The roots of intelligence?


Figure 1. The power-of-two based permutation logic for governing the specific-to-general wiring and computational logic of cell assemblies. (A) The equation defines the size of a cell assembly; the numbers of neural cliques within a cell assembly. By following the permutation equation of N = 2i–1, the cell-assembly motif exemplified here consists of 15 distinct neural cliques ([I]N1…15), which cover all possible connectivity patterns in order to process four distinct inputs ([I]i1, [I]i2, [I]i3, [I]i4). The specific-to-general neural cliques shown in this subpanel illustrate the logic for wiring non-recurrent networks (e.g., the hippocampal CA1). (B) Schematic “bar-code” illustrates the specific-to-general cell-assembly activation patterns, which can be measured by electrodes or imaging techniques, from the 15 distinct neural cliques (N1–15), processing four distinct inputs ([I]i1, [I]i2, [I]i3, [I]i4). The orange color represents the stimulus-triggered activation above the baseline state (in blue). The arrow on the right side illustrates the number of distinct neural cliques exhibiting specific, sub-combinatorial, as well as generalized, responsiveness. The cartoon illustration was adopted from Tsien, TINS, 2015. Specific neural cliques encode specific features, whereas various permutation rule-based neural cliques encode various convergent patterns, representing relational memories and generalized concepts.

[/I][/I][/I][/I][/I][/I][/I][/I][/I]

Some algoritms for raytracing(if I am in the correct) like Metropolis ‘remember’ where the ray has shoot. Is a basic form of convergence, but is a good example. Probably the human never thank that a machine can think, we allways create them, but the point is make easier our lifes. Now we design, program, test and use machines, and now we make AI’s.

this single most important thread in BA and it gets 1 response

we may have AI smarter than a human in under a year…

this is good, terrible and amazing.

To be fair, if you want more people to respond to the “single most important thread in BA”, then you should put more effort in the original post instead of putting up a random link, copy/pasting random bits from said link, and hoping other people would provide the discussion.

AI smarter than a human in under a year, neat. The real question is when do we get to take the pill from Morpheus and rebel against our Skynet overlords?

I am still trying to understand it all*

I just thought that it was really really interesting.

If you don’t understand it why are you posting it without any context?

If you really think it’s important you should at least make an effort to explain why you are posting it and try to start a discussion.

You have a habit of posting random links and treating this forum as facebook, hoping for likes and comments.

The problem is that not everyone uses the same metric as you do when judging a thread’s importance (heck the metric used may even be unique to each member).

If I wanted, I could deem my Cycles idea threads as the most important on BA, but doing that would be egotistical and assert the idea that I’m the superior person when compared to everyone else on the site (people who disagree then are either ignorant or just dumb).

Yeah, this is interesting and it poses great significance to anyone whom is a human if it is true.

People here write complex algorithms, they may shed light on this for me.

I Imagine now that we know how it works, we can begin building methods to enhance a humans intellect.

Can any coders here simulate a ant or something using this?

I am interested ai.I find this interesting.I would like to see more advanced human like androids.You have one more response.Plus i would like to see videogames use advanced ai in them.I tend to disagree that this would not be useful in videogames.

The problem won’t be if this goes into videogames. There’s lots of algorithms from the AI cosmos already implemented in games, and there’s absolutely nothing that show us that will stop.

But the big picture here is other. For example as Pedro Domingos states in his ‘The Master Algorithm’ (a must for any interested), the ability to have machine learning, will be a jump into knowledge without precedence. Machines will be able to choose different paths, based on previous assumptions or data, find equivalences that normally difficult for a biased human mind, propose predictions and solutions. And all this to be used in all fields of knowledge.

A new variety of games will definitelly emerge, with almost real AI, not only in enemies actions, but in most every part of the game (Physics, Graphics, Scenery Generation, adaptative game logic, full interaction ambients, complexity over software size, to name the more visible factions of the change)

It would interesting to have a npc have something close to real creativity in a videogame.At one time they made a ai that could make videogames.But it was not very good at it.It would be interesting if ai algorithms could make a videogame via description.You could input a story or book even.And it would make a great videogame of it.

For that, there is still the need for powerfull computers (as there is much more to human mind than just some reasoning), but if we solve the machine learning problem, we could design better consoles in a fraction of time we do it nowaday, specifically build for handling these kind of algorithms in an efficient time span.

They say molecular computers are a lot better than quantum computers.Maybe it could help
make computer algorithms think similar to humans.Maybe you should read upon on the recent developments in artificial general intelligence.I read a title about that but i did not read the article.I do not know what websight it is on.Because i am into videogames and 3 modeling.It probably would be long read.

Part of the reason the human mind is so powerful is it’s ability to change; work from a problem backwards to forwards, visualize a problem in terms of past knowledge, metaphors, or from the perspective of an entirely different discipline, etc. I’m not a programmer but from my knowledge of algorithms and code there is always only one mode of thought going on. In a super intelligent ai there would need to be transcendent algorithms constantly monitoring all processes below it with the ability to change them on the fly. The same way the human mind can work a problem from different angles. Even then, it still needs to be able to develop new transcendent algorithms above itself when a change is needed, essentially creating levels to itself that are never defined.

The biggest problem with super intelligent ai is that when we push “run”, we are no longer in control. Without predetermined parameters and thresholds like our genetic makeup in humans, there is no telling what logic will be generated or if it will even make sense to humans.

The only way, I think, we can develop effective super intelligent (human like and beyond) ai is by giving them a fluctuating multilevel algorithm like I mentioned but with extremely strict parameters to start and slowly giving the ai more reign of it’s own intellect.

I think we would be better off enhancing our own minds, than booting up new ones.

Just add new hardware that the brain recognizes and migrates into by connecting synapses etc.

That is probably the safest bet for ensuring human continuity. It’s definitely possible, but we might be a long ways from understanding how to translate information into something the brain can directly read.

This is imaginative, but maybe the future of humanity will have real upgrades for purchase to enhance our abilities how we so choose. I’ve already heard of working optical enhancements for nightvision and seeing more of the electromagnetic spectrum. There are definitely some moral and philosophical hurdles to get past before we can get there collectively. Still, I think there is a place for human minded ai in our future, especially for emergency situations or autonomously exploring conditions that people simply cannot go and then relaying information back.

But what if people could turn off your emotion with them or gain control of your brain through them.You probably could get cancer of the brain or other health issues from implating electronics in the brain.Because they already said that cell phones give you brain cancer.

First off you are already a computer, second off if we understand the mechanism, we can extrapolate and build our own, and no I would not add wifi or nearfield, or even an os, your brain is the os, and its just more “wetware”

Think of replacing parts with better parts that just don’t age.

The main problem is that we still don’t understand our brains.
We have some lights about some of the basic mechanisms but that’s almost nothing. It’s like trying to build a flying jet plane by looking at some photos.

The second problem is that our minds are not computers in the sense of performing the same operation and get the same result or even just being logical. It just doesn’t happen. We fail much more than we are capable to admit or even remember (some fails don’t even get to count since people die rigth at the moment).

The third problem is that machines will need to learn, and humans are still very bad teachers. We have lots of difficulties separating logic from reasoning and reason from feelings.

And this list doesn’t stop here. Aside some small amount of good information available, the most information we can find is (if not totally, then at least partially) wrong!