AI Research

How do I start? How do I build upon it? I want to know everything

Picture mostly unrelated

Attached: ClipboardImage.png (2816x2112, 7.2M)

Other urls found in this thread:

khanacademy.org/computing/computer-science/algorithms
openclassroom.stanford.edu
youtube.com/watch?v=P5iqYuFmzqg
youtube.com/channel/UCWN3xxRkmTPmbKwht9FuE5A
twitter.com/AnonBabble

Go to school. Get a job.

Math (Calculus, Linear Algebra, Stats)
Learn to program.
Read SIEGE FAGGOT

Any neuroscientist will tell you that AI will never even *begin* to match a chimpanzee, let alone a human in terms of sapience. Intelligence is not artificial. In general, we tend to exaggerate the uniqueness of our age, just as previous generations have done. The very premise of the singularity theory is that progress is accelerating like never before. I argue that there have been other eras of accelerating progress, and it is debatable if ours is truly so special. The less you know about the past the more likely you are to be amazed by the present.

You start here: khanacademy.org/computing/computer-science/algorithms . Don't worry if that looks like too much maths for you. The Pajeetacademy has got you covered with everything you need to understand all the way back to elementary school. Once you got this you can go on to openclassroom.stanford.edu .
Of course, you'll never do anything like this. Because you are too retarded to search for free education resources.

Fuck off.

In addition to what everyone says you probably want to try to build a chan and run it too because you always need new data sets on which to operate on. A chan should provide you with that.

The singularity in nothing more than a bunch of jews fellating each other over moving away from nature and truth ever more.

I think the only way to beat materialism and scientism is war at this point. God has failed or at least the Christian God has. It's war or Jewish techno golems.

Nah, we don't need a real (civil)war. A good old pogrom will do. It's a tiny minority of (((people))) who are causing most of the problems of Western civilization, kill them and revoke their (((laws))) and things will go back to normal in no time at all.

>tiny minority of (((people)))
They number around 15-20 million around globe. That's a war-scale pogrom. Not to mention shabbot goys would rather protect them than live happy. It's war.


Never. Christ never fails. Have faith. Problem is rapidly changing demographics, which gives us preciously little time to fix things. Without that, we know that secular materialists aka Jewish techno golems like you put it self-terminate their genetic and thus memetic lineage within next two generations. I know people who have sterilized themselves to "combat climate change" and "take global responsibility". It's the years 1960-2080 that are years of ever increasing danger and hardship, then Kali goes back to sleep. The meek shall inherit the earth.

Tell us all about the great calculating machines available in the past. You are truly delusional if you think every single kid and soccer mom carrying around the equivalent of a supercomputer in their pocket, instantaneous and globe-spanning communications, and the ability to jet anywhere in the world are not a game-changing combination the likes of which have never been seen before in history.

The current AI craze is a huge joke. Soon it'll burst and we'll reach the same conclusions we already did in the 80s: AI is good for solving certain kind of problems, but can't approach general intelligence.

...

this. proof is this thread

png is less bloated than jpg

Get into applied stats, do a course on machine learning and read the man page on strstr().

>photos measured in megabytes are less bloated than perceptually assuming the image wasn't lossy compressed before hypersperg png'd it identical images measured in kilobytes
Also, shame on Chodemonkey for not adding JP2/WebP/SVG capability.

>>>/prog/3034


it could also be that the substrate itself is not fit for consciousness, that silicon switches can't replicate the *physical* behaviour of our brain orchestration (quantum consciousness/neocartesian dualism)


Soap box, ballot box, jury box, ammo box, lunch box, and now for Zig Forums, the open box

For a long time we have also had machines capable of performing superhuman tasks. Think of the clock, invented almost 1,000 years ago - it can do something that no human can do: tell how many hours, minutes and seconds elapse between two events.

Right. I do not want to be interpreted as minimizing the progress of neuroscience and/or AI/ANN researchers but the hype surrounding this second (or third?) AI revolution is way overboard.

From a scientific and philosophical standpoint we're completely in the dark on what consciousness ``is``. We're so uncertain that some of the most prominent advocates of the view that we understand it go so far as to claim that ``it doesn't exist``. Maybe Dennett, himself, is an actual manifestation of the philosophical "zombie" so often used in thought experiments.

The difficulty from a scientific point of view is that there is no way to turn conscious experience into a symbolic relationship capable of being scribbled down on a piece of paper. What would an equation signifying the relationship between the canonical variables of physics and a qualitative, subjective state of consciousness look like? The problem is that every equation we build in the hard sciences is a relationship between two strings of symbols, each representing something outside of the "nexus" that is consciousness. The creative thought that goes into relating the symbols can, maybe, best be identified with the "=" symbol. I'm not a great philosopher so maybe my imagination isn't strong enough to chart this territory but, in any case, I don't believe it is possible for that which gives rise to symbolic representation to be represented symbolically, itself. It's the Logos. The Ouroboros.

Even in the case that we grant that the impossible task above has been accomplished we don't know at which level of physical analyais the relationships might hold. Molecular? Atomic? Subatomic? Subsubatomic (quarks)? Deeper than we're even able to describe right now? The nature of the physical systems involved is difficult to imagine even if we grant that we're able to imagine it somehow.

It's a real shame that the utility of science has been weaponized against society. Sophists with political aims are all too willing to deceive the lay public, who often do not have the time, means, or desire to learn about the history of our knowledge. "Trust us," they say, "throw away the traditions and beliefs of your ancestors. Throw away the beliefs which have actualized in the most advanced, free, and contemplative civilization yet known. Those beliefs are no longer required (in fact they're toxic!). Trust us! The scientists know what's best... now get back to laboring and consuming!"

Notice that modern "AI" devs have categorically rejected the Turing test as outside the scope of anything they have even the vaguest plans of attempting. We're making all sorts of progress, but not in anything intentionally related to reverse-engineering human consciousness.
Dualism a shit. Whatever it is, it's entirely at the molecular level, since the complete designs are stored in DNA that we can artificially synthesize from scratch, and constructed exclusively using molecular chemistry (granted, chemistry complex enough we haven't quite managed artificial abiogenesis of a whole organism... yet). If atomic or lower transmutations were occurring, the amount of energy needed would be quite conspicuous.

continue... not even remotely in the same category of advances i described. you'll need to do better than that to convince us we're no different than the dark ages today user.

kek. this.

I like the way you think. I'd suggest that insofar as the mystical nature of the human soul and human spirit, there surely exists some 'membrane' or interface between that which is physical, ie the human brain and all it's bio-electro-chemical nuance thereto, and that which is non-physical, namely the soul. The spirit is much deeper still.

Descartes was pretty smart, but the Biblical writers were wiser still, having the Spirit of God Himself.

I never brought dualism up. If you believe that it is implied by anything I wrote and would like to rebut that then you're free to. Also, you may have missed that the portion you quoted re: dualism was part of a case that I don't even consider to be important due to the preceeding paragraph. If you think it is possible to model the conscious mind in the same way we model observation and theory in the hard sciences (i.e. symbolically) then perhaps you might give an example of what this would look like.

Christcucks will be forever btfo when general AI is developed.

Or we could go in the opposite direction, and attempt to create AGI comparable in capability to the human mind on a physical substrate with little or no similarity to the human brain.

Either way, the main problem is simply that the system in question is so incredibly complex that we have yet to scratch the surface of its operation.

A much simpler example of the same problem is the genome. Our understanding of how DNA works has been complete for decades, yet the exact function of all the DNA in our genome remains the extremely incomplete subject of intensive study, because we're essentially attempting to reverse-engineer and fully document 700MB of the densest and most labyrinthine spaghetti code ever written.

I'm an atheist, but threadly reminder that every mainstream church in the entire world (Catholic, Orthodox, Anglican, every major Protestant church) was ambivalent to or soon endorsed evolution after Darwin, and all of them have retained that position to this day.

Religious opposition to evolution (much like young earth creationism) is an extremely recent phenomenon, dating back no further than 1923, when a burgerstani fundie cult called the Seventh-day Adventists began publishing pseudoscientific screeds, whose circulation eventually spread beyond burgerland via the Anglosphere by the 1930s, thence to other fundie sects around the world by the 1950s.

I don't know whether invoking some sort of membrane or any other interface between consciousness and observable phenomena is possible or necessary but I doubt both.

The problem in specifying an interface is, as mentioned above, the nature of the physical sciences itself. There are only three possible states that a scientific factoid is able to take (and only two of these have ever been encountered):

1) A statement about the world is "understood" by means of further, "deeper" statements about the nature of the physical world. E.g. copper is a good conductor of electricity *because* of the "deeper" fact that this element has a band structure (further caused by the Pauli exclusion principle) such that many electrons are forced above the Fermi level and may easily move from atom to atom. Note that this is recursive until...

2) A statement about the world is empirically verified, is not described by any existing theoretical framework, and a modification of an existing framework or the creation of a new framework is both required and expected. This is the cutting edge of science. A potentially absurd goal of progress in this direction is what we haven't yet seen:

3) A statement about the physical world that can be used to "explain" all other statements about the physical world and which itself needs no further explanation. The Theory Of Everything. This seems absurd because it would stand as a bare statement/assertion that is not understood in the same way as all other scientific statements. It falls prey to the problem of induction since it cannot be deduced which means that every statement deduced from it (all of science) also falls to the problem of induction.

We get either "turtles all the way down" or an impenetrable mystery. Attempting to square this with the problem of consciousness is... I'm not sure how one would even find a point to jump off of from one to another. Where does one even make a stand?

I suppose this implies dualism or idealism (inb4 somebody decides to mince words and bring more jargon) but that doesn't matter much to me. The only thing I believe to have learned is that a physical description of things doesn't suffice for consciousness. Sort of wrapping back around to my first post now so I'll stop.

You're missing the point. Take a piece of paper and write down a high-school physics version of Newton's law of gravitation. One side will contain a force and the other will contain some constants, a mass, and an acceleration. Every symbol in this equation can be measured or defined in terms of other measurements. Springs/rulers/stopwatches/etc can be used to determine whether an equation holds for a given experiment.

Now try to do the same thing for consciousness. Obviously you can't spell it all out. I mean that you probably can't even imagine what to begin writing opposite whichever physical state you choose for measurable brain description:

Phyiscal description of brain state = (What do you write here? What do you measure it with? In which units?)

For clarity: when I used the terms induction/deduction above it is not in the logical sense. In that sense all science is inductive. The terms are used relative to the scientific method itself. That is: inb4 nitpicking terms rather than arguing in good faith.

Molecular arrangement of brain at given points in time, reactions seen to occur
Chemical formulas
That's a very good question, for such a large, delicate, and complex biological device. Hopefully we'll be able to build such instruments.

But as I said, I think the most challenging quandary will be attempting to detect and understand patterns in the data we gather, rather than the task of gathering that data itself.

You're still not following. The left side of the equation in that example *is* the physical brain state. The positions and momenta and electrical charges etc etc of all of the particles involved. The right side is *consciousness*. Literal consciousness. What do you write for that?

Do you recognize an absolute objective distinction between the scribbles on a piece of paper and the subject an artist intended to depict? What about between the iron grains in an HDD and the data magnetically encoded on it? Both "sides" are merely patterns in physical objects of arbitrary significance to human observers, there is no inherent distinction between the two.

Re: first example: yes there is a distinction. The physical marks made by an artist can be measured (position, wavelengths of reflected photons, etc.) but the intention and interpretation are subjective (read: they pass through consciousness, whatever that might be).

Re: second example: irrelevant. The data contained on the disk is physically measurable whether defined in terms of positions of magnetic domains or defined in terms of binary data after being passed in serial underneath a drive head. Identical.

This is also tangential (at best) to my point. You're avoiding. The real challenge in "the hard problem of consciousness" is trying to figure out what the RHS of the equation I wrote above would look like. Will you take a shot at that? On one side there is a collection of symbols which describe the physical state of a brain and on the other side there is a set of symbols which represent consciousness itself. Impossible in my opinion. Brain state can be measured and described just as well as any collection of particles. Consciousness isn't a physical thing in that sense. It isn't a matter of complexity but a matter of type/category/something. To simply state "consciousness is physical" is to beg the question.

Goin to sleep now. Will check back tomorrow.

I think the hype over AI is slightly justified because this is the first time their algorithms have worked because of the vast amounts of information the internet offers and hardware has finally caught up. It is sort of like the first time you have sex but for AI researchers.

Eh, I'll try to pivot my argument a little.

Consciousness is not something produced by the brain, any more than gravity is something produced by an object's mass. Consciousness is LITERALLY the material arrangement of the brain, and the brain itself is literally consciousness.

Taking this perspective back to those examples, the drawing is material, the molecules and reactions in the artist's brain are material, and the subject (whether something abstract, or something representational like a tree or friend) is also material. The same with the HDD, extending beyond data stored on it, including its syntactic meaning, is wholly physical. There is nothing external, nothing "embodied" nor contained, these things simply EXIST, totally indistinct.

Your self, thoughts, hopes, dreams, every mote of your humanity are wholly described by the state of interlocking mechanical systems, evolving across spacetime, physically composing your brain, your body, the universe.

G'night user

Either I'm doing a poor job of explaining or you're not quite getting the point I'm trying to make. You claim that "Your self, thoughts, hopes, dreams, every mote of your humanity are wholly described by the state of interlocking mechanical systems" but are unable to create a mathematical bridge (the only type of bridge which exists in the hard sciences) between a physical description of a brain and the subjective experience of whichever person that brain belongs to.

The path you've started down with that statement has, in at least one paid academic's view, ended at the idea that consciousness doesn't exist at all: see the second paragraph of

Prof. Daniel Dennett has backed himself into a corner via this line of argumentation and now claims that consciousness doesn't exist. I believe the reason he has done this is that this tack allows him the flexibility to deny all counterarguments without doing much work.

I don't argue that brains do not exist as configurations of physically measurable things or that these brains are not necessary for and closely intertwined with consciousness. I argue that nobody has yet been able to suggest that there is a way to mathematically (symbolically) equate the physical state of a brain, which in theory is completely amenable to measurement, with the subjective experience of a person. This doesn't imply dualism any more than it implies a certain quality of matter which we can't yet measure (though if it turns out that in theory we aren't able to measure whatever this quality of matter is then it turns into dualism or idealism I suppose).

The entire point of the arguments I've put forth is that there is no way to set up two statements of different types such that we can call them equal to one another. A brain state can be perfectly mapped to physical, measurable quantities given advanced enough technology. The experience of a perception cannot be. For you to claim that these things are the same is to skip over the entire problem of mapping a physical configuration of matter to the nature of thought/perception.

For example: we can in theory map out every physical detail of a brain; the locations, momenta, electric charges, etc etc of every particle within a brain. We can do the same for a toaster or a potato or any other configuration of matter. What sort of theory allows us to map a given configuration of brain molecules/atoms/leptons+quarks/etc into the experience of one who experiences? This is the point of my attempt to get you to build an example equation:

(Physical configuration of brain state written down in mathematical notation as every current physical theory can be) = (what goes over here? how do you quantify a thought or an instance of conscious sense data as perceived by the receiver?)

To claim without an argument that everything is physical is to ignore what the precise definition of "physical" is. Physical stuff is measurable. The fact that physics and the other hard sciences have been so successful at manipulating the world is deceptive in a philosophical context as mentioned in my first post--nothing in the measurable physical sciences has anything (at least anything demonstrated thus far) in common with first-person experience. This doesn't mean that there isn't some sort of connection. It only means that the language which we use to describe our world scientifically (mathematics) isn't suited to make statements about subjective states of consciousness.

Can you provide a mathematical statement that encapsulates the conscious state of seeing a red bird against a blue sky? We can measure the things out there in the world (the bird and the sky) and we can measure the brain state of a person viewing these things, but we can't measure the subjective "scene" of a person experiencing it.

I admit I haven't read Dennett, but from what a brief search yields, that seems like a mischaracterization of his position. Consider the following: Assume for sake of argument the fate of the universe is wholly predetermined, robbing us of truly free will, but the Heisenberg Uncertainty Principle makes it impossible to take advantage of this fact. Is that different in any practical way from possessing free will in an unfated universe?

How can two systems be "connected", without them in actuality being subsystems of the same system? The flaw of dualism is that if we see and tinker with one "side", we must therefore be simultaneously interacting with the other "side" in the same ways through it. There are no one-way interactions (or at least there is no evidence to suggest such events, Occam's Razor).

A more obvious example of this fallacy is in a lot of modern "science fantasy" fiction, where "magic" and "physics" operate on "separate" sets of rules (or magic "has no rules"). To someone with philosophical grounding, that is of course a meaningless statement.

>the physical state of a brain, which in theory is completely amenable to measurement
This is the crux of the argument. Everything we're able to measure about the brain with each advancement has corresponded to conscious experience, so it logically follows that as we get closer to completely measuring the brain, we will get closer to measuring the experiences that brain feels. All that's needed is better instruments, and better mathematical models resulting from said instruments' data.

First I'll write: thanks for hanging around and sticking with this. Apologies for coming across as glib or condescending.
I've tried to make this point since my very first toast in the bread. If you're willing to indulge me go back and skim them all. No IDs but you'll probably be able to discern for yourself. First one is . Tangential yet important once we've figured out how to deal with the crux is .

I agree. My argument is that the nature of this correspondence is different than every correspondence in the hard sciences. This is the point I've tried to make by using the example of the two statements on either side of an equation. They are different in a conceptual sense but also the same in the sense that each side of the equation can be put in terms such that the "=" sign may be put between them and the equality can then be subjected to evidence and falsified or (in at least one instance) verified. [For details about falsification vs. verification just google "karl popper vienna circle" or something like that. The point here is that there was a framework, "logical positivism", which held that scientific truths could be "verified". Not the case and, surprisingly, this wasn't academically argued until the first half of the 20th century.]

My argument re: brain vs. consciousness is that, despite the implication in your own words which you might not be completely aware of, the two things cannot be called equal in the same sense that every other relationship in the hard sciences may be called equal. This is why I stuck on the point of the RHS of the thought-experiment equation I posted earlier. Imagine that the LHS is a perfect description of the physically measurable properties of the brain. What goes on the right side?

(All physical details of the brain specified in measurable quantities [think: springs, rulers, stopwatches, etc. The positions and momenta and charges and everything else we can measure!] live on this LHS of the equation.) = (What the fuck can we put here lol? This represents the conscious experience of a living person. I can't see a way to begin measuring anything that lives here--if it is some detail about the physical brain then it is already accounted for on the LHS)

This is where you're making the mistake. You (perhaps) inadvertently implicitly equate measuring the physical state of the brain with measuring the experience of the brain's owner. Measuring an entire brain perfectly has nothing to do with measuring what the brain "feels" other than incidentally. How do you physically measure the experience of a blue sky? How do you physically measure the experience of a person who has put their hand on a hot stove burner? The physical changes in the brain can certainly be measured... but if I set my hand on a stove the sensation that I (whatever I am) experience cannot be described in terms of the chemistry of cell death or the physical principles of thermodynamics.

It is probably clear by now but my argument has nothing to do with the subtlety or complexity of the brain or the physical world. The idea that I am trying to get across is that the measurable world and the world of consciousness are of a fundamentally different character. Again--it is possible to measure what happens to my hand, nervous system, and brain if I put my hand onto a hot stove top, but it is not possible to physically measure the sensation of pain in the units of position, momentum, charge, etc., etc., as all physical things are able to be measured in. There is a difference of type.

I guess I can also address:

This ties into my first reply pretty well.
I can't give you a list of all of the ways that systems can be connected. All I can offer is that if two systems are connected, in say... physics, then the connection can be mapped by a mathematical equation. I can't wrap my mind around every possible type of connection (even as a thought experiment) but I can wrap my mind around the types of connections that are meaningful in science. Not all scientific connections are instantly summarized by a mathematical equation but, in principle, they all can be from the first instant.

For example Michael Faraday was not trained in mathematics but he was very interested in electromagnetism. He obsessively sketched and described the phenomena he was able to generate with currents through wires and ferromagnets. James Clerk Maxwell was trained in mathematics and turned Faraday's observations into a mathematical framework (the nature of the physical interactions even inspired him to create a new sort of product in R^3, the "cross product". Nowadays mathematics has outpaced physics but this wasn't always the case.). This set of equations is now known as, creatively enough, "Maxwell's Equations". The point here is that, in this case, the mathematical relationships were not immediately known but that they were in principle discoverable. Everything in Faraday's drawings and descriptions could be measured and described in terms of physical entities (position/displacement, force, charge, etc.).

This isn't true of perceptions. I can't measure my field of vision. I can measure things *in* that field of vision, but the measuring apparatus, itself, then is part of my field of vision (or tangible in some way but I state it this way to make it clear).

Going back to my words:
This is the same thing you've alluded to by invoking "correlation". If there is a correlation yet we are not able to describe this correlation in terms of mathematics then I'm not sure where to go. It is some sort of connection/correlation but it can't be analyzed with the same tools we use to analyze every other question in the hard sciences.

transhuman theosis
soviet union

Over all a good response. My answer in short you've already said it:
that's it, and always will be to some degree in this creation. In the new creation we'll apparently have far broader capacities to fathom reality. As you seem to understand science is no do-all-be-all-end-all of understanding. As has been shown repeatedly since the early 1900's (notably Godel, et al) we simply are incapable of understanding many areas fully, and very fundamental ones such as logic and math themselves.

Python gets in the neck this time.

This appears to be about the quandary seemingly imposed by squaring your measurements of humans (such as yourself) with the experiences they (or you) relate verbally (or know), causing you to place it on a pedestal above those measurements.

The problem I've always had with such positions, is why the consciousness of others would be privileged at all, if it is in turn ingested by you exactly like those measurements? Because of this, dualism has always seemed to me a hobbled middle ground between solipsism and materialism. Once you've dismissed solipsism (a very simple matter, owing to its nihilistic pointlessness) and now believe your consciousness is similar to others', why would you still entertain the possibility that consciousness could be separate from the material world at all?

I might point to other scientific challenges, where rather than just lacking an appropriate conceptual language in which to describe the two sides of the "system" you're attempting to understand, you are physically cut off from measuring anything on the other "side" of the "equation". Black box reverse engineering, in other words.

The most trivial example would be attempting to cleanroom duplicate an encrypted piece of software, or crack into a remote computer system using an architecture you aren't familiar with. Probably the most daunting example would be fields such as astronomy and astrophysics, where we can only get a partial glimpse in our telescopes (even if indirectly, in the case of objects such as black holes and dark matter we can only guess the existence of through their effect on what can be seen). In all such cases, your only viable approach is to build the "other half of the equation" from scratch, by creating the simplest possible model that matches up with the input and/or output you actually can see. Whether or not your construct matches up with the actual thing you're observing isn't your priority, just that it matches up with your observations.

In the case of consciousness, that "simplest possible explanation" has so far been materialism.

logic, philosophy, and linguistics.

Minsky, pls. Aren't you dead or something?
youtube.com/watch?v=P5iqYuFmzqg

Get a gaming rig, learn Python/Numpy/SciPy/Conda, learn Keras.
It is that easy (((newfags))), so go back to >>>/machinecult/

This -> is the answer you are looking for, OP.
Also, learn lisp and prolog (especially if you want to do AI stuff the "old way". Nowadays most soydevs just use TensorFlow) But see : the current AI/machine learning hype is just a big meme


But does anyone even use it?

Chrome-based browsers including "Opera" and Brave. Mozilla won't add it to Firefuk'd even though Palemoon has. Apple & M$ of course are in a perpetual 3-way slapfight with Google.

youtube.com/channel/UCWN3xxRkmTPmbKwht9FuE5A