|
Recent Accomplishments (pg. 8)
|
View this Thread in Original format
| r5a |
| quote: | Originally posted by Enigmatik
I went out once in the past year (which was the re-opening of Stereo Afterhours)... So no, not really. :sadgreen:
The "new me" is no fun at all, I've been sedated. The party animal is dormant until further notice.
What about you? Update me. |
check yo pm's |
|
|
| w_ashley |
| quote: | Originally posted by fayraree
what other kind of AI is there? |
AI is a genralized term for theoretical system that have the capacity to adapt intuitively.
"Brain Inspired" AI might be generalized to Neural Network based AI, although to say brain is to imply something far more involved.
Ex. Brain inspired would entail not only data transfer and arrangement based upon inputs, but it would also entail a system to provide for survival and continuance of the system based upon environmental inputs, and any characteristics inate with the system.
AI can be as simple as aparatii can ajust in N ways, AI system will adjust until N_? acheives success flag rather than failure flag.
Likewise it could be as simple as having a way of recording a state based upon last state or all states leading to a given flag in a specific locii of operation. Then an ability to discern the loci the desired outcome and the needed state. Simple AI systems would then use a matrix or array data arrangement to track multiple entities or variables. Adding complexity to an otherwise basic system.
Not all AI need be "brain based" there are other forms of intelligence that are created by artifical means. Also brain based is a generalization which could entail a form a CI or HI (conscience intelligence or human intelligence).
So yes there are non brain based AI systems. Some are very complex - to say that human intelligence is house in the brain alone though would be a false dichotomy human intelligence is univeral and not wholely biological. The central processing point of that system is seen as the brain but once again that is only currently tied in for biological existence, non scientific cultures have had other means of definition of the totality of human capacity and state.
I can only hope everyone here is not a slave to science and can view the discusion at an intellectual other than solely science based reality indoctrinated level. I think it is essential to really even understand existence or intelligence. Anything less is just ego denoting satisfaction of what you know about something rather than knowing the full extent of what you know. |
|
|
| pkcRAISTLIN |
| quote: | Originally posted by ********
I can only hope everyone here is not a slave to science and can view the discusion at an intellectual other than solely science based reality indoctrinated level. I think it is essential to really even understand existence or intelligence. Anything less is just ego denoting satisfaction of what you know about something rather than knowing the full extent of what you know. |
ie, I hope people don’t realise I have no ing idea what I’m talking about. |
|
|
| chlola |
| quote: | Originally posted by EricB.
Oh also I stopped in with dumb bimbos and raised my standards, started to deal with girls with more substance and who are a lot smarter. Its working out for me and chicks no longer stress me out. Well I take that back.. They stress me out less. |
Awesome! |
|
|
| zoogla |
| quote: | Originally posted by ********
Not all AI need be "brain based" there are other forms of intelligence that are created by artifical means. Also brain based is a generalization which could entail a form a CI or HI (conscience intelligence or human intelligence). So yes there are non brain based AI systems. |
why are you so ing stupid? |
|
|
| gehzumteufel |
Lost 20lbs
Got a fantastic job where I work with friends and make much better money
Just got back from Tokyo
Going to visit some family in August
Hoping to see a shuttle launch in Sept
Getting a motorbike in the next couple weeks
Gotten in contact with some family I haven't really been in contact with for about 7 years. |
|
|
| tachobg |
| quote: | Originally posted by fayraree
what other kind of AI is there? |
That's a good question. Technically, all AI is "inspired" by the brain, because one of the original motivations of AI has been to achieve human-level intelligence (and then some day surpass it). However, the methods people have tried to develop have changed quite a bit in the past 50-60 or so years.
From what I understand, AI research has gone through a few major paradigm shifts. Initially, people were focused on building logical systems that manipulate symbols based on precise rules to produce a result. This is very far from what the brain actually does.
Later on, neural networks became popular. But neural networks had a lot of problems and generally failed to deliver on the hype. In addition, despite being "brain-inspired" most models had little to do with how the brain works. To give you an analogy, it's like putting a simple circuit together to model the computational properties of a digital computer. It does enough interesting things to keep you interested for a while, but it doesn't come close to what a digital computer does.
Most recently, people have been developing mathematical and computational theories of learning. These are mostly based on statistics and are most commonly known as "machine learning" or "statistical learning theory."
The approach that I'm involved in right now differs from all of these. In this approach, one says, "well, if the brain is really good at telling apart dogs and cats, then why don't we look at the brain and figure out the important pieces of the brain that make that happen?" Surprisingly few people are doing that. The brain is either ignored in favor of simpler, mathematically tractable models, or people who study the brain only look at tiny pieces of it rather than the big picture. In the latter case, I think is a symptom of overspecialization in academia, where to be successful, you have pick a sub-sub-subfield and focus on that. In a (well-funded) company, you can afford to take more risks.
This is definitely not a comprehensive overview of the history of AI, and you might want look for some more authoritative materials on that. But to sum up, historically, AI has most definitely not been very close to the neurobiology of learning and intelligence. |
|
|
| couch-potato |
| quote: | Originally posted by tachobg
The brain is either ignored in favor of simpler, mathematically tractable models, or people who study the brain only look at tiny pieces of it rather than the big picture. |
The big picture is a rather complex thing :p |
|
|
| tachobg |
| quote: | Originally posted by ********
I can only hope everyone here is not a slave to science and can view the discusion at an intellectual other than solely science based reality indoctrinated level. I think it is essential to really even understand existence or intelligence. Anything less is just ego denoting satisfaction of what you know about something rather than knowing the full extent of what you know. |
Correct me if I'm wrong, but I think you mean that artificial intelligence is worthy of discussion on a philosophical level rather than solely on a scientific level. Both of these have their roles. In machine learning, determining generalization guarantees for a classifier is a purely theoretical thing. In the philosophy of mind, discussing the nature of consciousness is a purely philosophical thing (for now).
It is necessary to know the distinction and to combine them cautiously.
Consider this. Science is subject to evidence. This hold up to evidence or they don't. Philosophy is different. You can have scientific endeavors inspired by philosophical considerations, but philosophical considerations don't give you scientific results -- the ultimate judge is whether the theory reasonably holds up against scrutiny given the evidence. On the other hand, scientific results *can* have philosophical implications. For example, as we have gradually discovered that we humans are not the center of the universe, this has caused a shift in perspective and a lot of philosophical discussion.
What I'm trying to say is this -- scientific results can sometimes invalidate (or make irrelevant) certain philosophical questions. But philosophical ideas can *never* invalidate a scientific theory, which is subject to experimentation and hard evidence only.
Ultimately, what this means for consciousness is that scientific theories of the brain, intelligence, and consciousness will inform and direct the philosophical discussion on these topics. |
|
|
| tachobg |
| quote: | Originally posted by couch-potato
The big picture is a rather complex thing :p |
Absolutely :p This is where one needs to come up with simplified models that capture the important things that are going on in a system. For example, modeling the complex physics of every neuron in the brain, to me, seems like saying "I don't know wtf is going on so I'll just model EVERYTHING!" Picking the right level of abstraction to model things (the right level of "zoom" into the system) is extremely important.
For example, knowing that certain parts of the neocortex use a certain mechanism to remember patterns is a lot more useful than knowing how the voltages at each neuron membrane vary with time. Once you know what parts of the system do at a high level, you can treat them as a black box and ignore how their functionality is internally implemented. |
|
|
| Comrade Stalin |
I've constructed an options trading model that made 7% last week in a test and will soon use real capital with it. I am now developing a currency trading model to integrate into my macro-economic grand model. This thing is going to be a beast.
EDIT: Hah, looks like I can and will be combing my option, macro, and currency models into one!:D |
|
|
| zoogla |
| quote: | Originally posted by tachobg
Technically, all AI is "inspired" by the brain, because one of the original motivations of AI has been to achieve human-level intelligence (and then some day surpass it). However, the methods people have tried to develop have changed quite a bit in the past 50-60 or so years. |
This was my understanding, and I was also aware of the variety of methods to develop it. Your approach seems very interesting, although all approaches are looking for ways to simulate independent thought through logical, heuristic, sensory, etc. perception and focus on a single or multiple parts of the brain (like those lifelike robots in Japan who develop goosebumps when it's cold--mainly for sex lol). |
|
|
|
|
