Dronez

[1338 h4x0r]

http://news.bbc.co.uk/2/hi/south_asia/8115814.stm

So, apparently a formation of Predator drones flew in and launched missiles at the funeral of a deceased Taliban member. At least 43 people died, of whom a number were alleged militants.

Heavy-handed? Morally ambiguous?

There's probably more where that came from:

http://www.darpa.mil/darpatech2005/presentations/jucas/francis.pdf

If you don't want to read the paper, which is about progress in autonomous combat UAVs, as distinguished from drones like the Predator which are glorified RC planes, you can just watch the video:

http://www.youtube.com/watch?v=cA8EQ3dyB8c

The X-45, which has been discontinued in favor of the X-47, was tested on a number of functions. It was able to fly, land, fix its own malfunctions without the assistance or even the knowledge of ground crews and even deployed a test weapon with nominal approval from a person. (As you can see, he basically only clicked an "OK" button.)

Deployment of autonomous UAVs in combat is probably going to happen within the next ten years, which raises a lot of interesting issues now. According to the Geneva Convention, an operator must be "kept in the loop" if weapons are to be deployed from these autonomous platforms, to make sure someone is "responsible" for the decision to deploy.

So, even if it is technically possible to run a whole UAV mission from takeoff to landing, including weapons deployment, with no intervention from a ground crew, the military will probably let someone rubber-stamp the UAV's decisions in order to comply with the treaty. But eventually, many ground vehicles will become autonomous as well* and, unlike aircraft, they may have to make snap judgments on when to use weaponry. So, if something goes awry in the case that no person approves the deployment of weapons, who is responsible?

(An even more interesting question of course is whether the apathetic public will care, or even know.)

* - The National Defense Authorization Act for Fiscal Year 2001, Public Law 106-398, Section 220 states that “It shall be a goal of the Armed Forces to achieve the fielding of unmanned, remotely controlled technology such that...by 2015, one-third of the operational ground combat vehicles are unmanned.”

[Aaron]

These aren't the droids you're looking for.

[1338 h4x0r]

Die Mensch Machine
Halb Wesen und halb Ding
Die Mensch Machine
Halb Wesen und halb über Ding

[mothergoose]

Pretty crazy... it's strange how this is becoming a forgotten war already within the US.

[Smokey D]

So, if something goes awry in the case that no person approves the deployment of weapons, who is responsible?

You could look at the question of liability as one related to the rule in Rylands v Fletcher, namely that there will be strict liability for failure to control a dangerous thing even if there was, technically speaking, no fault. Presumably the person to whom liability would attach would be the individual/organisation which should have been in control even if in actual fact they weren't.

[1338 h4x0r]

Pretty crazy... it's strange how this is becoming a forgotten war already within the US.

Yes, that's right, we're thinking hard enough as it is, shhhh, everything is going to be alright, don't worry about anything you're doing or anything that's happening around you, shhhh, just keep watching TV, everything is going to be alright ......

You could look at the question of liability as one related to the rule in Rylands v Fletcher, namely that there will be strict liability for failure to control a dangerous thing even if there was, technically speaking, no fault. Presumably the person to whom liability would attach would be the individual/organisation which should have been in control even if in actual fact they weren't.

I looked at the details in Rylands v Fletcher and don't see how an analogy holds between this case and that. If an unmanned vehicle shoots the wrong person, do the creators get blamed? Who among the creators? (Presumably there'd be teams for vision, reasoning, pattern recognition, etc. ...... all components working together to make a final decision.) Or does the military itself get blamed? Or is there no clear liability?

[Smokey D]

Rylands v Fletcher is specifically related to emanations from land, so obviously it is not binding. My suggestion we could use Rylands by way of analogy to explore and understand one way in which liability could attach.

Rylands says that the person who creates the danger should control it, and is strictly liable for any default. In this situation, we would say that the people who were meant to be in control of the drone are strictly liable for any damage it causes even if in actual fact they weren't in control of it. So, if an unmanned vehicle shoots the wrong person we would blame the person who is meant to be in charge of the vehicle just as we would blame the land owner who brought dangerous/unnatural materials on to his land which subsequently escaped. That is, liability would lie with the putative operators, not with the creators of the drone.

Presumably ordinary negligence liability could also be relevant but Rylands means there will be liability even in the absence of negligence (or other form of fault), which seems to be an appropriate way to induce people in charge of especially dangerous things to act with special care.

In a civil suit, the doctrine of respondeat superior (vicarious liability) would apply , meaning the military itself or at least the controlling branch would be liable.

I don't think it would be appropriate to blame the creator/manufacturer, unless some defect could be discovered in the manufacturing process which led to the default, in which case ordinary manufacturer negligence liability would probably attach.

[1338 h4x0r]

What about the drone itself

[Smokey D]

Um, probably not. I doubt any court would recognise a drone as capable of decision making in a way that would attract liability.

[1338 h4x0r]

If they have memories, at least, they can presumably be differentiated in terms of decision making ability

[Smokey D]

Possibly although no court would be willing at this early stage to recognise a machine as having free will capable of attracting liability. At the most, I can see an analogy being drawn between the drone and a wild animal.

[1338 h4x0r]

lol, so we'd be letting wild animals loose?

[Aaron]

Animals aren't tried in court, their owners are.

[1338 h4x0r]

Yes but the analogy is kind of silly since we don't let wild or trained animals loose on the battlefield, that's nuts

[Smokey D]

lol, so we'd be letting wild animals loose?

Well, it's an analogy not an identity, but there are good grounds for the comparison.

Yes but the analogy is kind of silly since we don't let wild or trained animals loose on the battlefield, that's nuts

I thought we were talking about situations where the drone took action in non-battlefield situations.

[Aaron]

You're acting like a drone is any different from another piece of machinery that malfunctions; if my work's printer burns me then my work is liable. Same situation.

[1338 h4x0r]

Well, it's an analogy not an identity, but there are good grounds for the comparison.

Kind of

I thought we were talking about situations where the drone took action in non-battlefield situations.

No

Or, at least, I wasn't

That's why I emphasized the combat role they will have

You're acting like a drone is any different from another piece of machinery that malfunctions; if my work's printer burns me then my work is liable. Same situation.

Printers don't make lots of decisions

[Aaron]

Printers don't make lots of decisions
They make decisions within the constraints of their programming, as do drones.

[1338 h4x0r]

As does everything else, derp

[Aaron]

Nope. Humans think, machines respond.

[1338 h4x0r]

Most AI and neuroscience experts would beg to differ

There is no essential difference between artificial and natural reasoning, only a difference of degree

[Aaron]

...and most legislators would disagree with them. You can't hold a tool accountable for doing something that it has been designed to do.

[1338 h4x0r]

...and most legislators would disagree with them.

lol

[Aaron]

What's lol about that? They're the ones actually doing things in society, unlike the academics.

[1338 h4x0r]

What's lol about that? They're the ones actually doing things in society, unlike the academics.

Come now, don't go all DBoons Ghost on us

What they do in society isn't relevant to the validity of their opinions on what constitutes thinking

[Aaron]

Sure, but academic discussion and opinion doesn't change legislation of society's perception of events and issues. They're just a conduit for reflecting the results of research, they don't actually apply or change ideals.

[1338 h4x0r]

Yes they do

Just slowly

Albeit powerfully

[Smokey D]

Kind of

No. Both are potentially dangerous things which have the capacity to operate more or less autonomously.



No

Or, at least, I wasn't

That's why I emphasized the combat role they will have

So why do you care about liability or whatever?


Most AI and neuroscience experts would beg to differ

There is no essential difference between artificial and natural reasoning, only a difference of degree

Debatable.

I'm not quite sure where Aaron's going with this but he's right in that courts and policy makers are not going to recognise machines as anything other than tools any time soon. Even if the difference between artificial and biological reasoning is only one of degree, the fact remains that the degree is vast.

[1338 h4x0r]

No. Both are potentially dangerous things which have the capacity to operate more or less autonomously.

Ok.


So why do you care about liability or whatever?


Just asking


Debatable.


Debated mainly by people who don't know what they're talking about.

I'm not quite sure where Aaron's going with this but he's right in that courts and policy makers are not going to recognise machines as anything other than tools any time soon. Even if the difference between artificial and biological reasoning is only one of degree, the fact remains that the degree is vast.

I think the graph in this picture says it best:

http://i40.tinypic.com/23jmoaf.jpg

Our opinion of "tools" is going to change dramatically

[Aaron]

Debated mainly by people who don't know what they're talking about.
You do know that I work in risk management and deal with liability distinctions all the time right?

I'm not quite sure where Aaron's going with this but he's right in that courts and policy makers are not going to recognise machines as anything other than tools any time soon. Even if the difference between artificial and biological reasoning is only one of degree, the fact remains that the degree is vast.

Yep, that's what I was alluding to.

[Smokey D]

Debated mainly by people who don't know what they're talking about.

Maybe. I dunno.


I think the graph in this picture says it best:

http://i40.tinypic.com/23jmoaf.jpg

Our opinion of "tools" is going to change dramatically

Maybe. Personally, I think you just got a hard on for the idea of AI.

The point remains the same. No machine currently in existence will be treated by policy makers as anything other than a tool.

[1338 h4x0r]

You do know that I work in risk management and deal with liability distinctions all the time right?

Uh yeah but how does that override scientific opinion

Maybe. Personally, I think you just got a hard on for the idea of AI.

So does the US gubbermint

What about augmented soldiers? Aren't they going to change how we see ourselves wrt technology?

[Aaron]

Uh yeah but how does that override scientific opinion
Because I understand how legislation surrounding these issues works. Whether the scientific community thinks a robot can think is irrelevant outside of their discussions if legislation in place deems them to be machinery, etc etc.

[Smokey D]

So does the US gubbermint

What about augmented soldiers? Aren't they going to change how we see ourselves wrt technology?

Probably. I don't really care. I don't even know if they're feasible or the fantasy of sci-fi nerds.

[1338 h4x0r]

There'll probably be laws about it eventually, Luke, here's a good start, very readable. You can get a head start:

http://wtec.org/ConvergingTechnologies/1/NBIC_report.pdf

Pages 97~274 (sections Expanding Human Cognition and Communication and Improving Human Health and Physical Capabilities) are of most relevance

[Smokey D]

Yeah there probably will be laws. So what. I don't care if robots get as smart as humans. It's certainly not relevant to whether the robots we have now ought to be treated like humans.

[Aaron]

But Luke, The matrix could happen.

[1338 h4x0r]

The resemblance of current research to science fiction doesn't mean anything. It's not like science fiction hasn't become true before.

Yeah there probably will be laws. So what. I don't care if robots get as smart as humans. It's certainly not relevant to whether the robots we have now ought to be treated like humans.

Well this is talking about modifying persons

[Smokey D]

Well this is talking about modifying persons

And?

[Aaron]

Wearing shoes is modifying a person.

[Mr. Ron]

Laws on what constitutes "human rights" will be interesting since the next logical procession in human evolution is to integrate some sort of machinery into our bodies, imo.

[1338 h4x0r]

Luke, that will ostensibly change our attitude towards what qualifies as a person

EDIT - yes, Mr. Ron gets it

[Aaron]

Isn't a person, you know, a body, and stuff in them is stuff in a body?
A contact lense doesn't have rights now does it?

[1338 h4x0r]

Contact lenses aren't the same as modifying the functions of a person's brain substantially

"DARPA has a brain-machine interface program about to start. This program seeks human ability to control complex entities by sending control actions without the delay for muscle activation."

(This is a little different)

[Smokey D]

I really don't see what any of this has to do with a drone as used by the US military. Yeah it's potentially true that we're going to enter some weird cyborg age but that there will be laws on this and that our understandings of what it means to be human might change aren't exactly earth shattering conclusions.

[1338 h4x0r]

Well what do you think will happen

[Aaron]

Changing someone's perception of their world around them, ie their sight, is the same as changing their brain in result.

[1338 h4x0r]

Yes but it's not very invasive is it

Hell even heart replacement is less of a threat to vulgar views of personhood than direct brain modification or interfacing

[Aaron]

How isn't it?

[1338 h4x0r]

It's on the outside of the brain, not on the inside

[Aaron]

Presenting information to someone differently is the same as changing how they interpret it in so far as it's invasive.

[Smokey D]

Well what do you think will happen

I'm not really interested in speculating. This conversation was a lot more interesting when it was about who would be liable for semi-autonomous drones randomly killing people.

[1338 h4x0r]

@ Aaron

Not really. The neural infrastructure for vision processing doesn't change because you have glasses or contacts

@ Luke

Then keep talking about that

[Aaron]

@ Aaron

Not really. The neural infrastructure for vision processing doesn't change because you have glasses or contacts
No but you're changing the information recieved [ie more accurate information].

[Meatplow]

Yes, that's right, we're thinking hard enough as it is, shhhh, everything is going to be alright, don't worry about anything you're doing or anything that's happening around you, shhhh, just keep watching TV, everything is going to be alright ...... ?

don't forget the beer

[Gattsu347]

The robots have arrived.

[1338 h4x0r]

No but you're changing the information recieved [ie more accurate information].

Alright, in the views of, say, legislators, which is more likely to challenge current laws and views of personhood: contact lenses or brain modification and interfacing?

[Meatplow]

robots

[1338 h4x0r]

Wir sind die Roboter

[Aaron]

Skynet is operative!

[1338 h4x0r]

没有 ...... 我在战后长大 ...... 在废墟中成长

"No ...... I grew up after [the war] ...... in the ruins"

I'm re-posting that video to tudou, hope it gets a lot of views

[Aaron]

Haha. This is more fun than usual cause the threadstarter [you] can take a joke.

We don't know who struck first, but it was us that scorched the sky.

[1338 h4x0r]

Of course that's cuz all my threads are basically practical jokes mixed in with some hard facts

[mph4ever]

sure there will be some merging of technology with humans. i'm sure today it is possible to remove the brain from a body, change the nerve connections to something delivering a similar electrical signal and make limbs move and hearts beat and whatever. pacemakers do this type of thing today. but the machine will never replace the mind, it will never be conscious

and with regard to invasive machinery and the brain. augmented reality will deliver all our needs in the future and the debate will be about the morals of merging man and machine. this debate will fall in favour of man every time. reason is that the software and hardware running the proposed machines will be managed and maintained by a corporation and people will want some indepedence from intel and google

lol would you like linux, android or windows on that part of your brain.

error: XXX101001110101 can't get erection at right time. fatal error. dick runtime routine failed due to corrupt diskpsace, have you got the driver disk or would you like us to search the internets for the latest version,

run routine audio output "tell bitch to wait, we'll be online in t mnus 3 minutes"

1338, smokey's right, stop getting alll horned about ai and sci fi and go shoot your load to some noodz man

we are not borg, and we never will be

[1338 h4x0r]

sure there will be some merging of technology with humans. i'm sure today it is possible to remove the brain from a body, change the nerve connections to something delivering a similar electrical signal and make limbs move and hearts beat and whatever. pacemakers do this type of thing today. but the machine will never replace the mind, it will never be conscious

As an atheist, what kind of (non-metaphysical) appeal can you make to back up this distinction

And do you realize that almost no person genuinely qualified on the brain agrees with you

I mean there's a minority in the AI/neuroscience debate with people like John Searle and Roger Penrose but their arguments are widely considered to be elaborate appeals to absurdity

EDIT - and analogies to the BSoD in Windows are just stupid, you can't compare crappy commercial software running on a von Neumann architecture to purpose-built neuromorphic hardware that doesn't even have an operating system ...... that's a horse of a different color

[mph4ever]

As an atheist, what kind of (non-metaphysical) appeal can you make to back up this distinction

which distinction? the one between humans supplementing their bodies with machines that replace brains?

And do you realize that almost no person genuinely qualified on the brain agrees with you
of course they don't. if they did then they would be all out a job now, wouldn't they. don't you get reseach? blue skies baby. give some money to prove some stuff that i fantasise about. i'll publish some research backing up what i knew we could prove at the outset and then come back for more and do the same.

i'm under non disclosure with a number of university based research organisations. we are looking to monetise their output, particulary in the area of geo spatial and augmented reality. they start out with a "wouldn't it be great if we could" but never think about "what practical use would we put it to"

[1338 h4x0r]

which distinction? the one between humans supplementing their bodies with machines that replace brains?

Sure

of course they don't. if they did then they would be all out a job now, wouldn't they. don't you get reseach?

Come on ...... this argument is just ****ing stupid, and I'm not going to entertain it.

You're dodging the point, which is this: what is the essential difference between thoughts made of electrical impulses and thoughts made of electrical impulses

Oh yeah

There is none

BTW: does this look like "castle in the sky" **** to you?

http://www.tudou.com/programs/view/V1N6dcDDJGY/

[1338 h4x0r]

I wouldn't worry unless you live in Waziristan

At least not for the time being

EDIT - wow moderator decision cut a lot of dumb peanut gallery **** out of my thread and actually seems favorable this time

[Light Flantastic]

the whole thread would have been a better idea

i mean, i think robots will take over soon

[1338 h4x0r]

Yeah you know the ENIAC is useful for calculating trajectories and stuff but it doesn't have real thoughts, and besides it takes up a whole room and is breaking all the time

......fast forward to the present day

Even the most outlandish cybernetic revolt scenario is more likely than anarchism and everything being about "property rights" (which you blather about constantly by the way), and a number are entirely plausible. You'd see that if you had any ability to extrapolate from the past and present.

[Smokey D]

Dude, quit drooling. It's all well and good to have a discussion about it but when you talk about it like it's guaranteed it gets really annoying.

Given that we really don't know much about the relationship between the brain and personhood, it's far too premature to be talking about sentient or semi-sentient machines let alone computer persons.

[Light Flantastic]

Even the most outlandish cybernetic revolt scenario is more likely than anarchism and everything being about "property rights" (which you blather about constantly by the way), and a number are entirely plausible. You'd see that if you had any ability to extrapolate from the past and present.

i can tell you're mad because you introduced an unrelated topic

!

[1338 h4x0r]

Given that we really don't know much about the relationship between the brain and personhood, it's far too premature to be talking about sentient or semi-sentient machines let alone computer persons.

I think Hans Moravec's argument (which is uncannily similar to mathematical induction) is good.

If one neuron in a person's head died, and you replaced it with an equivalent component, you wouldn't notice a difference between the old person and the new.

What if the person got brain cancer and you kept doing this?

Presumably there wouldn't be any difference.

[Smokey D]

It's an interesting argument, but it depends on the idea that the difference is really only one of degree. Surely we cannot know this at this stage.

Also, it assumes its own conclusion by suggesting that we could fashion a component which could perfectly replace a neuron.

[1338 h4x0r]

Kind of like how the jury is still out on climate change

I'm just articulating mainstream views, not sure if you knew

[Light Flantastic]

the 'argument' you presented mentions nothing of our ability to do it, its just a philosophical musing on what makes humans human

[Smokey D]

Kind of like how the jury is still out on climate change

I'm just articulating mainstream views, not sure if you knew

I'm no neuroscientist, but from what Amit used to say, I struggle with the idea that neuology has reached a level that we can conclusively what constitutes a person.

[1338 h4x0r]

the 'argument' you presented mentions nothing of our ability to do it, its just a philosophical musing on what makes humans human

Analogous to "nonconstructive proof"

Derp

I'm no neuroscientist, but from what Amit used to say, I struggle with the idea that neuology has reached a level that we can conclusively what constitutes a person.

No but the state of the art is like a half-assembled jigsaw puzzle and you can pretty much tell that the message on the puzzle is: "It's just signal processing n shi"

[Smokey D]

I think you're overstating the where the science is at.

[1338 h4x0r]

I'm being a little glib in describing the state of the art, but I'm really not far from the mark actually

[Light Flantastic]

Analogous to "nonconstructive proof"the question was of the certainty of it happening, which is wholly dependent on our ability to do it

it wasnt really a proof of anything anyway, it was an opinion on what makes a human

[1338 h4x0r]

That's why I said "analogous"

For someone who goes around making outlandish arguments by analogy pretty regularly, you sure do a poor job at recognizing analogies

And why wouldn't we be able to build an artilect

People raise a ton of dumb objections to everything in AI and then keep raising dumb objections until the thing they were dumbly objecting to actually happens

Then they start raising new dumb objections

[Light Flantastic]

That's why I said "analogous"

its still not applicable to the question of certainty

[1338 h4x0r]

The eventual creation of artilects is virtually certain

Every naysayer I've come across hasn't read a Goddamn thing about current research and applications

I'm not willing to bet you're an exception

[Smokey D]

And why wouldn't we be able to build an artilect

The point is that the argument assumes that there is such a thing as an artificial equivalent to a neuron, and, if you're trying to show that artificial personhood is inevitable, depends on that for its conclusion ie it begs the question. The argument, as Chad point out, is rightly deployed not to show that artificial intelligence will happen but that if it does happen then there will be no difference between humans and artilects.

[Light Flantastic]

The eventual creation of artilects is virtually certain
an impossible statement to prove

[1338 h4x0r]

The point is that the argument assumes that there is such a thing as an artificial equivalent to a neuron, and, if you're trying to show that artificial personhood is inevitable, depends on that for its conclusion ie it begs the question. The argument, as Chad point out, is rightly deployed not to show that artificial intelligence will happen but that if it does happen then there will be no difference between humans and artilects.

Yes but there's really no argument in the relevant sections of the academic community about whether artificial neurons are equivalent to vanilla neurons

To raise that question would be kind of like asking whether the Universe is really billions of years old and would probably provoke equally disdainful reactions in some people

Didn't notice I was begging the question in your view: I thought that premise would be patently obvious

an impossible statement to prove

Proof is for alcohol and mathematics

[Smokey D]

There's no point in any of us trying to prove a negative, but I'm skeptical of the claim about where the science is at.

[1338 h4x0r]

Why

I have a broham in neuroscience I can ask about this, I should bring the issue up when I see him

[Smokey D]

Because neuroscience and ai research are incredibly broad fields with a huge array of opinion in them, and I don't think there is anything like the level of consensus you are suggesting.

As far as I'm aware, you haven't even defined AI, although I'm assuming you're talking about it as appoximating human intelligence (strong AI?).

[1338 h4x0r]

You've really missed the boat, then, Luke

(look up "cognitive revolution")

[Smokey D]

Um, so you're saying there isn't a huge diversity in opinion in AI research?

[1338 h4x0r]

There are disagreements about the ethics, few disagreements about the ontology

In fact you'd probably be scoffed at if you suggested that an artificial person could not exist to many people in AI or neuroscience. I know because I learned the hard way

[Smokey D]

Yeah I bet there are heaps of people who think it's inevitable. The point is not that there isn't a large group of researchers who fervently believe in artificial personhood but that they do not comprise the entirety of scholars working in the field, and, from what I gather, do not even comprise a majority, not least because there is considerable disagreement about what personhood even means.

[1338 h4x0r]

Like, who else are we talking about here

There tend to be philosophical concepts shared in common by these researchers (like materialism, empiricism, etc.)

[GorgeousGabe]

I'll be laughing at you ****ers when it becomes an issue and you look like an idiot because you decided to say 'prove it can happen' on an internet forum because you were that stubborn about proving some point that you have no educated opinion about.

AI will come and it will become an issue in the near future

I'm disgusted by unmanned combat vehicles and I think they should all be destroyed

[Light Flantastic]

ill be laughing at you for denying god when he smites you thats what you get for scoffing at him and asking for proof


no what

[1338 h4x0r]

I'm disgusted by unmanned combat vehicles and I think they should all be destroyed

Yeah personally the reason I bring them up, again, is not because I want us to get wiped out by robots (that's shtick), but in the hope that a few people, maybe not many, but a few, will be shocked out of their complacency

What's most likely is some synthesis of man and machine. If "takeover" happens it will be gradual and we'll barely notice.

Regardless, you have to admit that bombing run was pretty cool. The ground crew only hit the "OK" button.

[GorgeousGabe]

Sure it was pretty cool. I mean I know you're with me when I say that a I just can't wait!!!! for these ****ers to be on every street corner, protectin american lives here on home soil!

[1338 h4x0r]

That'd make an interesting science fiction idea

Homeland Security patrol cyborgs are given the directive to round-up potential terrorists

Accordingly, they put most of the Republican base in internment camps

[Mr. Ron]

(and minorities too)

[1338 h4x0r]

Me, I'm keeping my eye on those Spaniards

2 / 15 / 1898

NEVAR FORGET

[Mr. Ron]

lol

[Smokey D]

Like, who else are we talking about here

There tend to be philosophical concepts shared in common by these researchers (like materialism, empiricism, etc.)

Like, for example, the Chinese Room.

It's hard to carry on this conversation when you haven't really defined AI. Are we talking about strong AI or any sort of intelligence?

Also, I see no problem with automated or remote killing machines. At least, no more problem than I see with human operated ones.

[1338 h4x0r]

Like, for example, the Chinese Room

That's what I'm talking about: Searle's expertise isn't even in AI or neuroscience and his argument is a long appeal to absurdity. That's why he's not entertained outside of undergraduate philosophy texts anymore.

It's hard to carry on this conversation when you haven't really defined AI. Are we talking about strong AI or any sort of intelligence?

Strong AI - is a mind

[Smokey D]

That's what I'm talking about: Searle's expertise isn't even in AI or neuroscience and his argument is a long appeal to absurdity. That's why he's not entertained outside of undergraduate philosophy texts anymore.


Explain why it is an appeal to absurdity.

But either way, he is only an example. Other obvious problems include the diverse array of abilities we would need to engineer (parts of mind), potentially problematic metaphysical assumptions about the nature of information, mind and personhood etc. I don't mean to suggest I've done anything remotely approaching a review of the literature, or that I know the ins and outs of AI research. I'm not saying that strong AI is impossible -- although I still think Searle provides a pretty powerful critique -- I'm merely questioning the automaticity of it.

Strong AI - is a mind

Well, that's part of the problem. We don't really even know what a mind is. I mean, we can probably recognise a mind when we see one but I doubt we have identified all the various parts of mind let alone worked out how to replicate them.

[GorgeousGabe]

Who says we need to? We just need to make computers that can think on their own. They don't necessarily need to think like us

Though I do think they'll try to model AI after human thought

[mph4ever]

Sure
[/quote

Come on ...... this argument is just ****ing stupid, and I'm not going to entertain it.

You're dodging the point, which is this: what is the essential difference between thoughts made of electrical impulses and thoughts made of electrical impulses

Oh yeah

There is none


i don't get that. you claim that a load of scientists reckon that they will develop intelligence equivalent in every way to the human brain and mind, right? and you base this claim on the fact that we use electricity and so do ai machines? but what you are forgetting is that the machines are mimicking based on speed, slight of hand, deceptive power to make its reasoning appear human like. do you claim that they have invented something that has a mind of its own. that has free will, that has evolved through millions of years and adapted at each stage of evolution to survive and has recorded these developments so that they are handed down to the next generation, a mind that can relate things within the layer of broad content that humans can, or can you? if you can then you must make your case for it. not post some toys and technologies that push the boundaries of progressive thought in the "what if" situation. in all the threads started on this topic i have yet to see someone contribute a single post that leads me to believe that we even understand the mind let alone create one.



BTW: does this look like "castle in the sky" **** to you?

http://www.tudou.com/programs/view/V1N6dcDDJGY/

I wouldn't worry unless you live in Waziristan

At least not for the time being

it looks like an angel of death

EDIT - wow moderator decision cut a lot of dumb peanut gallery **** out of my thread and actually seems favorable this time
i dislike the idea of posts being entirely deleted but i guess reality is that some pigs are more equal than others

[Smokey D]

Those posts are still there, but they were annoying me when they were visible. If you want to keep them for your diary, I can forward them to you.

[mph4ever]

Nah you're ok.

[GorgeousGabe]

do you claim that they have invented something that has a mind of its own. that has free will, that has evolved through millions of years and adapted at each stage of evolution to survive and has recorded these developments so that they are handed down to the next generation, a mind that can relate things within the layer of broad content that humans can, or can you?

Having evolved through millions of years and having adapted at various stages of evolution are not prerequisites for (what we consider to be) free will. Regardless of how it came to be naturally, the final product is a biological machine which communicates with itself using electrical impulses fired in unfathomably complex patterns.

Computers certainly can do the same thing. The question isn't whether it's possible or not, the question is when will we be able to create it

[1338 h4x0r]

Explain why it is an appeal to absurdity.

"lol digitul kompewturs kant have sumantiks (except that our brains r digitul kompewturs)"

lame argument by him full of question begging

http://www.kurzweilai.net/meme/frame.html?main=/articles/art0495.html?m%3D19

i don't get that. you claim that a load of scientists reckon that they will develop intelligence equivalent in every way to the human brain and mind, right? and you base this claim on the fact that we use electricity and so do ai machines? but what you are forgetting is that the machines are mimicking based on speed, slight of hand, deceptive power to make its reasoning appear human like

In the 21st century, there is a stronger focus on copying (not "mimicking") neural activity, which is what makes thought as we know it

i.e., artificial brains in development and in existence (like the X-45A autopilot you saw in action) do not simply rely on exhausting every single possibility in massive search trees

You're using grossly outdated information


do you claim that they have invented something that has a mind of its own. that has free will, that has evolved through millions of years and adapted at each stage of evolution to survive

Natural selection is irrelevant here. Why would you need it?

[Smokey D]

"lol digitul kompewturs kant have sumantiks (except that our brains r digitul kompewturs)"

lame argument by him full of question begging

I don't know why you're so hostile towards him.

I guess his answer would be that if computers can have semantics, then it shouldn't be hard to find a computer with semantic capabilities.

Also, speaking of question begging: "except that our brains r digitul kompewturs"

lol.

[1338 h4x0r]

Also, speaking of question begging: "except that our brains r digitul kompewturs"

lol.

Digital - working in essentially discrete phenomena (I'm glossing over some details here but axons typically "fire" or "don't fire")

Computer - device primarily suited to calculation

ta da

[Smokey D]

That's a fairly cretinous simplification, I think.

You still haven't shown how an artificial computer can have intentionality, which I think is a pretty important part of mind for strong AI.

Moreover, I don't think you're fairly representing AI research, since strong AI is only the goal of some researchers.

[1338 h4x0r]

That's a fairly cretinous simplification, I think.

Why

You still haven't shown how an artificial computer can have intentionality, which I think is a pretty important part of mind for strong AI.

Why wouldn't it

Moreover, I don't think you're fairly representing AI research, since strong AI is only the goal of some researchers.

Their goals, or current goals at least, say nothing about what they think is possible

Some of 'em are just personally interested in robot vision or NLP or whatever and that's fine

Really the only way you can say that only "weak AI" is possible is through metaphysical explanations (which are considered especially ridiculous in this whole community) or by ideas that have had holes shot through them, like the question-begging Chinese Room scenario and Roger Penrose's Goedelian/Turing machine-based argument (which is actually pretty clever---too bad it's invalid)

(And notice that most of the criticisms of strong AI come from outside the field: Searle is a philosopher and Penrose is a mathematician.)

[Smokey D]

Why

Well, partly because it assumes that just because computers and neurons can do some of the same thing that they are the same thing.

Why wouldn't it

Um... the Chinese Room.


Their goals, or current goals at least, say nothing about what they think is possible

Some of 'em are just personally interested in robot vision or NLP or whatever and that's fine

Presumably some of them aren't interested in strong AI because they don't think it's feasible.


Really the only way you can say that only "weak AI" is possible is through metaphysical explanations (which are considered especially ridiculous in this whole community) or by ideas that have had holes shot through them, like the question-begging Chinese Room scenario and Roger Penrose's Goedelian/Turing machine-based argument (which is actually pretty clever---too bad it's invalid)

Okay but there are serious problems with assuming axiomatically that brains are just a type of computer. Within the research structures of AI it makes sense, but it doesn't necessarily speak to a settled truth.

The Chinese room is no more question begging than the claim that brains are computers.

And you still haven't shown how AI researchers understand all the relevant parts of mind which they'd need to replicate to generate human level consciousness and intentionality.



(And notice that most of the criticisms of strong AI come from outside the field: Searle is a philosopher and Penrose is a mathematician.)

Um. And?

[1338 h4x0r]

Well, partly because it assumes that just because computers and neurons can do some of the same thing that they are the same thing.

And how can we account for an essential difference between them without invoking Jebus?

Um... the Chinese Room.

Didn't you see my rebuttal? It's a shitty argument.

Presumably some of them aren't interested in strong AI because they don't think it's feasible.

Maybe some, but some climate scientists don't believe in AGW. There's no accounting for some opinions.

Okay but there are serious problems with assuming axiomatically that brains are just a type of computer. Within the research structures of AI it makes sense, but it doesn't necessarily speak to a settled truth.

Everything is a kind of computer, but some computers are more powerful and concentrated.

As John McCarthy put it:

My thermostat has three beliefs: "It's too hot in here," "It's too cold in here," "It's just right in here."

The more you think about it the more it makes sense.

And you still haven't shown how AI researchers understand all the relevant parts of mind which they'd need to replicate to generate human level consciousness and intentionality.

Why would they have to do everything like the human brain?

Um. And?

They're frequently not fully informed.

[Smokey D]

I totally already responded to this wtf.

And how can we account for an essential difference between them without invoking Jebus?

I'm not going to speculate. The point is that it's too early to make reach anything conclusive.


Didn't you see my rebuttal? It's a poopty argument.

I don't think you really rebutted it. You haven't shown that machines have intentionality.

And, to be fair to Searle, he doesn't deny that at some point computing machines might gain intentionality. He just says that there needs to be a qualitative shift in the current hardware before that's possible.


Everything is a kind of computer, but some computers are more powerful and concentrated.

As John McCarthy put it:

My thermostat has three beliefs: "It's too hot in here," "It's too cold in here," "It's just right in here."

The more you think about it the more it makes sense.

Intentionality?

Why would they have to do everything like the human brain?


I doubt they need ot replicate everything to create intelligent (responsive) AI, but they need to replicate a lot parts of mind more than we understand in order to approach human level intelligence.


They're frequently not fully informed.

That's not an argument.

[1338 h4x0r]

What is intentionality

[Smokey D]

From what I understand, it refers to the ability to conceive of object as separate from the decision making process about the object.

So a thermostat doesn't have intentionality regarding the temperature, it merely responds according to certain physical criteria. A person on the other hand conceives of temperature as something quite distinct from input data producing an automatic response.

I guess the ability to know what you're doing rather than simply the ability to do it.

[GorgeousGabe]

It's all just as real as we are. You'd think that actually an intelligent species would be MORE capable of creating sentience than random occurrences in nature.

Think about it.

It took billions of years for man to evolve out of the dead space of the universe.

It's only been thousands, and we may be on the verge of creating life and creating sentience.

It makes perfect sense when you think about it.

[1338 h4x0r]

From what I understand, it refers to the ability to conceive of object as separate from the decision making process about the object.

So a thermostat doesn't have intentionality regarding the temperature, it merely responds according to certain physical criteria. A person on the other hand conceives of temperature as something quite distinct from input data producing an automatic response.

How do you know that?

What if you're a bot, Smokey D

I don't trust you