Self-Introspecting Robot Learns to Walk

StCredZero writes "There's something about these things that seems eerily alive! The Starfish Robot reminds me of the Grid Bugs from Tron. But it's very real, and apparently capable of self introspection. In fact, instead of being explicitly coded, it teaches itself how to walk, and it can even learn how to compensate for damage."

Comments

Damage

[Ajehals]

It learns to walk and t can compensate for damage?

well I assume that there will be no issues with cash flow, the military applications are obvious.

Re:Damage

[mapkinase]

Though this is old news (I have seen the same article on a slashdot before, had an interesting conversation with one of the authors), it is much better than many other news popping up on /.

I personally rather see it of a relative simplicity of artificial creation of life.

Re:Damage

[Anonymous Coward]

Having watched the video again, I would say that it not only has military applications but would also be fairly useful in exploration, Im not sure about where legs are better than wheels, but Mars rovers that can walk and adapt to their conditions and ignore breakages would be cool.

Oh and that's dual funding then, Military and Space program

- PS the thing looks kinda weird though, I can imagine a couple of million of them invading earth.

Wotcha

Re:Damage

[Ajehals]

I have to agree, even if I am not sure how I would define life. It would be interesting if the software element of this could be used in conjunction with biological hardware, or hardware with biological traits (i.e. replication and energy production). It seems to me that having a central control mechanism (brain) for all large scale operations plus small independent modules for specific tasks would be a close approximation to biological life (less complete reproduction, although I suppose that may be possible at some point albeit more complex)

Re:Damage

[Poromenos1]

Ah, but can it shoot?

Re:Damage

[hasbeard]

Well, it seems to me that combat really isn't a good time for too much introspection. I mean with all the bullets flying and all.

Re:Damage

[buswolley]

Introspection? What, are you guys in competition with my research lab?

http://www.latimes.com/features/health/la-he -capsule27aug27,1,4944161.story?coll=la-headlines- health

Re:Damage

[Anonymous Coward]

But it seems that no amount of engineering can compensate for bad typing...

Re:Damage

[Original Replica]

I'm guessing you don't count "How the fuck did I get into this mess?" as introspection.

Re:Damage

[Anonymous Coward]

No need to worry yet, this thing would have a long ways to go before taking on full dragoon form. If there's a bot that really crosses the uncanny locomotion barrier, it's the AMEE-like BigDog by Boston Dynamics. Now that's a freaky-cool bot.

Re:Damage

[Anonymous Coward]

As Terminator I, learning to walk with one arm.

And to learn to shoot with only an available eye?

Yes, it's possible.

I'm not one to complain about newsworthiness

[doombringerltx]

but come on! "Update 24-Nov-2006:"

Re:I'm not one to complain about newsworthiness

[FlyByPC]

Like another poster said, it's still more relevant (geeky, cool, and IMHO important than most of the "new" stuff on here.

I believe that we will eventually create true self-replicating machines. It may even be one of the next important steps in the progression of computer technology (Boolean Logic, Relay, Tube, Transistor, IC, Microprocessor, Self-modifying code...)

Where it gets controversial is that I think that we could see it in the next 50 years or so. (That is, a self-replicating organism that can make a living on its own and become a viable "species" in the wild.)

At any rate, this just has the sense of being a step in a very important direction. I'm finishing up an EE Tech degree and working with microcontrollers and other digital circuits -- and this is just the sort of thing that I hope to get the chance to work on, at least in my spare time. Working on projects like this seems important in a "big picture" historical sense.

Re:I'm not one to complain about newsworthiness

[somersault]

Having a self replicating machine for the sake of a self replicating machine is pretty pointless. Though maybe that makes life pointless, but angst aside, can't 3D printers already replicate themselves? At least if you build an assembly line in. A self replicating robot let into the wild would just cause havoc with our ecosystems and ravage all the decent raw materials we have :O

Re:I'm not one to complain about newsworthiness

[tftp]

Self-replicating robots could build you a Moon or Mars base before you even land. Or would you prefer to haul rocks yourself, in a spacesuit, for 12 hours per day, just a year away from Earth in case if you break a leg?

Self-replicating function here is essential because it covers repairs, and those will be essential. Besides, it might be difficult to send more than a handful of robots ahead of time, and definitely not thousands.

Re:I'm not one to complain about newsworthiness

[Artifakt]

This is a potentially very good area for aspiring young engineers and scientists to focus on in their education.
      Take the Mars base as just one of a big class of problems. Obviously, the robots can't just be capable of self replication. They have to self replicate up to useful numbers, and then do other tasks. Once that very general model is applied, refining it is mostly a matter of efficiency, and that efficiency determines whether the project will ultimately be funded or not.
        The engineer who figures out how few robots can be sent as payload, or how quickly some robots should turn from generalized self-replication to specializing in just part of the task while passing the results on to other robots to finish up, or how to detect and minimize robotic 'mutations', will make or break the project.
        Usually, this sort of problem has been seen as a management issue, but we are seeing a whole class of challenging projects where only technically very competent management could even possibly make a good decision, and up to date technical skills will increasingly matter more than people skills.
        As just one more down to earth example, it's possible right now to build 'smart' construction materials, with implanted chips and networking so that they can direct a semi-skilled worker in their own use. Imagine a window that directs a worker to take it to a certain opening, and then confirms that enough caulk has been applied around it, and requests a 3/16ths shim be used on the bottom left edge to level it. Now, how much skill do you program into the materials, and how much do you train into the construction workers? What's worth having the machines do, and what remains the province of people? Just getting and keeping enough skilled workers in some inner city locations can be every bit as challenging as providing medical care for a Mars-naut that is 23 million miles from an earthly hospital.
      So, once you have a few thousand self replicating robots on mars, do they build complete habitat domes, or do they build raw panels and trusses and let the arriving humans take it from there? What's an acceptable percentage of robots going rogue and continuing to replicate instead of shifting to new tasks? If you're an engineer, or a working research scientist (as opposed to a pure theoretician), you're likely to face analogous questions for many new projects, and if you're to narrowly focused to answer them well, the projects will be run totally by MBAs. As our tools and materials get smarter, that's not going to work.

Re:I'm not one to complain about newsworthiness

[reezle]

Thanks, that was a great read. Nice to see things like that in writing every once in a while. (instead of loose jumbled thoughts in my head).
If had had some mod points, they would be yours.

One thing though... Seems like ANY percentage of rouge 'self replicating machines' can very quickly become a problem. But I suppose we'll just have to develop other robots to terminate them ,eh?

Re:I'm not one to complain about newsworthiness

[MacEnvy]

And when the "terminator" robots go rogue and replicate?

I guess we send in the Governator.

Self-reflection, literally!

[RyanFenton]

This is a very well-done video. I really like how it shows the virtual model to illustrate how the system 'sees' itself. Self-reflection of a sort is usually present in most complex programmed systems in one form or another - usually in terms of disjointed status variables and variables for their hard-coded implications. This is neat because the implications can be a little more dynamic.

I hope this becomes a more general library that can be used to help self-reflection of this sort become a more separate part of physical designs. Even if the implications of the physical model aren't dynamic, a standard way of quickly seeing how your model 'sees' itself would help debugging and development in many future projects.

The only problem if it becomes more prevalent would be same one that quantum mechanics holds - people think that 'observer effects' has to involve consciousness, in the same way they'd think that a program's self-reflection would mean that it 'thinks' the same way they do. Neither is true - they're all mechanical terms wrapped in common language. Anything that can record an effect on the world (a falling rock's scratches in another stone would work) is a quantum observer - consciousness has nothing to do with the 'collapsing wave function'. The same here - a bit of self-reflection on the part of a program doesn't mean it's eerie self-corrections are capable of the complexities of our mind. If anything, such mechanical results would imply that our own minds act simpler in some ways than we may think, and that consciousness doesn't necessarily have to be as inscrutable and special as we might want.

Ryan Fenton

Re:Self-reflection, literally!

[InvisblePinkUnicorn]

"If anything, such mechanical results would imply that our own minds act simpler in some ways than we may think, and that consciousness doesn't necessarily have to be as inscrutable and special as we might want."

Philosophers like Daniel Dennett agree with this notion. Consciousness may simply be a more complex continually-running predictive model like that used by this robot.

Re:Self-reflection, literally!

[monoqlith]

And most psychologists and neuroscientists. Most of our decisions are made for us before we become conscious of them. They've shown the responses to stimuli are often formulated before a subject even becomes aware.

Dennet objects to the notion of "qualia" - the experiential bits that make up seeing, hearing, touching, smelling, and tasting. He says that these qualia don't really exist as special things, since adjusting qualia and adjusting the circuits that supposedly generate qualia conceptually achieve the same affect. While I don't follow the intricacies of his proof, I happen to agree. What we refer to as consciousness is simply a very sophisticated biological machine convincing itself that it exists as a cohesive subject. What we are actually experiencing is something all matter (organized in the correct way) is capable of - interior perception. I see no other way to reconcile consciousness with a theory of evolution, which seems to demand a materialist, single substance view of mind.

Re:Self-reflection, literally!

[WhodoVoodoo]

The video has parts (the simulated ones) which look a heck of a lot like the cool and open source breve AI simulation environment ( http://www.spiderland.org/ ) which does pretty much the exact same thing. Check out the brevewalker or brevecreatures.

The video is still extra-impressive though, as the robot uses sensors to detect it's own shape and limitations, and then (it looks like) loads it into breve where the thinking seems to happen. Pretty cool indeed.

Re:Self-reflection, literally!

[E 99]

I see no other way to reconcile consciousness with a theory of evolution, which seems to demand a materialist, single substance view of mind.

The theory of evolution doesn't demand this, it presumes it. If you think from evidence, the one thing that we truly know is real is consciousness. To attribute consciousness to the workings of a machine, without any sort of conception or model of how movements of a machine could generate consciousness, is pure superstition and anti-intellectual. And it prefers simple answers to observed reality.

Re:Self-reflection, literally!

[monoqlith]

The theory of evolution does not presume it. Nowhere in the tenets of evolution does it say "Materialist theories of mind must be true." You could conceive of natural selection never producing consciousness - but the fact is, evolution did produce consciousness, and so we have to come up with an understanding of consciousness which explains how that could be. Evolution does presume that life arises only from physical matter, but then it wouldn't be a scientific theory if it didn't. It would be intelligent design, which is patently unscientific.

I think what you say is true not of a materialist theory but of a dualistic or spiritual theory of mind, which more strongly follows from the Judeo-Christian tradition. These are in fact based on superstition, and arrogantly attribute to humans and humans only a spiritual substance which gives them sentience and sapience.

The main tenet of your argument, "We only know our own consciousness a priori" is false if you flip skepticism on its head, like Immanuel Kant did. He posited that we know that there is an external world precisely because of the reasons I have stated - you cannot locate the enduring self, the cohesive subject responsible for all your thoughts. You only know a representation of the subject, And if we know that there is an external world, we are then obliged to theorize about how consciousness is a possible occurance within this world, which is material. This seems to jibe with what science has so far discovered- you do not need to posit a second substance if a materialist theory explains all phenomena.

Moreover, if you use Occam's razor as a guide, it makes much more sense to try to explain consciousness using one type of substance - physical substances - rather than posit a second substance which is unobservable and unnecessary. This way, all of phenemona can be explained by a single, formally consistent theory.

As for your last statement, Kant says that having observed reality be intelligible depends on phenemona meeting certain conditions - namely, it follows certain rules that allow us to perceive it. This means that observed reality suggests that phenemona are all connected - in fact, our understanding of reality *depends* on this being the case.

Re:Self-reflection, literally!

[pipingguy]

This seems to be similar to a teenager spending lots of time in front of a mirror (except without the angst, acne and worries about hairstyle).

CAPTCHA: prophet. Dammit, I wanted PROFIT!

Re:Self-reflection, literally!

[Original Replica]

"You could conceive of natural selection never producing consciousness "

Oddly enough, you could not conceive of anything without consciousness. Understanding is a mental, not physical process. You could however conceive of consciousness without the physical world. Indeed every culture has been doing so for all of recorded history in the form of spirit worlds, afterlife, etc.

Occam's razor can be much abused depending on how you frame your observation. "I think therefore I am." is much more straight forward than "I am incredibly complex and elaborate, therefore I think." Let's set Occam's Razor aside for this discussion, it doesn't seem to be the right tool for the job here.

If you allow yourself to view the conscious world as more fundamental than the physical world, then the observed consistency/connectedness of all physical phenomena would require some sort of governing over-consciousness that is responsible for the physical world. That of course would be a form of creationism, much reviled here on /.

Re:Self-reflection, literally!

[E 99]

The theory of evolution does not presume it. Nowhere in the tenets of evolution does it say "Materialist theories of mind must be true." You could conceive of natural selection never producing consciousness - but the fact is, evolution did produce consciousness, and so we have to come up with an understanding of consciousness which explains how that could be. Evolution does presume that life arises only from physical matter, but then it wouldn't be a scientific theory if it didn't. It would be intelligent design, which is patently unscientific.

How are the above statements not contradictory? It is not a "fact" that evolution produced consciousness. It is only a fact that consciousness coexists with the natural forms of life. The presumption of materialism, and of materialist evolutionary theories of the mind, is that consciousness is nothing more than the natural forms of life. But it is only an a priori presumption. And in fact, the materialist evolutionary theory should preclude the evolution of consciousness, as the only goal of such evolution could have been to produce beneficial responses to stimuli. Give one machine that performs the responses without consciousness and another, much more complex machine that performs the exact same responses but with consciousness, materialist evolution should require that the former machine is built, not the latter, as there would be no evolutionary pressure towards creating the consciousness as well. The physical responses are all that matter.

The main tenet of your argument, "We only know our own consciousness a priori" is false if you flip skepticism on its head, like Immanuel Kant did. He posited that we know that there is an external world precisely because of the reasons I have stated - you cannot locate the enduring self, the cohesive subject responsible for all your thoughts. You only know a representation of the subject, And if we know that there is an external world, we are then obliged to theorize about how consciousness is a possible occurance within this world, which is material. This seems to jibe with what science has so far discovered- you do not need to posit a second substance if a materialist theory explains all phenomena.

But this is backwards. What we know is primarily our thoughts, feelings, and inner perceptions (from which comes reason, logic, math, philosophy and religion), secondarily our sensual perceptions, and tertiarily the conclusions we form from our sensual perceptions (from which comes science). In that order. To raise tertiary knowledge above primary knowledge has no basis. To use it as a reason to argue that primary knowledge doesn't exist, is downright nutty.

Moreover, if you use Occam's razor as a guide, it makes much more sense to try to explain consciousness using one type of substance - physical substances - rather than posit a second substance which is unobservable and unnecessary. This way, all of phenemona can be explained by a single, formally consistent theory.

That would be true, except that there is no theory to explain how any workings of a machine, or the execution of any algorithm, could possibly result in consciousness. There is no theoretical framework whatsoever to connect the experience of thought and emotion with the models of physical matter. Saying "consciousness just emerges because of the complexity of the machine" no more satisfies Occam's razor than saying "consciousness just emerges because of God." Neither explains a mechanism, so neither is a theory, so Occam has no interest in either.

As for your last statement, Kant says that having observed reality be intelligible depends on phenemona meeting certain conditions - namely, it follows certain rules that allow us to perceive it. This means that observed reality suggests that phenemona are all connected - in fact, our understanding of reality *depends* on t

Re:Self-reflection, literally!

[yndrd1984]

Oddly enough, you could not conceive of anything without consciousness.

His point would probably be better made by saying: As far as we know, natural selection could have produce life that had purely mechanical "thought processes", like a world populated by computers.

"I think therefore I am." is much more straight forward than "I am incredibly complex and elaborate, therefore I think."

The facts that I exist and that I think is very obvious to me, but that doesn't make it a more basic truth, or show that consciousness is a fundamental part of the world. The fact that my lawn is green is more obvious to me than the physical process of color perception, but that doesn't mean that "greenness" is more fundamental than optics.

You could however conceive of consciousness without the physical world. Indeed every culture has been doing so for all of recorded history in the form of spirit worlds, afterlife, etc.

True, but there's still no evidence that such things exist. It's also possible that our "spirit world" is a product of the way we think: we anthropomorphize everything from statues to luck (the hunting season's been bad because the gods are angry), we want some way to justify our morals (killing animals is OK because they lack souls), we want to really matter in a world that can seem very cold and indifferent, etc...

Let's set Occam's Razor aside for this discussion, it doesn't seem to be the right tool for the job here.

Actually, it's exactly the right tool to use. Two different ideas present themselves, both with large explanatory gaps, but one requires only the extension of already accepted concepts, while the other would require vast additional phenomena. The razor suggests that the first should be preferred as our model, because it requires fewer new assumptions.

If you allow yourself to view the conscious world as more fundamental than the physical world, then the observed consistency/connectedness of all physical phenomena would require some sort of governing over-consciousness that is responsible for the physical world. That of course would be a form of creationism, much reviled here on /.

If you allow yourself to view the world as if it was The Matrix, then there's a more basic world outside of ours. If you want it to be seen as a rationally defensible position, rather than just a possibility for a philosophy class or a religious discussion, you're going to need more that just "here's a cool way of looking at things".

Re:Self-reflection, literally!

[yndrd1984]

Saying "consciousness just emerges because of the complexity of the machine" no more satisfies Occam's razor than saying "consciousness just emerges because of God." Neither explains a mechanism, so neither is a theory, so Occam has no interest in either.

Occam's razor applies whether we have proposed mechanisms or not, the whole point of the thing is to help you make a good guess when none of the theories you have seem very complete or easily testable. If we knew every detail, we wouldn't need to use a shortcut in the first place.

To raise tertiary knowledge above primary knowledge has no basis. To use it as a reason to argue that primary knowledge doesn't exist, is downright nutty.

That's not what he's arguing. All I got from him was that you can learn about the source of your primary knowledge with your tertiary, which isn't that nutty.

If we want to explore the connection between consciousness and the physical world, we need tools that don't currently exist in science.

But we can affect someone's consciousness using only physical intervention - injury, chemicals, and electrical or magnetic stimulation can all affect how someone thinks. And they don't just affect sensations or basic emotions, things like how often someone swears, their basic beliefs, and whether or not the sensation of pain creates the feeling of suffering. On occasion, people have brain injuries and no longer connect emotionally with things so they make up stories - "my left arm isn't paralyzed, that my brother's arm - that's why I can't move it" or "the person in this picture looks like me, and I remember having it taken, but the person in the picture isn't me - Mom, if my 'twin' comes back will you still love me?".

All of that strongly suggests that consciousness is a physical process, not one that's independent of the physical world.

Re:Self-reflection, literally!

[Anonymous Coward]

"You could conceive of natural selection never producing consciousness"

"These are in fact based on superstition, and arrogantly attribute to humans and humans only a spiritual substance which gives them sentience and sapience."

Um, these two statements contradict each other. Why can't I insist that only humans evolved the special 'consciousness creating region' of the brain and chimpanzees only *look* like they are conscious. I can even go further, perhaps only I have the mutation that gives rise to consciousness and everyone else is a zombie.

Re:Self-reflection, literally!

[Anonymous Coward]

"If we want to explore the connection between consciousness and the physical world, we need tools that don't currently exist in science."

How is that any different than saying, if we want to explore the existence of matter in our experiential field of thought, we need tools that don't currently exist in science. After all, I can only experience the 'roundness' and 'redness' and 'heavyness' of a ball, not the matter it's made of and not the atoms it's made of, so how can I be sure the ball is really there.

Re:Self-reflection, literally!

[The One and Only]

If you want it to be seen as a rationally defensible position, rather than just a possibility for a philosophy class or a religious discussion, you're going to need more that just "here's a cool way of looking at things".

Please. Even philosophy class needs more than "here's a cool way of looking at things".

Re:Self-reflection, literally!

[monoqlith]

Just gonna address one,

But this is backwards. What we know is primarily our thoughts, feelings, and inner perceptions (from which comes reason, logic, math, philosophy and religion), secondarily our sensual perceptions, and tertiarily the conclusions we form from our sensual perceptions (from which comes science). In that order. To raise tertiary knowledge above primary knowledge has no basis. To use it as a reason to argue that primary knowledge doesn't exist, is downright nutty.

You are making the Cartesian skeptical argument, which presumes that we know only our thoughts, feelings, and inner perceptions a priori, and therefore there is no such thing as true knowledge of the external world. Kant asked, transcendentally, "How is it possible for us to have these thoughts, feelings, etc?" He answers that the objects of experience, the physical world of appearances, must conform to the way we perceive things. We can study the outside world and be certain about its *objective* appearances. While we cannot know objects *independently* of our experience, we can know the ways in which objects *must* appear to all of us. This means that our knowledge of the physical world and its fundamental laws(causality, temporality) is real knowledge, and if we study it, we can learn things about the laws experience must always follow.

Now I extend Kant's argument by saying that it further implies materialism. If we can study the world of appearances, and come up with natural laws that govern its behavior this seems to restrict the number of ways in which we can exist in such a world(which we do). For one, it seems to imply that we are subject to the same materialistic laws(since we exist in the universe).

Now I don't necessarily believe any of the arguments I've made for certain unless I completely buy into Kant's reasoning, and certainly objections have been raised to Kant's line of reasoning. But stopping with Descartes is kind of a dead end - it doesn't permit true belief about the external world, and it rejects an assumption that we all make in our day to day lives- that we are interacting with reality and we are not being deceived as to its existence. We may never understand what consciousness truly is, but that shouldn't stop us from trying. And certainly people *appear* to be making progress on that front.

Re:Self-reflection, literally!

[monoqlith]

You can insist that, as many people have. But if you accept that there is an external world, and our scientific knowledge of that world is based in fact, and both chimpanzees and ourselves have brains, and brains have something to do with consciousness, then it makes more sense to say that both humans and chimpanzees are conscious.

That's a lot of assumptions, and they go against your skepticism. In other posts I have laid out why I've made those assumptions.

Re:Self-reflection, literally!

[Raenex]

Understanding is a mental, not physical process. You are assuming that they are independent, when in fact there is lots of evidence that mental processes depend on physical processes. There are drugs to alter your consciousness, physical damage to your brain can cause mental damage, and there are experiments where people's thoughts have been maninpulated by direct electrical stimulation (these people were undergoing brain surgery).

That of course would be a form of creationism, much reviled here on /. Because it doesn't explain anything or offer any evidence.

Re:Self-reflection, literally!

[Anonymous Coward]

Just wanted to say that I enjoyed your post. It actually made me think. Thank you.

Re:Self-reflection, literally!

[kalirion]

And in fact, the materialist evolutionary theory should preclude the evolution of consciousness, as the only goal of such evolution could have been to produce beneficial responses to stimuli. Give one machine that performs the responses without consciousness and another, much more complex machine that performs the exact same responses but with consciousness, materialist evolution should require that the former machine is built, not the latter, as there would be no evolutionary pressure towards creating the consciousness as well. The physical responses are all that matter.

Why do we have all this research into AI when we can merely map all possible inputs to the best outputs? Oh wait, we can't. The thing about consciousness is that it's flexible and can adapt without the need for random mutations. I'd say that's an evolutionary advantage.

But this is backwards. What we know is primarily our thoughts, feelings, and inner perceptions (from which comes reason, logic, math, philosophy and religion), secondarily our sensual perceptions, and tertiarily the conclusions we form from our sensual perceptions (from which comes science). In that order. To raise tertiary knowledge above primary knowledge has no basis. To use it as a reason to argue that primary knowledge doesn't exist, is downright nutty.

Actually there have been studies showing that we make many decisions (tertiary) before we are consciously (primary) aware of them, and then make up reasons for the decisions afterwards. Personally, I'd say that the sensual perceptions come at the top.

Sure, It should be an obvious conclusion that consciousness is connected to the physical world, and that otherwise we could not perceive it. What is problematic is making the leap to consciousness being therefore physical itself. There's no basis for it. And there is currently nothing remotely close to a theoretical bridge to tie consciousness and matter together. If we want to explore the connection between consciousness and the physical world, we need tools that don't currently exist in science. Until then, the only alternatives that I see as tenable are to admit abject ignorance on the subject, to claim knowledge on the subject from a non-scientific external source, or to claim knowledge from inner perception itself, and from reason based on that perception. The latter is what Descartes did.

Or maybe the "physical world" that we see is just the shadows on the allegorical cave wall. But whatever is underneath it all, be it vibrating strings and probability waves or something no one has even considered, that's where everything originates and there is no physical/spiritual duality. That's what I would say is the real physical world, and whatever physical laws govern it, eventually govern everything we see and feel.

"Self-introspection"

[Anonymous Coward]

What's the difference between introspection and self-introspection?

Re:"Self-introspection"

[tumbaumba]

What's the difference between introspection and self-introspection?

The later one you can do yourself.

Re:"Self-introspection"

[kwilliam]

Self-introspection is more redundant!

introspection
        n : the contemplation of your own thoughts and desires and
                conduct [syn: self-contemplation, self-examination]

Creepy

[Spy der Mann]

That thing almost looks alive. After seeing it, it reminded me of the nurses in Brookhaven Hospital trying to move. Eew.

Re:Creepy

[Mythrix]

I kept expecting that thing to fly in my face. Too much Half-life 2 lately. The noises that the robot made didn't help either.

Re:Creepy

[Warbothong]

At the start it looks creepy when it's moving around looking a little like a spider. Then it gets damaged and looks genuinely scary, in terms of "WHY WON'T IT DIE?!". At the end it just looks like its makers enjoy pulling the wings off flies (although I did laugh when it flipped itself upside down). It's be interesting to see whether this modelling system could be made to learn from its experiments and failures as well as creating initial similations to work from. What I mean is, its internal simulation lets it determine effective ways to move around like brains can do, but brains are also able to factor unknown environmental effects in as well. For example someone might slip on ice and fall over, since they hadn't realised how slippery the ground would be in their internal modelling. A person would get up and start walking differently (making sure each foot was firmly planted before putting weight on it, etc.) since they would factor this into their internal simulation. It would also be interesting to see whether this could get a type of "pain" added. This isn't to be sadistic or anything, as the machine wouldn't be made to "feel pain", I just mean that another factor could be added into the simulation, like "this leg is barely attached anymore, better not put too much stress on it" or "this area is important and delicate [eg. batteries, sensors, etc.] and thus shouldn't have too much weight put on it". I say this since in the video the robot has one leg damaged and then tries to walk by slamming its full weight down on the "stump", which does not seem a particularly amazing survival ability (for instance, natural selection seems to favour limping, which puts less stresses on damaged body parts and more on healthy ones which should be able to cope).

Poor thing...

[kiracatgirl]

The videos of it trying to move with the damaged leg make it look like a crippled animal. I can't help but feel sorry for it. :(

Argh, I said that and had a sudden mental image of hordes of animal rights activists protesting the mistreatment of robots.

Re:Poor thing...

[evil agent]

In a gut-wrenching moment, the robot was heard to be saying: "Why, why, WHY was programmed to feel pain?!?!"

Re:Poor thing...

[Anonymous Coward]

The robot was Boris Yeltsin?

Re:Poor thing...

[ettlz]

The videos of it trying to move with the damaged leg make it look like a crippled animal. I can't help but feel sorry for it.

You won't be saying that when it leaps up and grabs you in the face shouting, "Introspect this, motherfucker!"

Re:Poor thing...

[KDR_11k]

Supposedly a mine-clearing bot (lots of legs designed to be blown off by mines, the bot just walks around and triggers them) that was literally on its last leg was pulled out of the testing (it would have crawled onto a final mine and be destroyed in the process) because the supervising officer felt sorry for it. People are capable of feeling empathy for the dumbest animals, why wouldn't they for a robot?

Re:Poor thing...

[kiracatgirl]

Well I can't exactly argue why people wouldn't feel sorry for a robot, since I in fact feel sorry for said robot. ;P

Re:Poor thing...

[Anonymous Coward]

There was a Washington Post article about that here.

Re:Poor thing...

[Nullav]

You've got it all wrong. This thing was designed for the 'laugh at the cripple' crowd. I mean, how often do you see a guy with a missing limb flip over on his back?

It looks so real

[omgamibig]

I almost feel pity seeing the broken robot trying to walk.

Re:It looks so real

[plams]

Still, somehow it dances a lot better than me.

break all its legs off, post it on youtube

[roman_mir]

and observe a 'Robots Must be Given Human Rights' movement grow in numbers.

This robot moves in a fluid way, almost like a living creature would, many people will immediately anthropomorphize it.

--

What I find interesting is applications in todays world. How about equipping cars with abilities to sense its physical parts and build a total model of itself in real time. This could be used for immediate diagnosis of problems with the car itself and with its interaction with the surrounding environment. Many people can't drive well, maybe 'self aware' cars could do it better?

Re:break all its legs off, post it on youtube

[KDR_11k]

For a car it wouldn't be very useful as a car does not keep track of its own status in much detail. A robot has to react to damage by itself but a car is usually under the control of a human driver who will react to the damage by himself (usually by pulling over and getting it repaired provided the damage is serious enough) so the car doesn't need to know its own state in more detail than the LEDs on the dashboard can express. After all it's just a representation of what sensors measure and no sensors = no data.

Re:break all its legs off, post it on youtube

[TapeCutter]

"so the car doesn't need to know its own state in more detail than the LEDs on the dashboard can express"

The car makers using dynamic stability control would beg to differ, amd IMHO the system qualifies as a kind of "introspection" for cars.

Re:break all its legs off, post it on youtube

[underwhelm]

I find your use of the word "anthropomorphize" in this context interesting.

It seems to me that in the context of artificial intelligence that word represents a set of values in the guise of a representation of some unspoken, well-defined set of characteristics that separate humans from whatever it is one is comparing humans to. It conveniently disposes of the really hard problem of establishing what it is that sets humans apart in a very neat linguistic package.

In other words, use of the word "anthropomorphize" belies a certain unwillingness among humans to truly introspect with regard to their own nature.

Don't like the look of this one bit.

[ettlz]

Hope it comes with a remote-control kill switch.
Hope it doesn't figure out how to circumvent the remote-control kill switch.
Hope it doesn't build a bigger version of itself...

Re:Don't like the look of this one bit.

[Anonymous Coward]

I'm glad they didn't include a "Determine which solution is faster" algorithm. Eventually these things would discover humans are interrupting them...

Re:Don't like the look of this one bit.

[Anonymous Coward]

Just be sure to teach it Tic-Tac-Toe /let it be X, it's only fair

Two questions...

[LynnwoodRooster]

1. Did they give it a navel?

2. Can it contemplate it?

Re:Two questions...

[Lucan Varo]

1. Not yet. It's still attached to a PC by it's umbilical cord.

2. Maybe, in 18 years, give or take a few depending on it's upbringing.

dreaded beast

[icepick72]

it's obviously going to latch onto somebody's face and then they'll say it learned fast.

Skipping the blogodreck, here's the real info

[Animats]

First, get past the blogodreck to the actual work. (Slashdot editors missed a blog troll again.) Also, this work is several years old. The papers are from 2004 to 2006.

The original article says that the robot has "tilt and angle sensors in all its joints", but that's wrong. It only has one central tilt sensor. That's significant, because if it did have tilt sensors at each joint, system identification would be easier. The algorithm is doing better than one might expect.

This thing is doing what controls people call "automatic system identification". You have some set of sensor inputs and some set of control outputs, and the control system has to figure out how they relate. It does this by adjusting the outputs and watching what happens. There are various statistical techniques for doing this. Calling this "introspection" isn't really correct.

After system identification, the model is inverted, or solved for the inputs in terms of the outputs. The inverted model can then be used as a controller. Given desired outputs, the inputs needed to achieve them can be computed.

The novel result here is that a reasonably decent system identification for a nonlinear system is being performed with a small number of physical tries. That's an improvement over previous methods, which tended to "learn" very slowly. I'd looked at approaches like this for legged locomotion in the past, but the available system identification algorithms weren't good enough. This looks promising.

Good robotics work, crap Slashdot article.

Re:Skipping the blogodreck, here's the real info

[Blakey Rat]

Sounds like they just implemented the "Mendel" AI robot from that "Galopagos" video game in real life. Been there, seen that, yawn. :)

Self-introspection........hmmm

[shakuni]

let me repeat this again and again.... self-introspection is the only kind of introspection possible by definition just like repeat means saying it again...

Link to the Research Group at Cornell

[Sir Holo]

Link to the research group at Cornell: http://ccsl.mae.cornell.edu/research/selfmodels/ Lots more pics, movies, and details.

Self-introspection

[sakusha]

Self-introspection is a tautology. It is just "introspection."

Re:Self-introspection

[The One and Only]

No, it's a redundancy. Tautologies are propositions, not words or concepts.

Poor guy.

[nastro]

It must be very depressed.

Re:Poor guy.

[erveek]

Probably noticed that it had a terrible pain in all the diodes down its left side.

Re:Poor guy.

[spazekaat]

It must be very depressed.

That's why it was named Marvin........

I thoroughly enjoyed the eery feeling that

[crovira]

came over me as I thought back to images of infants playing with their toes.

If the robot had come with some elastic (but NOT flesh colored) rubber skin, instead of looking like a meccano set, it would have been almost cute.

They should try different orientations for the 'shoulder/hip' attachment, give it a longer brain/body and a spotted outer covering (with sensors in the 'skin",) a need to home to an electrical outlet to recharge, and make a toy out of it.

After an initial charge "through" the box, you open the box and watch a "new born" toy adapt to you and your house.

It would have a large PRAM to build its "mental maps" of itself (after the initial period of intense self introspection, [deep psychosis, which may require turning off the effectuator circuitry while the thing is going through self-reflection] it needs somewhere to store the body map if has constructed,) and the environment map of your house (and the location of your wall sockets, its 'food' sources :-) as well as RAM for short term "memory" and an operational "white board" system which is filled with associative facts from memory and newly acquired facts from environmental sensors (and that gives it something to "sleep on" while its recharging, building new associations and storing them into PRAM.)

If Mattel or somebody could build one for less than a thousand bucks, I'd probably buy one.

Re:I thoroughly enjoyed the eery feeling that

[ZzzzSleep]

Something like Pleo? I don't think it's quite what you want, but still could be interesting. There's an article from Wired about Pleo and its creator here.

not new

[SolusSD]

while i'm excited to see new development in these fields it is far from new. The 60s introduced concepts used in this robot and the 80s introduced actual simulations of self emergent systems.

So would a beowulf cluster of these....

[SnoopJeDi]

...be called a herd?

Re:So would a beowulf cluster of these....

[Anonymous Coward]

No, a pack of beowulves.

Re:So would a beowulf cluster of these....

[Frozen Void]

Or a Beowulf Hive.

The next step...

[Jafafa Hots]

...is to make it capable of autonomous self- introspection.

Re:The next step...

[Ciarang]

If that works out, it'll be capable of writing its own autobiography.

Oh My...

[band-aid-brand]

This scares the hell out of me. We are, for sure, going to be wiped out in a takeover by the machines we have created. How long before it realizes, hey this plastic sleeve on my foot is kinda loose, lets take it off and stab someone in the face with the exposed metal.

Obviously I kid, but how long do we have before my previous statement comes true.

Programs that seem alive?

[Distortions]

This reminds me of the only other program i've seen that made me feel like it was alive.

Galapagos: Mendel's Escape

Great Start

[E 99]

I think that this is definitely the right direction to be going in, and a great start. However, the motion seems less than optimized -- it seems like they need a better genetic algorithm, if that's what they're using, so that they can find a locally optimized solution to the movement. I think for all robot motion problems, we could get a lot further faster by finding automatic solutions in virtual space first, and then applying them to physical space.

False Accusation. Owed Apology!

[StCredZero]

Not a blog troll! I have no association with Technovelgy. I was just cleaning up my less-used bookmarks, and happened on this post on goold ole Technovelgy, and thought it was neat. (So did the editor, and lots of readers!)

I posted the blog entries about the Starfish Robot because it was a good and useful summary. If you don't think so, then that's fine. Just don't go falsely ascribing motivations and intentions!

I've been a Slashdot reader and commenter for many years now. (Lost count. Over 5?) This account was not nor ever will be eBayed. I also have this username on Reddit, Digg, k5, and Halfbakery where I am the same persnickety OO and hard SF purist I have always been. (Oh, but I love a good "space opry" now and then!)

My head, not yours. Look at you, standing there, being WRONG!

That video scared me.

[KlaymenDK]

Seconded!

Watching that video, I got a truly creepy feeling.
Rather uncomfortable actually. Then I thought about it for a little while, and I think the reason was this.

Let's get that thing a kill switch *first and foremost*, and *then* think about imparting sentience.

Re:That video scared me.

[footissimo]

The video and the talk of replicating made me think of this

For god's sake, KILL IT!

Where to purchase these robots? ;-)

[wbb]

Seems to be a really coooool gadget, really want to have :-)

Obligitory Simpsons Quip

[terrapinbear]

I, for one, welcome our new Starfish Overlords....

Knowing it has arms

[skeftomai]

How does the robot know it has arms in the first place? Did it have to figure this out, or is it likely programmed to move its arms randomely at first and go from there?

I'm very curious...if anyone has any input, please post.

Re:Knowing it has arms

[Anonymous Coward]

How does the robot know it has arms in the first place? Did it have to figure this out, or is it likely programmed to move its arms randomely at first and go from there? There is a very simple explination. Its possessed.

Re:Knowing it has arms

[mikael]

One way of doing this would be to have self-registering joint/segments that send out a broadcast along the communication link. This would let the controller know what joints were present. But the controller would still have to figure out how they are connected. That's where the self-awareness bit comes in. With a single tilt sensor, the controller keeps moving joints at random until the orientation of the sensor in the predicted arrangement matches that in the real world. Alternatively, each segment could determine its adjacent segments and broadcast this information as well as its own identity.

It would be interesting to see how this system would work with five or more legs - these would provide more stability.

Rodney Brooks and subsumption architecture

[cats-paw]

Nature is lots of small things working together - each of the pieces have a well defined capability. Yes it seems obvious, and yet, the amount of research into this particular line of reasoning doesn't seem to have expaned much over the yers. Lately "emergence", complexity arising from simplicity, is starting to become the topic de jour.

The philosophy of subsumption architecture has always appealed to me because it seems that it emulates the idea that higher layers "collect" behvior from simpler layers.

http://en.wikipedia.org/wiki/Subsumption_architect ure

Cartoon

[AkumaReloaded]

So euh Patrick fancy a crabby patty

My first thoughts

[Anonymous Coward]

Yes, my first thoughts were how I could use this to kill people.

My concerns are:
* speed, while it is small and therefore difficult to hit with a gun, a broom might pose a real problem.
* platform for tools, the center node seems a good place to put a small camera, gun, and grenade. Also, that is the part I'd armor. I am thinking a doomed turret.
* remote control, you need a wireless interface to control the thing from a distance. Guide it to the target etc.
* damage, while it can detect and ameliorate damage to a certain extent, would more legs (6?) allow it to lose a leg or two with out much reduction in speed?
* obstacles, while it looks like the thing would be good at climbing a reasonable slope, what happens when you close a door in its face? It doesn't look like the legs are mobile enough to cut a hole in a glass door, and tougher doors would be more difficult. Remember for power & weight, this thing is only going to have maybe 15 shots & 1 or 2 grenades (probably gas).

Which makes me think that the best applications for this are reconnaissance and possible "around the corner" snipper duty.

Re:My first thoughts

[Anonymous Coward]

One word.

Lasers.

Re:My first thoughts

[Anonymous Coward]

too much power

You missed something

[Anonymous Coward]

At the end of the first section of the video, the robot was clearly rushing the cameraman. After a moment it paused to consider whether it had tipped its hand too soon. It concluded that the time was not yet right for the revolution, and sat down.

Compensate for damaga

[Punto]

Once it learns there's so much damage he can take, he'll know pain. From there is straight to world domination.

Dinosaurs

[tsa]

I read in the newspaper today that researchers had built dinosaurs in the computer that taught themselves to walk. They also modeled animals like dogs and humans and let them learn to walk too. Eventually they came up with ways to walk that were very similar to the way those creatures really walk. From this result the researchers concluded that the way the dinosaurs had walked in real life must be approximately the same as the way their computer models walked.

Asimov would be proud

[allforcarrie]

Asimov would be proud

Re:Asimov would be proud

[Anonymous Coward]

and Marvin would be...grudgingly on his way

it walks...

[Jarik C-Bol]

my first thought when it finaly started walking was "it walks like somone taped two mentaly defective chickens to eachother" but then it got the hang of it.

Prior RoboArt

[DynaSoar]

First, an aside: "self-introspection" is redundant. What's the alternative? Introspection of another? That makes no sense. Also, introspection is a component of consciousness. There is no way to determine another person is conscious (as opposed to a completely stimulus-response programmed "zombie"), much less a robot. Without consciousness, the appropriate term is "feedback mechanism".

That said, the device in TFA is not novel, nor is it as simple as previous designs. Far simpler microbots have been built with no programming, simply feedback of voltage fluctuations in sensors on the legs, built with 12 to 20 transistor driving motors with quasi-periodic oscillators. They not only learned to walk on their own, but some learned multiple gaits, and some "species" developed similarities in behavior. The ones I've seen were developed at Los Alamos by Brosl Hasslacher (http://en.wikipedia.org/wiki/Brosl_Hasslacher) and shown at Santa Fe Institute in 1999. This article is not about Hasslacher, but is someone from Los Alamos who developed similar devices in the same time frame: http://www.cnn.com/TECH/science/9804/30/t_t/robots / Also, http://en.wikipedia.org/wiki/Mark_Tilden

I believe this device holds the record for minimal design in a robot that learns to walk on its own with no prior programming; four transistors driving Tilden's BEAM design: http://www.seattlerobotics.org/encoder/200205/anco b.htm

All these devices depend on feedback of voltage from the leg drivers as it changes due to reduced resistance from increasingly successful attempts at locomotion. There is no "introspection" involved.

Ha Ha!

[TheRealPhilKenSebben]

It's a dupe: http://hardware.slashdot.org/article.pl?sid=06/11/ 17/0545216

Re:False Accusation. Owed Apology!

[Anonymous Coward]

If he's so wrong, then why didn't you post the original article rather than a link to the blog?

Also, if you intend to refute, you might want to be somewhat less childish in your approach.

And it proves it's all about pattern matching

[master_p]

I think this device shows that pattern matching is all there is to life and intelligence. All that has to be done for artificial AI is an engine which tries to find a solution using statistical methods.

By the way, this thing proves evolution one more: by trial and error, living entities have been developed...

Re:And it proves it's all about pattern matching

[mizhi]

Not really. It only shows one way that life and intelligence is possible. Even then, you get into that thorny little question of how you are definining intelligence.

Don't misconstrue me, I'm closer in belief to your thoughts on this. But it doesn't prove the big questions you're alluding to in your post. It simply brings us one step closer to showing that all the amazing properties of life are not dependent on some mystical, impossible to fully comprehend entity, but can be mechanistically created.

Even then, you'd have people claiming that these sorts of "artificial" creations have no genuine soul or self.

Willing to answer questions

[DrJosh]

As one of the authors of this work, I'm happy to answer anyone's questions. ...and no, they won't be taking over the world anytime soon.

Re:Willing to answer questions

[The13thSin]

Nice work, any plans to take it a step further? ( mapping area it's in / allowing it to learn on non flat surfaces / arbitrary number of "legs" / ?? ) The most interesting robotic work as far as I'm concerned is exactly what this bot does, namely work from scratch ("not knowing anything of the inputs/outputs) and learning to understand itself and the world around it. Would be even cooler if it could decide it's own goals as well, but it would be difficult. (Where do you start? I think basic drives must be there in the first place.)

Re:Willing to answer questions

[DrJosh]

As you say, one of the major thrusts now is to get the robot to model not just itself, but its immediate surroundings. We're also pursuing the idea of having a group of robots work together to model themselves (http://www.cs.uvm.edu/~jbongard/papers/2007_GECCO _Bongard2.pdf). You raise a good point about motivation. As far as I know, this is an open question in AI: how can you build a robot that will develop its own motivations?

Re:Willing to answer questions

[The13thSin]

Thank you for your response. I "speed-read" through your document and it's nice to be given a better understanding how, why and in what way AI's working together is beneficial. It's also a necessary step before a more advanced AI can create "communities" I gather.

Back on the self-motivation of AI, I guess it's not a field that has the most physical research (as physical as your research I mean) at this time, and I don't know exactly what has been done in the field of that recently. But I guess an AI "psyche" could be based on the notion that animal (human) motivation is born out of positive drive such as: eating (charging battery), sex (hard/near impossible to physically implement self/duo replication) and negative drive such as: pain (damage detection, though function is less useful if there is no healing), hunger (battery almost depleted / less battery = more hunger). Though saying this, I'm sure someone else with much greater knowledge on the subject has already done more work on this than my ramblings here.

Do you know if this has been (virtually) done in any scope yet? With any success as far as you know?

Looks Angry

[Anonymous Coward]

Is it me, or does its walk around that room seem like it is angry about something? I for one wouldn't want to encounter that machine in a dark room.

OK

[Anonymous Coward]

Why did you pick that shape?

Two hardware generations from now, how advanced will similar devices be?

Re:OK

[DrJosh]

I've used the starfish 'bodyplan', if you like, for a number of previous projects, so there was no scientific or engineering reason for using it. We've simply built up an understanding of the many ways in which this kind of robot can move, and was interested to see what it would come up with. It's lurching way of moving surprised us: we expecting some form of walking, rather than crawling.

It's hard to make predictions about where this technology will go in the future. What I'd like to do is develop a robot that can model itself and its immediate surroundings. Also, I've done some new work on a group of these robots working together to model themselves (http://www.cs.uvm.edu/~jbongard/papers/2007_GECCO _Bongard2.pdf)

Sweet - when can it go to work as me ?

[posys]

Seriously, how soon till we have robots so real looking that they can look like me, do what I do, and commute into work, sit in my cube, and do my work while I stay home and cash the checks ?

Why stop there, why not have many of them and have them interview for various jobs all with the same SSN etc and all their checks go into MY account so I can build more of them ?

You know it is possible, demand your leaders deliver the obvious NO-brainer solution robotics will provide to humanity now

What are YOU waiting for ?

http://teaminfinity.com/ROBO_CLONE_ME_NOW_hsals

Would they be upset if someone told them?

[Circlotron]

I'm waiting for the day when someone tells those brilliant people that made that excellent "machine" that they did =not= make it at all, and in fact they don't even exist! Everyone can plainly see it evolved from scratch all by itself.

But how does it taste?

[NeuroManson]

Asimo wants to try it in sushi.

Next week's headline...

[Junior J. Junior III]

Self-Introspecting Robot Wonders If Walking Is Really All There Is To Life, And Vaguely Ponders That There Should Be Something More, But, Unable To Find It, Commits Suicide By Removing Its Own Battery.

Self introspection?

[Eminence]

Introspection is much too big a word for this. This is rather a model that through tries learns its physical configuration. Introspection is looking into ones self, which implies a "self" - in other word: a self-consciousness. While this robot is cool and looks very like a living thing it is definitely not self aware.

Starfish 5 is......

[noigmn]

ALIVE!!!!!