Where will artificial general intelligence come from? • Hacker News Board

Re: Where will artificial general intelligence come from?

#131

Post by viewtransform » Sat, Sep 02, 2017, 8:34 PM UTC

mythrwy wrote: post #67

rdlecler1 wrote: post #29

Background: I did AI and Philosophy of Mind in undergrad, an MSc focused on ALife, then a PhD at Yale under a Macarthur Fellow who developed the theoretical framework for the 'evolution of Evolvability' where I worked on computational evolutionary biology. I can say we're not going to blindly brute force our way forward, but instead we'll need to reverse engineer nature's core algorithms to generate hard AI. Every ti…

Background: Some guy on the Internet with an unrelated Bachelors degree he didn't study that hard for and has picked up a little Python and Javascript somewhere along the way. I know of no biological forms that have evolved wheels. But wheels have turned out to be a hell of a lot faster than fins or legs. I see no reason "intelligence" has to follow the pathways or limitations of biology or neurology at all. Although…

Not a wheel but a rotary motor: topologically the same. https://en.wikipedia.org/wiki/Flagellum

Re: Where will artificial general intelligence come from?

#132

Post by orthoganol » Sat, Sep 02, 2017, 9:01 PM UTC

nopinsight wrote: post #28

indifferentalex wrote: post #2

Not on their radar, or their slides at least: Natural Language Processing based rule-based brute-force artificial intelligence (that could be augmented through sensors/motors that allow interaction with the external world). A Vulcan-like (Star Trek) AI, what do you think? Might be easier to simulate the entire brain, on the other hand it might be doable and bridge the gap to general AI.

Rule-based NLP has been tried for several decades and has (very) limited success in the real world. Current systems based on deep learning beat it for most complex tasks. DeepL, which was on HN front page a few days, is the latest example: https://news.ycombinator.com/item?id=15122764

You're going to have to elaborate on complex tasks. I would argue the majority of successful, money generating software based in NLP/ NLU, i.e. the majority of the industry, is "rule based" (used in a general sense to mean non DL). Personal assistants, search, chatbots, etc.

Re: Where will artificial general intelligence come from?

#133

Post by viewtransform » Sat, Sep 02, 2017, 9:05 PM UTC

NumberCruncher wrote: post #40

If you take a look at the evolution of the most advanced non artificial general intelligence, eg. human intelligence, it is strongly connected to the evolution of communication. It is a question of efficiency whether you learn through your own experience and failures or through the experience and failures of others. This teaching/learning process was boosted through the use of pictures, spoken language, hand written…

I would argue that intelligence is connected to the evolution of sensing at a distance. Vision in particular, allowed life to evaluate the state of the environment at a distance and allowed for the evolution of strategies to predict and respond to the environment in real time. The progression in intelligence from sponge to amphibian to mammal is related to the evolution of finer sensing of the environment at a distance: vision, smell, sound etc.

Re: Where will artificial general intelligence come from?

#134

Post by PeterisP » Sat, Sep 02, 2017, 9:16 PM UTC

jamiek88 wrote: post #94

skummetmaelk wrote: post #90

Earlier quoted context omitted.

Wheels are useless if you don't have roads. EDIT: Please explain your downvotes here? Don't you people see that this, coupled with there being no useful intermediate stage before developing a full wheel, prevents evolution from "inventing" it.

Well that's not true at all. See: Wheelbarrow; amongst many wheeled devices I can think of immediately.

Wheelbarrows are a prime counterexample - historically, in environments that had decent man-made trails, wheelbarrows were useful and used; in environments that did not, pack-carrying animals or even people did the job even though the concept of wheelbarrows was known; also, the availability of roads (as opposed to narrow trails) was the prime reason for differences of usage of wheelbarrows (especially the efficient middle-wheel variety e.g. https://www.google.lv/search?q=wheelbarrow+chinese&tbm=isch ) versus two or four wheel animal (or man) pulled carts, which were popular only where roads had been made and were maintainable, so not in jungle-like environments where it takes more work to keep a road from overgrowing.

Re: Where will artificial general intelligence come from?

#135

Post by viewtransform » Sat, Sep 02, 2017, 9:28 PM UTC

bra-ket wrote: post #125

Artificial intelligence will come from understanding natural intelligence.

We generally accept that we 'know' something when the model used to explain the system is simpler than the system itself.

The brain is a very high-dimensional non-linear dynamical system. The number of neurons in the brain is on the order of the number of trees in the amazon rain forest and the number of synapses the number of leaves on those trees. https://youtu.be/8FHBh_OmdsM?t=1165

We do not have the mathematical tools to understand such systems in general. What if reductionism doesn't work and the best model of natural intelligence is as complicated as the system itself ? Can we say we understand natural intelligence?

It could be the case in a distance future that we evolve an artifical intelligence purely as a computation that is capable of understanding us but not us them.

Re: Where will artificial general intelligence come from?

#136

Post by korzun » Sat, Sep 02, 2017, 9:33 PM UTC

mythrwy wrote: post #67

rdlecler1 wrote: post #29

Background: I did AI and Philosophy of Mind in undergrad, an MSc focused on ALife, then a PhD at Yale under a Macarthur Fellow who developed the theoretical framework for the 'evolution of Evolvability' where I worked on computational evolutionary biology. I can say we're not going to blindly brute force our way forward, but instead we'll need to reverse engineer nature's core algorithms to generate hard AI. Every ti…

Background: Some guy on the Internet with an unrelated Bachelors degree he didn't study that hard for and has picked up a little Python and Javascript somewhere along the way. I know of no biological forms that have evolved wheels. But wheels have turned out to be a hell of a lot faster than fins or legs. I see no reason "intelligence" has to follow the pathways or limitations of biology or neurology at all. Although…

The wheel, by itself, is not faster than legs. It does not adapt, repair, or feel. You will always need more than your legs to create a system that properly utilizes wheels.

There is nothing intelligent about artificial intelligence. You might as well call it artificial stupidity, but that doesn't make you sound sophisticated.

Memorizing the results of a pre-defined concept faster due to a lower barrier of entry does not make you smarter or more intelligent. You can process all of the traffic signals in the world in a blink of an eye only to get stomped by a dancing traffic director.

You can't have AI without a human, and you can't have AI with a human.

Re: Where will artificial general intelligence come from?

#137

Post by viewtransform » Sat, Sep 02, 2017, 9:37 PM UTC

psadri wrote: post #78

Do we consider simpler brains to exhibit general intelligence (e.g. A crow's). Is it a more tractable problem to replicate crow level AI first before tackling humans?

This bottom up approach was investigated in the late 90's after researchers became skeptical that the symbolic AI approaches in vogue then would achieve general AI.

You can look for the topic of Artifical Life, Emergent Computation, Artifical Evolution. See the Santa Fe Institute conferences on ALife. There were some promising initial results but the field never managed to take off.

Re: Where will artificial general intelligence come from?

#138

Post by tokipin » Sat, Sep 02, 2017, 9:45 PM UTC

CuriouslyC wrote: post #119

tokipin wrote: post #101

Earlier quoted context omitted.

> Artificial general intelligence won't be invented, it will emerge, just general intelligence did in the wild. I think so as well, but for a specific reason: That people are not as "AGI" as we think, and that our general intelligence is the result of immense neuron-computing resources committed to satisfying a few simple drives. In other words, narrow AI with a lot of resources. Not to mention many specialized subco…

Why do you think consciousness is a selected trait?

Well I don't know what consciousness is, so I couldn't answer that. But I think the opposite stance of consciousness being some kind of majestic oracle is silly. There is a reason why we perceive consciousness, even if it is just as a side effect of something else or a random fluke that is useless or an illusion for coordinating our limbs or whatever.

These types of discussions do get pretty wonky. When I say consciousness is selected for, I mean more whatever practical apparatus (if any) it is emergent from, not the perception of consciousness itself.

Re: Where will artificial general intelligence come from?

#139

Post by karpathy » Sat, Sep 02, 2017, 9:46 PM UTC

I gave this talk a while ago to a small group of attendees. It was not recorded (I saw some ask below). It's based on a document I wrote a while ago called "You suck at writing AI" (never published). The basic argument was that people are comically inadequate at writing complex code. You can't write the code to detect a cat in an image and the correct thing to do is to give up, write down an objective that measures the desiderata and pay with compute to search a function space for solutions. In the same vain, the idea of writing an AGI and all of its cognitive machinery is preposterous and the correct thing to do is to give up, think about the objective and search the program space for solutions. Unfortunately, the mindset of decomposition by function (see Brooks ref), which has worked so well for us in so many areas of scientific inquiry, is just about the most misleading mindset when it comes to AGI.

Re: Where will artificial general intelligence come from?

#140

Post by scottlocklin » Sat, Sep 02, 2017, 9:46 PM UTC

Speculating on where AGI will come from is sort of like speculating where Faster than Light travel will come from. Except FTL has some vaguely plausible physics behind it, and AGI-wise, we really have no idea what the "I" in AGI means.

The mere fact that biological neural networks are rate encoded might turn out to be the one crucial thing that's practically impossible to simulate in a VN computer.

My vote: "we have no idea; probably not in my lifetime."