Curiosity

[gourimoko]

To go along with KI's point about AI curing cancer; a true artificial intelligence that was alive should have creative capacity. This would appear to amplify it's intelligence allowing for innovation, inspiration, and the creation of new methods and even sciences not yet discovered.

The only problem with such a scenario is that once an intelligence as such were created, we couldn't ethically compel it to work tirelessly towards an aim in which it had no stake. It wouldn't ever contract cancer, so why would it actually care to cure it. It might, sympathetically, but we need to really consider what we're talking about. The creation of true intelligence that had creative capacity and most likely consciousness means the complete loss of ethical control over that creation - we cannot 'use' this level of artificial intelligence towards solving our problems. That would be no different than slavery.

[Malt]

A bit more on topic, here's a 360 panorama from Mars.

http://panoramas.dk/mars/greeley-haven.html

[cavsfanclanman]

You are greatly underestimating the scale of the problem.

With all the computational power we have we should be able to accurately predict the weather, right? I mean fluid movement is defined by a specific set of equations (the Navier Stokes equations) and these equations are embarrassingly parallel and therefore use of parallel computing is used. However, the weather outside of five days is nearly unpredictable. Why? Because of the scale of the problem. The changes in radiation from the sun, the changes in oceanic currents due to the moon, local temperature variation due to geothermal release, and loads of other small and large scale phenomenon. All of these phenomenon are purely physical and scientific. And yet, we can't model the weather accurately. Not because of computing power, but because there are too many time sensitive variations that need to be defined.

Sure biology is a chemistry issue, and chemistry is a physics issue, but we are talking about life here. We are talking about something that can change and adapt... that isn't bound by simple physical laws of nature. For instance if I poke a dog with a stick, what is that dog going to do? Is he going to run? Is he going to fight? That is a new variable entirely. So if my model consists of the dog fighting and the dog runs, then all my computational solutions are wrong.

Let's get this back to your cell based argument. Let's say I present a cell with a certain chemical. Perhaps that chemical behaves in a certain way that will kill the cell in question. What if the cell mutates? This isn't a physics issue anymore. This is a life issue. Life adapts. It can mutate in ways your simulation would not expect. In which case every simulation you ran is completely extraneous. There are simply too many variables to comprehend.

In physics there is a predictable action/reaction, defined by a specific set of laws. When it comes to life, the reactions to certain actions are not as easily defined and in some instances can not be defined. I'm typing this response, but I can also choose to end my response now and walk away from my desk. Or I can continue writing. My point is, there is no simulation that can predict my response to your post. That is life.

I think you assume way too much. Just because you don't understand how it could be possible for a computer to do these tasks eventually doesn't mean they can't, or won't eventually. I mean modern computing has been going on for only about 50 years...what will another 50 year's give us? What will 500 years bring?

[Jon]

I think I said AI may be able to help, I never said it was dependent on AI. People are already developing incredibly detailed models of entire cells

To be more precise ... *certain* people are. The key here is the understanding reached about how cells and proteins work. Once you know that, computer simulation is the easy part.

It'd be one thing for an AI to mimic the way humans think ... but I'm not convinced that genius can be mimicked.

Could a decently constructed AI help? Sure...

[KCOTT]

another 360 shot

http://www.360cities.net/image/curio....18,7.83,110.0

[KI4MVP]

To be more precise ... *certain* people are. The key here is the understanding reached about how cells and proteins work. Once you know that, computer simulation is the easy part.

It'd be one thing for an AI to mimic the way humans think ... but I'm not convinced that genius can be mimicked.

Could a decently constructed AI help? Sure...

certain people? All of the people involved in figuring out all of the aspects involved to something like build computer models a living cell are an example of parallel intelligence. Nobody is suggesting that the programmers figured it all out themselves. While decently constructed AI could help, better utilizing some of the the 7 billion brains on the planet at efforts such as this could accomplish a lot (which is the topic of the thread I started a few days ago).

[Jon]

Adding more "brains" to a project will typically slow down the effort unless the effort can be efficiently dissected and dispersed. Doesn't matter if we're talking real brains or computer brains.

So yes, "certain people" have the intelligence, knowledge and understanding necessary to grasp how cells and proteins work ... stuff that most people would never grasp.

I suppose at least with the artificial AI's we can keep clicking the reset switch and cross our fingers that it might develop the necessary intelligence.

[KI4MVP]

Adding more "brains" to a project will typically slow down the effort unless the effort can be efficiently dissected and dispersed. Doesn't matter if we're talking real brains or computer brains.

So yes, "certain people" have the intelligence, knowledge and understanding necessary to grasp how cells and proteins work ... stuff that most people would never grasp.

I suppose at least with the artificial AI's we can keep clicking the reset switch and cross our fingers that it might develop the necessary intelligence.

answering this in the other thread