Eureka Machine Discovers Newton's Laws in 2 Hours

Captain Kremmen said:
@Pincho
Your example of making a tree with Morse Code is a good one.
And a good way of learning Morse Code should you have the desire to do so.
http://forum.thegamecreators.com/?m=forum_view&t=194548&b=2
Click to expand...

Well I posted it to show that I work on these things in some ways, and posters are trying to teach me my own subject. Missing the point of my replies in the problems with it. In that you need my theory to actually use it, else you are forcing it to use Newton, and Einstein just by giving an electron mass.
 
The main problem I see here is that a human has too program the data into the machine. The machine can't infer by itself what to study. Which means that although it is great for making a process more efficient, new discoveries are unlikely and biased by what a person may think a number represents. In the case of studying the universe it would be hard to tell the machine exactly what to study. We can hook this thing up to antiquated detectors, satellites, and telescopes and only be able to show what we already knew in a much simpler context.

What we need is a self sufficient machine capable of viewing every form of motion deep inside structures at the atomic level and simultaneously follow galactic movements. With something like that not only could we confirm our discoveries but we would have more insight into vague ideas such as black holes, DNA, and atomic transmutations. That way we could tell not only what something does or what characteristics it has and instead infer why and how these basic and simple objects work.
 
By Newton's laws, I suppose the law of gravity is the one meant.
I'm surprised it would take two hours.
It was the idea of some mysterious force making masses attract each other that was the breakthrough, not the calculations.

What else could it come up with?

As for Galileo's work, ie, that speed of descent is independent of mass and accelerating.
5 minutes should have done for that.
 
If I fed a machine the data "A", "B", "C", "+", "=" and "2" then told it to spit out every possible combination it would eventually "discover" the Pythagorean theorem. It doesn't mean that it comprehends trigonometry or even basic arithmetic. I don't know what the Eureka machine is doing but it's nothing outside the capabilities specifically designed by the scientists.

I'm not saying it isn't possible to discover the "Intelligence equation" but its implementation would require much more than pattern recognition. I'd be interested in seeing which results were disregarded and why (nonsense maybe?). What would happen if we fed it data that had no logical coherence whatsoever? What about data on many different subjects? Could it differentiate between them? Or would it instead try to find a pattern than doesn't exist?



On a side note, I dislike how loosely singularity is used these days... but you might be interested in the Omega point, it seems to run along the same thought.
 
I know you guys are way beyond this but awesome quote Pandaemoni!!! I'm gonna check that book out now!
 
impaJah said:
I know you guys are way beyond this but awesome quote Pandaemoni!!! I'm gonna check that book out now!
Click to expand...

It was also made into a movie
Colossus_the_forbin_project_movie_poster.jpg


And there is supposedly a remake in the works starring Will Smith.
 
et al,

(MY QUESTION)

The "Eureka Machine" appears to dates back to about early 2009; Cornell University in New York. It is now 2012.

  • What is the Eureka Machine doing now, three years later?
  • Is it still producing? (If so, what?)

Most Respectfully,
R
 
Chipz said:
http://fora.tv/2012/01/19/Cognitive_Robotics_and_Artificial_Intelligence

Here's a much longer discussion by some of its developers. It is primarily focused around the end.
Click to expand...

RoccoR said:
et al,

(MY QUESTION)

The "Eureka Machine" appears to dates back to about early 2009; Cornell University in New York. It is now 2012.

  • What is the Eureka Machine doing now, three years later?
  • Is it still producing? (If so, what?)

Most Respectfully,
R
Click to expand...

I posted this link to a discussion, it's from January 2012.
 
You must log in or register to reply here.
Back
Top