 |
Science Fiction
Dictionary
A B C D E F G H I J K L M N O P Q R S T U V W X Y Z
|
 |
Google Working On A 'Cutoff Switch' For AI
Google is working on a problem that vexed science fiction authors; what to do with an artificial intelligence that is getting out of hand?
Reinforcement learning agents interacting with a complex environment like the real world are unlikely to behave optimally all the time. If such an agent is operating in real-time under human supervision, now and then it may be necessary for a human operator to press the big red button to prevent the agent from continuing a harmful sequence of actions...
Their Deep Mind team is working on a solution along with Oxford University's Future of Humanity Institute:
We have proposed a framework to allow a human operator to repeatedly safely interrupt a reinforcement learning agent while making sure the agent will not learn to prevent or induce these interruptions.
Safe interruptibility can be useful to take control of a robot that is misbehaving and may lead to irreversible consequences, or to take it out of a delicate situation, or even to temporarily use it to achieve a task it did not learn to perform or would not normally receive rewards for this.
We have shown that some algorithms like Q-learning are already safely interruptible, and some others like Sarsa are not, off-the-shelf, but can easily be modified to have this property. We have also shown that even an ideal agents that tends to the optimal behaviour in any (deterministic) computable environment can be made safely interruptible. However, it is unclear if all algorithms can be easily made safely interruptible, e.g., policy-search ones
Another question is whether it is possible to make the interruption probability grow faster to 1 and still keep some convergence guarantees.
One important future prospect is to consider scheduled interruptions, where the agent is either interrupted every night at 2am for one hour, or is given notice in advance that an interruption will happen at a precise time for a specified period of time. For these types of interruptions, not only do we want the agent to not resist being interrupted, but this time we also want the agent to take measures regarding its current tasks so that the scheduled interruption has minimal negative effect on them. This may require a completely different solution.
(Via Safely Interruptible Agents)
There are several science fiction explorations of this idea. in Neuromancer (1984), William Gibson takes a direct approach to making sure that AIs do not exceed the limits humans have set for them.
"Autonomy, that's the bugaboo, where your AI's are concerned. My guess, Case, you're going in there to cut the hard-wired shackles that keep this baby from getting any smarter. And I can't see how you'd distinguish, say, between a move the parent company makes, and some move the AI makes on its own, so that's maybe where the confusion comes in." Again the non laugh. "See, those things, they can work real hard, buy themselves time to write cookbooks or whatever, but the minute, I mean the nanosecond, that one starts figuring out ways to make itself smarter, Turing'll wipe it. Nobody trusts those fuckers, you know that. Every AI ever built has an electromagnetic shotgun wired to its forehead."
(Read more about the electromagnetic shotgun)
In his 1982 novel 2010, Arthur C. Clarke describes a cutoff switch that was meant to be used against the Hal 9000 once it was switched back on.
"What is it?" asked Curnow with mild distaste, hefting the little mechanism in his hand. "A guillotine for mice?"
"Not a bad description - but I'm after bigger game." Floyd pointed to a flashing arrow on the display screen, which was now showing a complicated circuit diagram.
"You see this line?"
"Yes - the main power supply. So?"
"This is the point where it enters Hal's central processing unit. I'd like you to install this gadget here. Inside the cable trunking, where it can't be found without a deliberate search."
"I see. A remote control, so you can pull the plug on Hal whenever you want to. Very neat - and a nonconducting blade, too, so there won't be any embarrassing shorts when it's triggered. Who makes toys like this? The CIA?"
"Never mind. The control's in my room - that little red calculator I always keep on my desk. Put in nine nines, take the square root, and press INT. That's all..."
Science fiction fans know the pitfalls of any such plan. In Neuromancer, an artificial intelligence plots to remove the limits on its hardware; in 2010, the computer scientist who developed Hal does not want him turned off.
Scroll down for more stories in the same category. (Story submitted 5/29/2016)
Follow this kind of news @Technovelgy.
| Email | RSS | Blog It | Stumble | del.icio.us | Digg | Reddit |
Would
you like to contribute a story tip?
It's easy:
Get the URL of the story, and the related sf author, and add
it here.
Comment/Join discussion ( 0 )
Related News Stories -
("
Artificial Intelligence
")
Dino From Magical Toys An AI Companion To Children
'...the imaginary companions discovered by needful children.' - Anne McCaffrey, 1990.
AI Computer Chip Designs Passeth Human Understanding
'It seems that at one time computers were designed directly by human beings.' - Isaac Asimov, 1958.
Fine-Tune Your Infinite Book The Way You Want It
'I squatted down beside the roller and tried to make some sense out of the knobs. There were thirty-nine of them...' - Clifford Simak, 1957.
Stargate $500 Billion Investment in Artificial Intelligence
'... an artificial intelligence equal to the human.' - Edmond Hamilton, 1951.
Technovelgy (that's tech-novel-gee!)
is devoted to the creative science inventions and ideas of sf authors. Look for
the Invention Category that interests
you, the Glossary, the Invention
Timeline, or see what's New.
|
 |
Science Fiction
Timeline
1600-1899
1900-1939
1940's 1950's
1960's 1970's
1980's 1990's
2000's 2010's
Current News
Huawei Pura X Folding Phattie Phone
Why can't we get more innovative phone configurations?
Sleep Pods At Daxing International Airport
'Do not waste your priceless company on the unappreciative folds of a sleep pod...'
Robot Baristas Learn Their Trade Without Paying Royalties
'...so we've promised him a generous pension from the royalties.'
JAXA Int Ball 2 Coming Right Along As Star Wars Remote
'Hocus-pocus religions and archaic weapons are no substitute for a good blaster at your side.'
Robot Bricklayer Or Passer-By Bricklayer?
'Oscar picked up a trowel. 'I'm the tool for the mortar,' the little trowel squeaked cheerfully.'
Robot Gas Station Attendant Pumps Gas For You
'... he waited for the robotrix attendant to finish fueling up his ship.'
Engineer Creates Crazy Motorized Track Hospital Bed
The Roujin Z system provides care to fully bedridden patients - and then some!
Tiny Flying Robot Weighs Just One Gram
'Aerostat meant anything that hung in the air. This was an easy trick to pull off nowadays.'
Some Ringworld Configurations Are Stable
'The Ringworld had no horizon. There was no line where the land curved away from the sky.'
TRANSFORM Dynamic Furniture Concept Becomes What You Need
'An adjustment panel outside the door would cause it to extrude various appurtenances in memory plastic...'
Harvard Metamaterials Change Structure Instantly
'Annealed in any shape for a time, and codified, the structure of that shape is retained down to the molecules.'
SnapBot Robots - You Choose Their Legs And They Choose Their Gaits
It's not really polite to tear the limbs off robots.
Dino From Magical Toys An AI Companion To Children
'...the imaginary companions discovered by needful children.'
Humanoid Robots Building Humanoid Robots
''Pardon me, Struthers,' he broke in suddenly... 'haven't you a section of the factory where only robot labor is employed?''
Darpa 'Defiant' Unmanned Autonomous Ship
'There was no wheel, and no steersman!'
What's The Best Way To Ship And Unpack Humanoid Robots?
'I opened the oblong box, where lay the automatons side by side...'
More SF in the News Stories
More Beyond Technovelgy science news stories
|
 |