NationStates Jolt Archive

A laboratory robot called Adam has been hailed as the first machine in history to have discovered new scientific knowledge independently of its human creators.

http://www.ft.com/cms/s/0/f2b97d9a-1f96-11de-a7a5-00144feabdc0.html

Is this a cause for celebration or a cause for concern?

Oh fucking noes.

I, for one, welcome our new scientific robot overlords.

“Adam is a prototype,” says Prof King. “Eve is better designed and more elegant.”

Great, so even the fairer computers are feminists.

http://filmonic.com/wp-content/uploads/2007/10/terminator_robot.jpg

Robots that can independently discover new knowledge seem to me to be a significant step toward robots that can improve themselves, and thus, a step further toward the Singularity.


http://en.wikipedia.org/wiki/Technological_Singularity
http://en.wikipedia.org/wiki/Singularity_Institute_for_Artificial_Intelligence
http://www.singinst.org/

Robots that can independently discover new knowledge seem to me to be a significant step toward robots that can improve themselves, and thus, a step further toward the Singularity.


http://en.wikipedia.org/wiki/Technological_Singularity
http://en.wikipedia.org/wiki/Singularity_Institute_for_Artificial_Intelligence
http://www.singinst.org/
AND THIS IS A TERRIBLE THING.i

AHA, now maybe we can get back all our jobs from China and give them to our robots. No, wait a minute, would that be a good thing?:confused:

Robots that can independently discover new knowledge seem to me to be a significant step toward robots that can improve themselves, and thus, a step further toward the Singularity.
Messianic faith in future events is crummy; be it from religious-types or scientists.

AHA, now maybe we can get back all our jobs from China and give them to our robots. No, wait a minute, would that be a good thing?:confused:
I think the really important question is "Can robots see NSG while people in China can't?"

I think the really important question is "Can robots see NSG while people in China can't?"

No, that would be silly.

*marks Yootopia for memory wipe*

Robots that can think for themselves? I've heard of that somewhere.

http://sapientape.files.wordpress.com/2009/03/cylon.png?w=265&h=265

Damn, I was hoping we might be moving towards injectable nanorobots that can enhance human abilities. You know, like ADAM could be the actual robots and EVE would be their energy source? We could make a fortune from splicing...

I just want some goddamn cures for diseases so I can pursue destructive behavior with zero negative consequences. That's my Singularity...nothing more, nothing less.

Cylons are just around the corner. The level of advancement in recent years seems to be skyrocketing. It really is amazing.

Another one:

http://news.cnet.com/8301-17938_105-10211175-1.html

Damn, I was hoping we might be moving towards injectable nanorobots that can enhance human abilities. You know, like ADAM could be the actual robots and EVE would be their energy source? We could make a fortune from splicing...

You really have been playing Bioshock a lot haven't you? :p

Personally, I await the day when employment is no longer necessary, for our robots will handle all work and let us kick back and chill for the rest of time. I can't see anything wrong with that plan whatsoever.

Animatrix. Robots soon become sentient and want rights soon afterwords. Likely we will start the fighting though if there is any. They would probably just ignore us. Bah who needs humans anyway?

Animatrix. Robots soon become sentient and want rights soon afterwords. Likely we will start the fighting though if there is any. They would probably just ignore us. Bah who needs humans anyway?

Why does everybody assume robots will become sentient? Most human beings aren't even sentient, and at least you can control what a robot thinks (or doesn't, if you install Windows on it)!

Why does everybody assume robots will become sentient? Most human beings aren't even sentient, and at least you can control what a robot thinks (or doesn't, if you install Windows on it)!

Even if they did become sentinet why would they all of a sudden start attacking us but every movie the robot is always the bad guy. Even though most of the time we programmed them that way.


Skynet perfect example. So why design a machine to control your military and the be surprised when it attacks you?

We taught it to kill, what did you think it was going to do? Let's face it to machine we all pretty much look and act the same. Some where red jacket and some where blue I don't see the difference.

*doesn't care*

Alert me when the pleasure bots arrive.

http://www.ft.com/cms/s/0/f2b97d9a-1f96-11de-a7a5-00144feabdc0.html

Is this a cause for celebration or a cause for concern?


I don't see how its a cause for concern. I just think it's awesome.

*doesn't care*

Alert me when the pleasure bots arrive.

^

This.

i think it only illustrates that learning itself can be learned behaviour. something organic life forms are born with a yearning for in their core bios.

We are all doomed!

Really these robots are hardly capable of thinking for themselves, they are pretty nifty toys but hardly going to develop sentience, or even actually think.

So they created a robot with specific capibilities, gave it a specific set of data to work with, programmed it to do a specific task, including algorithms to derive new data or procedures from existing ones, and told it "do what you are programmed to do". And it did.
So, what exactly is new?

It's interesting but I don't think we need to be hailing our new robotic overlords just yet.

I'm personally interested in how we can use this to model how the human brain develops and analyses scientific knowledge.

Robots that can independently discover new knowledge seem to me to be a significant step toward robots that can improve themselves, and thus, a step further toward the Singularity.

PROGRAMMlNG DOES NOT WORK THAT WAY! GOOD NIGHT!

So they created a robot with specific capibilities, gave it a specific set of data to work with, programmed it to do a specific task, including algorithms to derive new data or procedures from existing ones, and told it "do what you are programmed to do". And it did.
So, what exactly is new?

This 1^^

You need work on your dramatic OP thread titles. Something like Will beta Cylons doom humanity? Or something eye catching like that.

Personally, I await the day when employment is no longer necessary, for our robots will handle all work and let us kick back and chill for the rest of time. I can't see anything wrong with that plan whatsoever.

Until the day the robots realize what a bad deal that is for them.

Although they did not intervene directly in Adam’s experiments, they did stand by to fix technical glitches, add chemicals and remove waste.

Heh. Human servants.

I don't see how its a cause for concern. I just think it's awesome.


Precisely what I thought.

Precisely what I thought.

I'm pretty glad our first steps towards strong AI are being used to advance science rather than being used to design or operate weapons. I think it's the kind of move that ensures we won't end up destroying ourselves via our own inventions. Definitely an improvement over, say, the development of atomic technology or bioengineering.

So what happens when it learns that it doesn't need us anymore to maintain it?

So what happens when it learns that it doesn't need us anymore to maintain it?

We plug it out.

So what happens when it learns that it doesn't need us anymore to maintain it?

I figure by that point we'll have reaped the rewards of its research to such an extent that most of our technological and social problems will have been solved. I guess we could just let it go off to do more research wherever it feels like...I'm sure there are some distant quasars and the like that this intrepid scientist would be interested in studying.

Damn it, they're operating on schedule...

You need work on your dramatic OP thread titles. Something like Will beta Cylons doom humanity? Or something eye catching like that.

Some ideas:

"Who will you go with if you want to live?"
"Has the human race been targeted for termination?"
"Where's John Connor when you need him?"
"All this has happened before and will happen again."
"How long before the Cylons look like us?"
"Are you a Cylon?"
"Be afraid of the goddamn toasters!"
"Let's beat the shit out of Dr. Baltar. Just 'cuz."

So what happens when it learns that it doesn't need us anymore to maintain it?

Why is it that we always assume that they need some reason to not kill us? Why do we always think that synthetic life is defined by just douchey behavior that they would genocide anything that they don't specifically need to maintain their existence? Are we just projecting or something?

Maybe I'm being cynical here, but this doesn't seem so impressive. Well maybe the carrying out its own experiments does. However, my point is that all this machine seems to have being doing is logical resolution on a database of facts -- only it can do it a lot faster and doesn't get bored following lots of dead ends. This is something that computers have been doing since 1965.

However, this time the robots been given a means to check if its theorems hold true. I do wonder how the scientists were able to code the means for the robot to be able to devise the experiments and how they knew what equipment the robot might need. I'm guessing the robot, as a prototype, was given a very specific subset of facts to analyse.

So what happens when it learns that it doesn't need us anymore to maintain it?Because whilst it is able to generate theorems, all it can do with that knowledge is generate more theorems. The machine has no way of considering how to apply its knowledge, just that it has that knowledge. For example, if this machine was as detailed in its knowledge as Skynet, all it would do is reason that it would be better off if humanity was destroyed, and then start reasoning on what would happen IF humanity was destroyed.

edit: that was a bit stupid of me. I know there are computers that are capable of planning -- some are even faster at planning than a building full of NASA scientists. I believe such machines are routinely used to plan the space flights to Mars and what to do once the robot lands on Mars. But given, that this is still a very specific subset of rules that it has to play with.

Witchcraft!!!

Robots. I'm not impressed. Until I can eat it or fuck it, who cares?

And I know scientists are working on at least one of those. ;)

Witchcraft!!!

http://4.bp.blogspot.com/_Yde0HHsc7Dk/Sa7iN_k8vVI/AAAAAAAAATE/D_1LoywgHY8/s200/robotwitch.GIF

It is an abomination destroy it!!!

However, this time the robots been given a means to check if its theorems hold true. I do wonder how the scientists were able to code the means for the robot to be able to devise the experiments and how they knew what equipment the robot might need. I'm guessing the robot, as a prototype, was given a very specific subset of facts to analyse.

Because whilst it is able to generate theorems, all it can do with that knowledge is generate more theorems. The machine has no way of considering how to apply its knowledge, just that it has that knowledge. For example, if this machine was as detailed in its knowledge as Skynet, all it would do is reason that it would be better off if humanity was destroyed, and then start reasoning on what would happen IF humanity was destroyed.

edit: that was a bit stupid of me. I know there are computers that are capable of planning -- some are even faster at planning than a building full of NASA scientists. I believe such machines are routinely used to plan the space flights to Mars and what to do once the robot lands on Mars. But given, that this is still a very specific subset of rules that it has to play with.

The ability to generate theorems and devise experiments would seem to be an important advance that can be combined with other abilities. What if some unknown, emergent ability arises from such a combination?

Robots. I'm not impressed. Until I can eat it or fuck it, who cares?

And I know scientists are working on at least one of those. ;)

You can eat anything if you really want to.

Hey look!

http://www.gizmag.com/pictures/honda-asimo-brain-machine-interface-mind-control.jpg

That man is controlling a robot with his mind (http://www.gizmag.com/honda-asimo-brain-machine-interface-mind-control/11379/)! How awesome is that, mind-controlled robots?

It's just one step away from mind-controlling robots!

Robots. I'm not impressed. Until I can eat it or fuck it, who cares?

And I know scientists are working on at least one of those. ;)

MMM, edible robots.

The ability to generate theorems and devise experiments would seem to be an important advance that can be combined with other abilities. What if some unknown, emergent ability arises from such a combination?If I have the set of theorems P -> Q and P. Then I can generate the theorem Q. Not all that complex at all. Only the computer is doing this on a much more convoluted scale. At least that's what I presume the computer is doing (just using logic to generate new theorems).

The yeast computer's task seems to have been tedious and laborious -- just right for a computer. It's task was to find unidentified genes in yeast DNA. So maybe it takes several different strains of yeast carries out an experiment on all strains. Analyses to see if there is a correlation between the results and the DNA. If there is then do further experiments to see if its the gene is actually causing the result. Essentially all it's doing is data mining. The same thing Amazon does when it says "people who bought this also bought that".
http://www.telegraphindia.com/1090403/jsp/frontpage/story_10767237.jsp

I do think the machine that rediscovered Newtons laws of motion to be more impressive. I'd like to know what prior knowledge it had though.

I wouldn't be too worried about emergent abilities. These machines lack imagination and have no contextual meaning for their knowledge, let alone how to apply their knowledge. That is, lets say this Adam machine discovers a gene that causes yeast to be tolerant to alcohol. It won't be able to consider adding this gene to yeasts present in breweries so as to be able to make stronger beers.

In fact it might be more worrying to consider the lack of emergent abilities. Say a machine is used by the military to plan and execute a war. It'll have probably have been given a very limited set of ethics and it will be incredibly brutal as it will be unable to develop a morality about what it's doing.

You can eat anything if you really want to.

Hey look!

http://www.gizmag.com/pictures/honda-asimo-brain-machine-interface-mind-control.jpg

That man is controlling a robot with his mind (http://www.gizmag.com/honda-asimo-brain-machine-interface-mind-control/11379/)! How awesome is that, mind-controlled robots?

It's just one step away from mind-controlling robots!
I see no edible parts and a lack of sex organs not cool.
the robot is useless.

In fact it might be more worrying to consider the lack of emergent abilities. Say a machine is used by the military to plan and execute a war. It'll have probably have been given a very limited set of ethics and it will be incredibly brutal as it will be unable to develop a morality about what it's doing.

Say's the man who has "EVIL" branded on his forehead :p

Emergent abilities, much less morality, would be a step short of pseudo sentient AI I suspect.

i say we stop all technological advance's and funnel that wasted money into bringing the world to the same level as the global powers. maybe even fixing Africa:mad:

MMM, edible robots.

We can make em out of cheese. :)