phenomena that involve non-negligible nonlinearities sometimes can’t be tackled with traditional PID controller based approaches and require nonlinear controllers based in state space models or using something like sliding mode control if you can map your system to a phase plane.

Point is, it takes money and resources to make controllers for robotic systems. Machine learning algorithms allow for the potential that infrastructure can be built and could optimize the control laws on its own without that additional labor. Many of these algorithms are readily available online and the science behind them and neural networks and all of that is all mostly open source. You can find many YouTube videos showing people putting them into various simple video games and making an AI play the video game and learn how to optimally play it, they are pretty cool.

quote:

---

We don't even understand consciousness and yet we are going to program a bunch of 1s and 0s to magically develop consciousness from nothing?

---

No. A.I. intelligence will come from quantum computing, similar to our own brains' abilities. It won't be programming, either. It'll be learning, the same as we do.

quote:

---

We will continue to develop machines that are very good at the tasks they are programmed to do.

A machine would not have a desire to exist and protect its right to exist unless it was programmed to do so... if that is even possible.

---

Quantum computing is in its earliest stages. As it advances, artificial intelligence will become a reality. When A.I. becomes self aware it will be a new form of life, not just a machine.

Last I heard IBM had an 8-qbit quantum computer working, it’s just nobody has a good way to scale the technology yet. Lot of research going on in that sector and it’s exciting what it’s gling to do in industry once the problem of scale is answered.

In my field CFD and other similar types of simulation work is going to greatly benefit. Specifically CFD is going to be able to get down to length scales to finally solve full turbulent fields without a URANS approximation. That is pretty exciting stuff.

The should make the robots with interchangeable penis or vagina (and bolt-ons) since they plan to make them self-aware. They should get to choose, amiright?

quote:

---

No. A.I. intelligence will come from quantum computing, similar to our own brains' abilities. It won't be programming, either. It'll be learning, the same as we do.

---

Late bump, but No.

We already have computers much more computationally capable than any human brain when it comes to specific tasks. The problem is that computers at their heart are just a series of instructions. Consciousness doesn't magically arise out of computational horsepower. We don't even understand the intricacies of how our brain processes, much less make a computational device that matches how we process data.

It doesn't matter if I have a computer that has a million more times computational power and memory of modern computers... it doesn't produce a conscious being. In the next 20 years we will have robots that can do a lot of human tasks and we can even make them look and act like people, but that doesn't make them any more conscious than my cellphone.

When IBM and Google talk AI, they are hyping up advanced pattern matching algorithms... not anything resembling true AI.

I'm going to write to him with suggestions about additional programming.

if he writes back I'll share my letter and the results.

Yeah, those Boston Dynamics terminator robot dogs aren't being created as robotic pets for human use. Those beasts are being perfected for use in the inevitable roundup of humanity and/or firearms.

quote:

---

You’re focusing on the literal meaning of words instead of the ideas they are trying to convey. 

Who gives a shite if it meets the literal definition of consciousness if the average human will be unable to tell the difference? 

We could even get bogged down over the word “artificial intelligence”, but that effort would only serve to avoid what we should really be discussing. 

This is a fascinating topic, primarily when it comes to the possibilities of what blending artificial intelligence with quantum computing could produce for us. 

Just as we find it remarkable how much smart phones have changed in the few short years they have been around, the growth and advancement of AI will be much more profound once we cross that unknown tipping point. 

---

No I am focusing on reality. Whenever someone says AI, they are really talking about machine learning... Which is a limited pattern matching algorithm. It's a fantastic thing, but much more limited in scope than pop culture imagines it.

The problem is the general public goes from hearing the term AI to believing we are on the verge of something seen in the movies. And then you hear Elon Musk spout nonsense about AI and everyone loses their minds.

Robots capable of walking, running, firing and engaging enemy combatants is something with possible terrible consequences that we will have to deal with before AI ever will be an issue.

Machine learning is definitely an awesome thing with a ton of untapped potential, but real AI it ain't.