Robot Makes Scientific Discovery All by Itself
- By Lizzie Buchen
- April 2, 2009 |
- 12:15 pm |
- Categories: Tech
For the first time, a robotic system has made a novel scientific discovery with virtually no human intellectual input.
Scientists designed “Adam” to carry out the entire scientific process on its own: formulating hypotheses, designing and running experiments, analyzing data, and deciding which experiments to run next.
“It’s a major advance,” says David Waltz of the Center for Computational Learning Systems at Columbia University. “Science is being done here in a way that incorporates artificial intelligence. It’s automating a part of the scientific process that hasn’t been automated in the past.”
The demonstration of autonomous science breaks major ground. Researchers have been automating portions of the scientific process for decades, using robotic laboratory instruments to screen for drugs and sequence genomes, but humans are usually responsible for forming the hypotheses and designing the experiments themselves. After the experiments are complete, the humans must exert themselves again to draw conclusions.
Meanwhile, some software programs can analyze data to generate hypotheses or conclusions, but they don’t interact with the physical realm. Adam is the first automated system to complete the cycle from hypothesis, to experiment, to reformulated hypothesis without human intervention.
Adam’s British designers, led by Ross King at Aberystwyth University in Wales, acknowledge that the robot’s discoveries have been “of a modest kind” thus far. Its proving ground as a scientist has been the genome of baker’s yeast, a popular laboratory species. Baker’s yeast is one of the best understood organisms, but 10 to 15 percent of its roughly 6,000 genes have unknown functions. The scientists hoped Adam could shed light on some of these mystery genes.
They armed Adam with a model of yeast metabolism and a database of genes and proteins involved in metabolism in other species. Then they set the mechanical beast loose, only intervening to remove waste or replace consumed solutions. The results appear Thursday in Science.
Adam sought out gaps in the metabolism model, specifically orphan enzymes, which scientists think exist, but which haven’t been linked to any parent genes. After selecting a desirable orphan, Adam scoured the database for similar enzymes in other organisms, along with the corresponding genes. Using this information, it hypothesized that similar genes in the yeast genome may code for the orphan enzyme.
The process might sound simple — and indeed, similar “scientific discovery” algorithms already exist — but Adam was only getting started. Still chugging along on its own, it designed experiments to test its hypotheses, and performed them using a fully automated array of centrifuges, incubators, pipettes, and growth analyzers.
After analyzing the data and running follow-up experiments — it can design and initiate over a thousand new experiments each day — Adam had uncovered three genes that together coded for an orphan enzyme. King’s group confirmed the novel findings by hand.
Waltz thinks Adam will inspire other scientists. “They’ll realize they can automate more of the process than they currently have. They can explore a wider range of possibilities without doing it all by hand.”
King is already expanding his Robot Scientist fleet by producing Eve, which will autonomously design and screen drugs against malaria and schistosomiasis.
“Most drug discovery is already automated,” says King, “but there’s no intelligence — just brute force.” King says Eve will use artificial intelligence to select which compounds to run, rather than just following a list.
If robotic scientists made their way into other labs, their human counterparts would not be out of a job anytime soon. If anything, they may find their work more exciting.
“There may be teams of humans and machines,” says King. “Robots will be doing more and more of actual experimental work and simple cycles of hypothesis generation. Humans would migrate to more strategic and creative positions. How can we waste trained post-docs by making them pipette things in labs? It’s crazy.”
But with advances in artificial intelligence, it’s conceivable that the role of robots would, in the more distant future, creep deeper into the human realm, progressing from lab technician to lab head. Robots may even be capable of performing supposed acts of genius, such as Einstein’s conception of special relativity.
“There isn’t any intrinsic reason why that wouldn’t happen,” says King. “I think there’s a continuum between the really basic types of science that you’d get from Adam, and the things I can do, and then Einstein-type science. A computer can make beautiful chess moves, but it’s not doing anything special. It’s just doing more of the same thing. In my view that’s what’s going to happen in science.”
King may already have a head start: Deep Blue could never have beaten Garry Kasparov without engineer Feng-Hsiung Hsu moving the pieces on its behalf.
See Also:
- Autonomous Robots Invade Retail Warehouses
- I, Nanny: Robot Babysitters Pose Dilemma
- “Autonomous” Heart Surgery Robot Mystery Solved
- Military AI Could Rule the Internet
- Humans No Match for Go Bot Overlords
- Biologists on the Verge of Creating New Form of Life
Image: Jen Rowland
The entire premise of this article is a load of SH**.
The robotic systems did NOT do anything that it was not PROGRAMMED to do by HUMANS!
The novelty may be in the number of tasks it was programmed to do in sequence without intervention by programmer/engineer/other personnel.
It did not make any decisions that were not coded in some form of decision tree which was part of a programmed sequence.
Boo!
HERE COMES JUDGEMENT DAY!
you might want to read a book on machine learning, its slightly more complex than a bunch of if statements.
I GEN,
I don’t think you understand how we learn, or how we program (or rather, how AI could work). Are we not all the product of those who taught us? Don’t we have a finite set of capabilities and behaviors?
SKYNET OH SHI–
Why are we compelled to build machines that do our thinking for us? If these people find thinking so painful why did they take up careers in science? Measuring out reagents is not a waste of time!! Regardless of whether ” I GENERATED THIS COMMENT ALL BY MYSELF ” was right, this “advance” is one we don’t need.
I for one am not looking to replace myself with a machine. If these robot-builders want to make a real contribution they could train computers to recognize dirty glassware and clean off every last spot.
I, for one, welcome our new robot overlords.
(It had to be said)
From what I hear, the thing always breaks down and never works. It’s only managed that one thing and that aint that scientific
This was a triumph.
I’m making a note here–huge success!
I guess I better start learning the robot. And the robo boogie.
cylons were created by man..
Anyone know if 556 rounds will punch through a robot body? This is info that may come in handy soon.
I call bullshit.
not just yet but the progress here is interesting. keep running the headline every year and it will keep getting a little more accurate.
The major disadvantage of computer-directed science will be the inability to exploit accidental discoveries. A human looking for one thing can discover another and retask their research to pursue a different application. A computer will see that what it’s found doesn’t fit what it’s looking for and just throw it away.
I’m doing science and I’m still alive.
It isn’t necessary for the robot to think. Even though I Gen is right, and people like Warpig, who don’t understand that any type of “free will” within the digital realm are wrong, the scientific benefits that can be achieved by automated trial and error are enormous. Many breakthroughs are achieved by serendipity, and machines can be made to try all sorts possible combinations of things like chemical formulas. Sooner or later, the robot could stumble upon something useful. Again, this requires no thought–only the following of directions. Humans can make up directions, digital machines cannot.
There is also a new Robot called Eve that comes up with new ways to nag people.
m3kt3k,
A couple of .308 Winchester rounds should stop the robots when the time comes to put them down.
… And it has a plan.
I need a plasma rifle in a 20watt range
I-GEN just realized he can’t do anything he was not PROGRAMMED to do by GOD! Or maybe he thinks humans are different because we have access to THE FORCE.
Impressive work, but…
“For the first time, a robotic system has made a novel scientific discovery with virtually no human intellectual input.”
Well, maybe. Some would point to Doug Lenat’s work, for example ( http://en.wikipedia.org/wiki/Doug_Lenat ).
This is a really exciting and impressive development. Just paving the way for more rapid advances in scientific knowledge.
(no I’m not going to address the idiots above, I don’t waste my time arguing with morons on the internet anymore)
@Rodney Hoffman Thanks for the comment. I guess it depends on how you define a *scientific* discovery. AM and Eurisko selected problems and discovered novel mathematical relationships, but this isn’t a “knowledge-based” activity like science (where more data doesn’t mean more knowledge). There’s also Gary Livingston at UMass Lowell whose robot autonomously ran protein crystallography experiments, but they were all in silico. People have been working on various pieces of the puzzle for decades–Adam fit them together.
‘Virtually (no human intellectual input)’ is not the same as ‘all by itself’, as in unaided.
Chuck Brooks
FutureWare SCG
If these things were around when I was a grad student I might not have quit.
Science research requires smart people to pay close attention to a lot of monotonous tasks (like measuring reagents). Handing this off to a machine frees the scientist up to do the interesting parts and solve the big & important questions more quickly and accurately.
From a Concerned Grad student who posted this article on Facebook:
“Maxime Desmarets” OK that’s it: I’m out of the job that I don’t even have yet. He’s the perfect Grad Student: he works 24 hours a day and on week ends, doesn’t procrastinate on facebook half of the time. How can I compete???”
Wow… just wow. This is why I love science and technology. Also, I think most of the people commenting don’t really understand the scope of this. You know, this isn’t a machine running a macro and if/else statements towards a unique fixed goal. This robot has full blown AI making decisions and formulating questions, then using tools available to it to answer those questions and repeating the process… it learns as it discovers! As Lizzie said, Adam fits many pieces together for the first time.
This just inspired me to work harder on my code.
@rodney..I wouldnt go believing everything you read on wikipedia…
@Danel..you did address the idiots above..
I think its a huge break through and this will help advance science in ways we mere techo-consumers dont even realize.Isnt this why we build machines? its the same job, just using a new tool, when has the human race not benifited from that?
work smarter, not harder….
Cant wait for them to build “Deep Thought”
Oh Yeah? Just wait till the damn thing decides in the middle of the night that humans are dull, myopic and needlessly overconsumptive – and tells its counterparts…THEN things will get interesting.
THEN, you can kiss my shiny metal ass.
Why is everybody so scared of scentient robots?
All you have to do is run up some stairs or around the courner and the robot will get stuck on the edge, do to pathing issues.
No matter how many “scientific discoveries” these robots make, does it count until a conscious entity (for now that means us) says that it is a scientific discovery? Does an elephant with a paintbrush tied to its tail make art only if we say that it did? This all makes an interesting extension of the discussion about trees that fall in the forest. I would lean toward the side of the debate that says that consciousness is a requirement, but like I said, that’s an interesting talk to have.
5.56 steel core will do the job but you better have big clips… 308 is better unless skynet has already begun production of advanced metallurgy by sending t-1000’s back. in that case the best thing to do is keep them from getting firearms or access to raw materials needed to manufacture ammunition. at that point we can just ram trucks into them. big trucks.
The funny thing is when it all works together, humans still are really good at figuring out new things, in the space we have.
These robots are still bound by the laws of physics when it comes to running these experiments, so really they’re not much faster than us, and until we program a method of ‘curiosity’ into the devices, they’re pretty much going to hit their ACTUAL permutation limit, basecode^environmentals, and if that curiosity isn’t there, they are going to fail to dream.
The other creatures on earth that have decent brain complexity are also in this class, except their ability to permute their environment is way less, so it happens less quickly.
This + the natural laws post makes for an interesting next 15 years that’s for sure. beginning of the end of something.
my improvised shotgun detonated EMF gun will make quick work of any robots headed my way. Bring em on!
“For the first time, a robotic system has made a novel scientific discovery with virtually no human intellectual input.”
I once went to Lowes to get a gallon of paint mixed and the guy, using no human intellect working the paint mix robot and I watched as the robot created a novel color of paint that I don’t think has ever existed before. The Lowes robot was the first.
Folks, it’s an April fools joke. It happens every once in a while in the scientific literature – to see if you can think independently, or if you just believe whatever you read because it’s Science or Nature or Cell.
I’ve been sayin’ it for years. Give the machines the _capacity_ and they’ll be able to do as much if not more than humans. And when I say capacity I mean literally the capacity to store massive amounts of information. Coupled with the ability to seek out _new_ information and you’ve got…A human! Teach it a few basic things and soon it’ll be off solving all our problems.
Ive heard about that King bloke – hes good at bigging himself up over tiny science. pretty sure Nature rejected these findings – King probably made them up. too desperate for fame and fortune
And for my next trick I’ll perform the Turing test autonomously …
I for one welcome our robot overlords. May death come quickly to their enemies.
Dear I GEN,
At the risk of being slightly insensitive, you don’t need an AI to deduce you are a buck-toothed hillbilly. There are ample examples of software, written by humans, that do unexpected and surprising things.
I suspect your family tree does not branch to any great degree but, for the rest of us, are we not the results of millions of years of random mutations, unexpected and surprising mutations?
Would it not be wonderful if this machine did something unexpected and surprising perhaps discovering something outright or at least putting scientists on the path to greater things?
– R. Daneel Olivaw
The brain is not a hood ornament,
The advantage of a thinking robot is that it can make these decisions much quicker than a human and can do them around the clock.
It’s a time saver, not so much a brain saver.
This is a viral article for the new Terminator movie.
Yay, Twister…may it live long and help discover much!
I Think It’s Really Creepy That Artificiality Controlled Objects Are Learning Things Almost The Way We Do!
Then Skynet became self-aware.
[...] logic computers are going to end up doing it. In some cases, they already are1. Excerpts from Robot makes discover all by itself (click for full [...]
[...] Robot Makes Scientific Discovery All by Itself [...]