[HS]

Of course! Just got Halloween to deal with today and then I'll have some time on the weekend.

[AndyGoode]

Of course! Just got Halloween to deal with today and then I'll have some time on the weekend.

Sounds good. By the way, I have reason to believe that nobody else is working on this idea, and I am personally interested in seeing results, but I don't have time for another project right now. We might even consider a collaboration on this sometime.

[ruebot]

I think that sensible is the opposite of mental, both physically and metaphorically. Sensible actions come from focusing on the senses, mental actions come from focusing on the mind. The camera  trying to see by looking at the monitor is being too mental. It gets stuck in a loop, it was using the wrong tool for the job. Being sensible is using effective approaches in your dealings with reality. (In this case, some kind of concept recognition instead of calculus.) The focus should be mainly on the senses, because they perceive the largest portion of reality.

It could possibly be a language barrier issue but that's not how I would define sensible. That's how I would define sensory.

Sensible is something like stop throwing good money away after bad. A "sensible action" would be to stop doing that.

When I look at a red hot stove burner I don't need to pick it apart into definite bits beyond "That's hot and will burn me". It would not be sensible to touch it.

When I'm listening to someone speak I pick it to bits by observing eye movements, facial expressions and body language, voice inflection, tone, and cadence etc. in addition to what they're saying to (hopefully) make a "sensible" decision based on my "sensory" perceptions.

[LOCKSUIT]

Picking things apart into definite bits is a waste of energy. AGI shouldn't have to break all curves into countless lines to make sense of them, it should only choose to do that if it's a sensible thing to do.

Please define "sensible" for me, ideally without picking things apart into definite bits.

I think that sensible is the opposite of mental, both physically and metaphorically. Sensible actions come from focusing on the senses, mental actions come from focusing on the mind. The camera  trying to see by looking at the monitor is being too mental. It gets stuck in a loop, it was using the wrong tool for the job. Being sensible is using effective approaches in your dealings with reality. (In this case, some kind of concept recognition instead of calculus.) The focus should be mainly on the senses, because they perceive the largest portion of reality.

Yes, if you are deciding about touching a hot stove, or not handing money over to somewhere, or trying to walk around a staircase, you'll need senses, so you can let your wisdom decide what to do with the inputs. Same for R&D.

But internal discovery/[R]&D, is enclosed in the noggin. You don't need external sensory input streams. So your sensible decisions will be internal again but no external sensory. You just decide, store te decision, decide, keep repeating.

[HS]

I think that sensible is the opposite of mental, both physically and metaphorically. Sensible actions come from focusing on the senses, mental actions come from focusing on the mind. The camera  trying to see by looking at the monitor is being too mental. It gets stuck in a loop, it was using the wrong tool for the job. Being sensible is using effective approaches in your dealings with reality. (In this case, some kind of concept recognition instead of calculus.) The focus should be mainly on the senses, because they perceive the largest portion of reality.

It could possibly be a language barrier issue but that's not how I would define sensible. That's how I would define sensory.

Sensible is something like stop throwing good money away after bad. A "sensible action" would be to stop doing that.

When I look at a red hot stove burner I don't need to pick it apart into definite bits beyond "That's hot and will burn me". It would not be sensible to touch it.

When I'm listening to someone speak I pick it to bits by observing eye movements, facial expressions and body language, voice inflection, tone, and cadence etc. in addition to what they're saying to (hopefully) make a "sensible" decision based on my "sensory" perceptions.

Yup, English is my third language, people don't really talk about the downsides of that. I think I'm happy with my definition as "using effective approaches in your dealing with reality". These approaches could still vary from person to person, different things might be sensible for different people. If my ultimate goal is making life as simple as possible, and you had a different ultimate goal or basic premise for life, eg being as informed as possible in order to make the best choices, we might act totally different, yet still both be sensible. I don't know if these basic premises for life can be judged against some definite metric.

[HS]

Yes, if you are deciding about touching a hot stove, or not handing money over to somewhere, or trying to walk around a staircase, you'll need senses, so you can let your wisdom decide what to do with the inputs. Same for R&D.

But internal discovery/[R]&D, is enclosed in the noggin. You don't need external sensory input streams. So your sensible decisions will be internal again but no external sensory. You just decide, store te decision, decide, keep repeating.

Yup, whatever works according to your situation. 

[goaty]

HS is asking the big question, I think on this one,  and i've got my hands full with applied so i've turned my brain off from the final speculation of intelligence... but of course I guess it is important to know,  as maybe the 500 year old aging *babies* in 'heaven' (if you would rather call it hell, maybe that's why its a common greeting in good old English!!) probably ALOT know by now.   Perhaps the little quick-start mortals such as us, who want A.I. in less than 30 years from birth or bust, In a desperation perhaps, will get no more respect than someone that just blankly stared into a corner his whole life, with no need for such brash and pointless development, or "proovence of oneself" which isn't needed at all in the whole scheme of things,  in existence, in this mystical magical realm, we have no true understanding of, as we are brought up in lies, perhaps as god (the mystical one whom no-one can guess why he ever does anything, even if it looks like you do.) wants it to be, in all its gory.

Im glad to see Locksuit has now realized that chess-ai is indeed an agent just the same as any stuff that open-ai are doing,  and remember back when ppl were telling us solving chess is nonapplicable to other ai tasks.  Take ur pick out of ENGLISH RUBBISH OR Yanky FUCKIN-A TRASH!

The way you are challenging Korrellan to say how to think without distinction, is equally a huge way out way of thinking, as thinking of how REAL intelligence works, both HUGE questions, if it ends up impossible, its good to know anyway even if it is.  But also, I would say, working out an impossibility is such a sadness, that you shouldn't so proud of yourself for putting yourself in a conceptual prison. and its *damn classical* to try and get out of horrible hells like that!

[ruebot]

Yes, if you are deciding about touching a hot stove, or not handing money over to somewhere, or trying to walk around a staircase, you'll need senses, so you can let your wisdom decide what to do with the inputs. Same for R&D.

But internal discovery/[R]&D, is enclosed in the noggin. You don't need external sensory input streams. So your sensible decisions will be internal again but no external sensory. You just decide, store te decision, decide, keep repeating.

And those sensible decisions will be based on what? Logic? If not, what?

[LOCKSUIT]

It run on the computer, as all others will too. If you can find an alternative to my computer, then 'my algorithm' can run on that non-logic based computing I think. When you say based on Logic, you mean transistor on/off or AND/OR/NOR ? Can a neural network run its own logic and host programs? How?

And then, he finally came to his senses ;p

[goaty]

It run on the computer, as all others will too. If you can find an alternative to my computer, then 'my algorithm' can run on that non-logic based computing I think. When you say based on Logic, you mean transistor on/off or AND/OR/NOR ? Can a neural network run its own logic and host programs? How?

And then, he finally came to his senses ;p

theres such a thing as analogue arithmetic, and it uses a range of power,  if you do it with capacitors your values are constantly leaking out. =)
Its a lot smaller doing it that way than putting the actual digital adder / divider in,  amazing technology, if only batteries didn't work by magic, then you could actually understand the situation instead of being constantly pestered with non cause and non effect.

A cool real world example is a line follower robot,   it doesn't have any digital logic and it works fine, and is not much bigger than an ordinary motor driver!  Imagine miniaturizing the circuit of this thing, it would be the smallest little dot you could possibly imagine its so simple.   Analogue pong is also HUGELY reducted over DIGITAL pong!


that's the whole thing there I think.

[AndyGoode]

There are many alternatives to 'computers', even beyond analog computers that goaty suggested. You also have cellular automata and neural networks, for example--non-symbolic processors.

P.S.--I totally forgot about Description Logic (DL) as an alternative to regular logic. DL is used in the semantic web, and deals more with descriptions than math, so that's another type of logic to consider. Unfortunately for me, I never understood DL or OWL very well, so I can't recommend DL or compare it to other logics.

https://en.wikipedia.org/wiki/Description_logic

[LOCKSUIT]

if we use physics - cells or NNs, that's universe based logic/math computin..............digital computers have artificial logic

Like korrelan, he uses a computer and has said he uses a physics NN (in the computer) to do the computation....for some purpose of course.....something like that....

[Korrelan]

I'm taking the road less travelled... In actual fact, there are no footprints in the sand ahead of me... I think I'm lost lol.

 

[LOCKSUIT]

Hehe I hinted at that idea before. Hehe. It's fun to be ahead of the pack by 980 miles. But you need to return, and share the bread to thy brethren and show them the way. Your footprints.

[Korrelan]

Like korrelan, he uses a computer and has said he uses a physics NN (in the computer) to do the computation....for some purpose of course.....something like that....

Ok, let’s break this down ‘logically’… ish.

A simulation is not magic, it can’t produce/ provide anything that standard human derived mathematics/ logic can, it is written in a Turing complete language after all.

The whole point of a simulation is to calculate functionality/ properties that are not easily definable using standard logic/ algorithms, usually in parallel and over space/ time where heuristics can be applied iteratively.  This also means that real world sensory data can be included in the calculations/ simulation driving/ changing the dynamics of the system on the fly.

If I wanted a processor that did calculations based on how many red/ blue balls bounced off its surface during a hurricane… I can write a simulation for that.  Any kind of system can theoretically be simulated on a Turing based machine using human derived binary logic… even the human brain.

The difference between my system and most other AI systems is that I’m not using the machines logic/ heuristics directly; I’m not using the computers standard logic to compute and outcome/ result.  I’m using the computer to simulate another type of processor that calculates on different principles, and that provides the results.

This is basically what we designed/ invented computers for.