News >> Browse Articles >> Science
Robot Scientist Able to Conduct Research By Itself
Adam (Source: Reuters)
Michael Barkoviak / DailyTech
April 03, 2009
'A new robot developed in the United Kingdom has the ability to conduct research by itself.' -
Researchers have successfully developed a new robot that is able to reason, formulate theories, and work on new scientific breakthroughs without the need of real scientists.
This marks the first time that a robotic system has been able to make its own scientific discovery with minimal intellectual input from humans. Researchers have been able to create software programs able to analyze data for hypotheses, but this is important because the robot is able to take a scientific experiment from a hypothesis, experiment, reformulated hypothesis -- if necessary -- all without the assistance of humans.
Despite researchers attempting to give robots human features, U.K. researchers are instead focused on workability and effectiveness, not looks.
"On its own it can think of hypotheses and then do the experiments, and we've checked that it's got the results correct," Aberystwyth University researcher Ross King said during an interview with Reuters. "People have been working on this since the 1960s. When we first sent robots to Mars, they really dreamed of the robots doing their own experiments on mars. After 40 or 50 years, we've now got the capability to do that."
Researchers behind the new robot, named Adam, admit that on the surface the recent discoveries have been "of a modest kind" so far, but have faith that the complexity of researched conducted in the future will be higher. To date, Adam has completed research into yeast metabolism, and has the ability to understand the results and plan what to do with the results.
Specifically, Adam was equipped with a yeast metabolism model and a gene and protein database, with human researchers only becoming involved when it was necessary to replace necessary solutions and remove waste. It developed advanced hypotheses then created experiments possible to test its hypotheses.
This is a vital step towards new research technology where humans do not need to be involved as much as currently necessary for experiments.
The research team is working on the next-generation robot, Eve, which will have higher brain power so it can help compile research on new medicines.
© 2009, DailyTech

Justen
7 months ago
212 comments
@ghendric: For the same reason that any other form of peer review exists, to verify that the results produced are accurate. It would be pretty stupid to assume that a process produces accurate results on faith alone without externally verifying those results, at least from time to time.
Justen
7 months ago
212 comments
@icpeace1: woah, settle down there Tanto. Computers have been thinking for themselves for quite a while now. Every time your AI opponents cleverly circle you or lure you into a trap on your favorite modern computer game that's exactly what they're doing. This is not "the appearance of thought", it is thought. It's the computer using its sensory perception to formulate an internal model of its environment, propose to itself various scenarios in that environment, and choose the one that seems to most optimally produce the results it requires. That is the same process you use when you think; some forms of AI even use the same neurological models as biological life. The system is primitive right now compared to the human brain in breadth and depth, but the differences in process and outcome are trivial. It will not be very long, in the span of decades, before the processing capability meets and surpasses that of the human brain; as the gap closes the AI that we have today will be as a protozoan to a reptile, then to an ape, then to a human being.
As for flaws, by what standard are you measuring? You admit that human beings themselves are flawed, but you hold AI to a standard of perfection. Do you suppose that a mind that follows essentially the same method of deduction but is vastly more powerful and has access to vastly more information than a human mind will make more or fewer mistakes? Furthermore, who is looking for perfection, or demanding it?
I find it amusing that your image of a perfect world is one in which humans and machines fight together against other humans and machines, as if the measure of a machine's usefulness is whether it can kill humans as effectively as a human can. The basic belief that this image stems from - that humans must kill other humans, and that all machines are ultimately measured by their success as weapons - is far, far more fatal than trusting a machine to make a decision.
icpeace1
7 months ago
4 comments
Um, excuse me. I'm no scientist. But since when did the human race considered computers and robots superior than human beings??? Computers and robots are tools to assist the humans with various work duties, not to replace humans. Computers and Robots do not think on their own. Computer programmers write software to make the computer and robots appear that they're thinking. But no way is the computer actually thinking. Only humans have thinking ability, not machines. And for those of you who believe in a perfect world where robot and machines fight the front lines? You had better get a serous reality check. Nothing is perfect in this world. NOTHING. No matter how hard and diligently humans work to develop perfect things, those things NEVER turn out to be perfect. Humans have flaws and those flaws manifest themselves into the products that humans develop. So don't think for one second that the world would be a happy, perfect place with "thinking" machines working for us. Thinking that is a bad mistake. Believing in that is FATAL!
digioz
7 months ago
112 comments
This has "Terminator" written all over it!
kartikay
7 months ago
4 comments
Oh yeah, and feedback loops :)
kartikay
7 months ago
4 comments
Now now, lets be fair guys. You think we would be fed crap by you guys, and we would relish it? This thinking robot is a bundle of crap. All you have done is added lots of switch cases and pluggable modules, and you have achieved a low grade machine which Prof.s call a computer. A very low grade machine...
Cruiser052
7 months ago
14 comments
As an aspiring programmer, I would very much like to take a look at the software in this. It is likely that they just need to add a new database into it for each different research projects. So it's just a matter of building the database, then adding it to the existing software. I am sure that, once most of the databases are completed, its programming will be modified so that it can switch between them when it needs to, without human interaction. However, I am sure that some things will need human interaction to be researched effectively. There is nothing like human ingenuity... yet... Hey, do I hear the Terminator soundtrack? Oh crap...
allenlawr1954
7 months ago
4 comments
Is this a True AI machine or just an expert system that needs someone to put in new parameters when moving to a different research project? What, if anything, can we expect to learn form this type of machine???
boltfox20
7 months ago
38 comments
There's only one thing I see going wrong with this... What name will they give the third robot? Snake?
jgatewood
7 months ago
2 comments
This is very interesting. It would be nice to see an explanation of the organization of the software which does all of this. Is it truly AI wherein the machine learns from mistakes and adjusts its reasoning accordingly, or is it more dedicated software to this particular problem - i.e., dedicated to just the yeast research with components directed by what scientists already know about doing this research. Either way it is a great start on freeing humans from tasks which heretofore only humans could do.
scbeacham
7 months ago
8 comments
This is truly amazing. Anyone who throws sarcastic comments is taking for granted the things we have nowadays. Honestly, it truly is an amazing thing when a machine can reason for itself. I mean, imagine if you're toaster were able to recognize when it had burnt the toast? From there making a hypothesis on how to improve, testing the hypothesis, and seeing the results. I mean, as small as this is, your toaster would always be able to give you your toast exactly as you like it.
In a larger spectrum, Automated Cars (like those in the movie "iRobot") are closer than ever. If the car could reason for itself how to accelerate and when to slow down, where to turn, it would be amazing.
I say, Kudos to those researchers, they are trying their hardest and it's paying off. We may only see a robot that does tests of yeast now. But in the future, we may see battleships that can fly themselves, and make the best judgment call, not like the ones shown in movie (where the robot enslaves mankind for the protection of mankind) but rather, where the computer can reason and see just as a human does, what is the best choice.
Not only could this mean perfect toast. It could mean the saving of thousands of innocent lives, those who fight for us now on the front lines. If we could replace them with robots, the world would be a whole lot happier.
Think, before you speak, that when you are making your jeers at these researchers, that you are critisizing those who would rescue those distraught mothers from the tears they will have when they hear that their only son died in the service of his country.
Goodluck with Eve.
ghendric
7 months ago
6 comments
If the scientists already know the answer that this robot is going to come too, then why do we need the robot?
Leave to the Brits to come up with this. :) They're always trying to make their software "alive"! lol... Maybe they should have named it Johnny 5.. lol..
cthenkhaus
7 months ago
102 comments
I like the idea and it's good to see the world moving forward but where's John Conner when you need him?
Nlazouzi
7 months ago
2 comments
Hmmmm. Something just doesn't feel right about all this. it sounds to me like a bunch of high expectations only. FORMULATE A HYPOTHESIS ON ITS OWN ????? ARE YOU KIDDING ME. NO WAY
rahul_rbp
7 months ago
14 comments
its really interesting. i am just thinking how is that in real..