ai
Latest
EATR robots are coming, this isn't funny anymore
Oh sure, we joke about rogue AI all the time, and we're aware that we'll probably pollute ourselves to death well before the robots get us, but who really thinks flesh-eating machines are a good idea? The (patently evil) scientists behind the EATR project -- no fair, they're making their own jokes now too -- have reached a new milestone in the development of the reconnaissance bot, successfully coupling a steam generator with a compact biomass furnace. It is now therefore possible for an autonomous machine to forage for and refuel itself with biomatter, otherwise known as soft, pulsating, yummy humans. They call it fuel versatility, as gasoline, diesel, and solar power may also be used if available, yet we'll offer no prizes for predicting which energy source these chainsaw-equipped robots will prefer. [Via Switched]
Researchers develop a robot that reads your intentions, says you're 'thick'
Robots won't be able to wrest control of the planet from us silly humans until they learn how to collaborate. Sure, they can mow the lawn or mix a drink, but only when you give 'em explicit instructions. Luckily for our future robot overlords, The EU's JAST project is studying the ways that humans work together, in the hope that it can someday teach robots to anticipate the actions and intentions of a human partner. "In our experiments the robot is not observing to learn a task," explains Wolfram Erlhagen from the University of Minho. "The JAST robots already know the task, but they observe behavior, map it against the task, and quickly learn to anticipate [partner actions] or spot errors when the partner does not follow the correct or expected procedure." This bad boy has a neural architecture that mimics what happens when two people interact, and the video below shows the rather melancholy automaton trying to convince his human partner to pick up the right pieces to complete a simple task. Watch it in action after the break.
IBM's Watson to rival humans in round of Jeopardy!
IBM's already proven that a computer from its labs can take on the world's best at chess, but what'll happen when the boundaries of a square-filled board are removed? Researchers at the outfit are obviously excited to find out, today revealing that its Watson system will be pitted against brilliant Earthlings on Jeopardy! in an attempt to further artificial intelligence when it comes to semantics and searching for indexed information. Essentially, the machine will have to be remarkably labile in order to understand "analogies, puns, double entendres and relationships like size and location," something that robotic linguists have long struggled with. There's no mention of a solid date when it comes to the competition itself, but you can bet we'll be setting our DVRs whenever it's announced. Check out a video of the progress after the break.[Via The New York Times]
Fallen Earth dev diary focuses on introducing new players to PvP
Fallen Earth is one of three post-apocalyptic massively muliplayer online games currently in development that are a far cry from the fantasy titles which have proven most popular in the MMO world. IGN scored an exclusive two part developer diary from Fallen Earth writer and content developer Wes Platt who discusses creating the PvP starter town of Terance. Namely, he explains how and why the Fallen Earth team has been putting so much work into Terance and the challenges and pitfalls faced in differentiating the PvP-centric area from other more standard towns in the game. The first part is "Building the Town of Terance". It paints a picture of a post-apocalyptic aftermath setting where a psychotic artificial intelligence, long since sealed away underground by its corporate progenitors and forced into a century of dormancy, is woken with dire consequences. Now powered up, the AI -- TETRAX -- prepares once again to work towards the extermination all human beings in its vicinity. Human beings in Terance may find themselves on the run, hunted by AI-directed zombies called Diggers, as well as mutants and vermin.
Artificial Intelligence solves boring science experiments, makes interns obsolete
Researchers at Aberystwyth University in Wales have developed a robot that is being heralded as the first machine to have discovered new scientific knowledge independently of a human operator. Named Adam, the device has already identified the role of several genes in yeast cells, and has the ability to plan further experiments to test its own hypotheses. Ross King, from the university's computer science department, remarked that the robot is meant to take care of the tedious aspects of the scientific method, freeing up human scientists for "more advanced experiments." Across the pond at Cornell, researchers have developed a computer that can find established laws in the natural world -- without any prior scientific knowledge. According to PhysOrg, they've tested the AI on "simple mechanical systems" and plan on applying it to more complex problems in areas such as biology to cosmology where there are mountains of data to be poured through. It sure is nice to hear about robots doing something helpful for a change.[Thanks, bo3of]Read: Robo-scientist's first findingsRead: Being Isaac Newton: Computer derives natural laws from raw data
GDC09: Havok gets smart, announces Havok AI
Havok is exploding like a red barrel into the world of artificial intelligence. With ragdoll grace, the middleware company announced its new Havok AI SDK during GDC 2009, promising "unique solutions" to various AI pathfinding issues faced by today's game developers. Like the folks at 1UP, we're not exactly sure what this means for gamers, though the new software is supposed to be fully compatible with Havok's other products and tools, such as Havok Physics. Perhaps now enemy patrols will get the good sense not to seek shelter behind things that go kablooey when bullets begin to fly.[Thanks David B.]
Brown University, DARPA give iRobot's PackBot autonomy
It's not easy to find research in the field of robotics without military applications (or military funding), and Brown University's latest is certainly no exception. Starting out with iRobot's PackBot (and some pocket change from DARPA and the Office of Naval Intelligence) researchers at the school have achieved several advances that will someday produce robots that follow both verbal and nonverbal commands from a human operator, indoors and out, without the need for a controlled environment or special clothing. The goal, according to Chad Jenkins, is to develop a robot that acts "like a partner. You don't want to puppeteer the robot. You supervise it, 'Here's your job. Now, go do it.'" The work is being presented this week at the Human-Robot Interaction conference in San Diego, but if you can't make it we've provided a video of the thing in action just for you (after the break). We for one salute our autonomous robot overlords.[Via PhysOrg]
Massively's Apocrypha expansion hands-on: The Sleepers
The Sleepers The inhabitants of these uncharted solar systems may take offense to your encroachment upon their territory, however. The Sleepers, an ancient race of NPCs known for their mastery of virtual reality and cryogenics, will provide the greatest PvE challenges players have ever faced in EVE Online through their guardian drones. Their AI is far beyond what players are used to going up against in PvE. The Sleepers will have varying levels of strength and adaptation to player threats. They do seem to have a particular hatred for their creators at CCP, given their ultra-violent response to Ward's arrival at a structure the Sleeper drones constructed in space. They move in on him, dishing out *all* damage types: beams sizzle for EM and Thermal damage; warheads obliterate for Kinetic and Explosive damage. Fortunately Ward's Proteus is set up to deal with this ("I'm going to put on a GM shield extender, or when we go through there I'm going to get wasted!") making his ship virtually impossible to kill, letting us witness the Sleeper offensive in safety.
Platinum Games discusses enemy AI in MadWorld
Bless the Platinum Games blog, as it's been a wonderful source of interesting stuff leading up to next week's release of MadWorld. Today, Hirono Sato takes over the driver's seat and talks about his role in the game: player interactions and the three different AIs governing enemies. He breaks the enemy units down to "grunt," "grunt leader," and "boss." Sato explains that the "grunts" are essentially pushovers and the challenge they present is not simply just to defeat them, but to defeat them in the most stylish way possible, for the most points possible. Sato says that in a game "where running around killing all the enemies is supposed to be fun, making them so hard that you can't kill them wouldn't be fun at all."However, for those that crave difficult combat, the "grunt leader" is always on hand to provide an ample spanking. This guy is a bit tougher to battle, so it's not as much about getting the most points possible as it is about just surviving his attacks. Then, there are the "bosses," which are pretty self-explanatory.Sato also tells us of a cool in-game item called the "Money Grubber," a briefcase stuffed with money that you throw at enemies. Once the "grunts" see it, they'll start clawing for it, and eventually fight each other over the money. While they're busy, Sato offers a few ideas, including tossing "a drum filled with gasoline their way" and even tossing the case "onto some busy train tracks." %Gallery-22964%
Gesture recognizing QB1 computer attends to your every desire
Scouting a computer that's "attentive to one's desires?" Good news, friends! Frédéric Kaplan's QB1, which was unveiled at the LIFT Conference in Geneva this past week, aims to be just that. Reportedly, the machine was designed in order to "alter the fundamentals of human-machine interaction," and rather than relying on the traditional mouse and keyboard approach, this one works entirely via gestures. QB1 is capable of recognizing inputs from both hands at once, with one example having a human select a record and adjust the volume by simply flicking their fingers through an on-screen album collection. We're told that the related patents behind the sophisticated 3D gesture interaction technology have been filed, but there's no word yet on when we'll be able to actually buy one. 'Til then, it's up to you to handle those "desires" yourself.[Via The Inquirer]
EVE Online developer explains new AI for the Sleeper race
One of the major new features to EVE Online's forthcoming Apocrypha expansion is the introduction of an ancient race of NPCs called the Sleepers. They're a breed apart from any NPCs ever seen in EVE, largely because of their AI. They react intelligently to threats, focusing fire on primary targets but diverting their attacks to counter whatever else they're faced with. They can 'spider tank' or protect and even repair one another as combat ensues. Added to their tactics is evasive maneuvering, making the Sleepers even deadlier. We recently showed you video footage of the Sleepers in action and while taking them on will, in some cases, be a daunting experience, this is not to say that they can't be defeated. EVE developer CCP Incognito wrote on the forums, "If you try to use the same-old, same-old tactics against Sleepers then you will have problems. Think out of the box and you will win." CCP Incognito's comments were in the context of his dev blog on the new AI, which discusses the challenges that the Sleepers will pose for players seeking fragments of their advanced technology. More than anything, the AI revamp is designed to make PvE more like PvP. Anyone planning to venture through EVE's wormholes should abandon their mission-centric ship setups and be prepared for PvP; that's what fighting the Sleeper NPCs will be like -- fighting other players.
Navy report warns of robot uprising, suggests a strong moral compass
You know, when armchair futurists (and jive talkin' bloggists) make note of some of the scary new tech making the rounds in defense circles these days it's one thing, but when the Doomsday Scenarios come from official channels, that's when we start to get nervous. According to a report published by the California State Polytechnic University (with data made available by the U.S. Navy's Office of Naval Research) the sheer scope of the military's various AI projects is so vast that it is impossible for anyone to fully understand exactly what's going on. "With hundreds of programmers working on millions of lines of code for a single war robot," says Patrick Lin, the chief compiler of the report, "no one has a clear understanding of what's going on, at a small scale, across the entire code base." And what we don't understand can eventually hunt us down and kill us. This isn't idle talk, either -- a software malfunction just last year caused US. Army robots to aim at friendly targets (fortunately, no shots were fired). The solution, Dr. Lin continues, is to teach robots "battlefield ethics... a warrior code." Of course, the government has had absolutely no problems with ethics over the years -- so programming its killer robots with some rudimentary values should prove relatively simple.
IBM develops computerized voice that actually sounds human
If there's one thing that still grates our nerves, it's automated calling systems. Or, more specifically, the robotic beings that simply fail to understand our slang and incomprehensible rants. IBM's working hard and fast to change all that, with a team at the company's Thomas J Watson research division developing and patenting a computerized voice that can utter "um," "er" and "yes, we're dead serious." The sophisticated system adds in the minutiae that makes conversation believable to Earthlings, and it's even programmed to learn new nuances and react to phrases such as "shh." The technology has been difficulty coined "generating paralinguistic phenomena via markup in text-to-speech syntheses," and while exact end uses have yet to be discussed publicly, we can certainly imagine a brave new world of automated CSRs.
EVE Online to introduce new NPC race with Apocrypha expansion
PvE gameplay in EVE Online can be profitable, but blasting apart NPC ships can also get repetitive after a while. Anyone who's run a solid week of missions knows this all too well, and the combat dynamic between players and NPCs is something CCP Games will start remedying in the next game expansion: Apocrypha.EVE dev CCP Ytterbium's dev blog "When Evolution Leaps Forward" has some big news about what's in store for players as of March 10th in Apocrypha. The devs tasked with revamping the game's PvE experience are making some substantial changes to the distribution of bigger and badder NPCs. They're also introducing a new NPC race to New Eden: the Sleepers, an ancient race with technology beyond what is currently available to players, but whose secrets can potentially be unlocked.
Basil the Robot trained for symbolic recognition, beer toting
Though not much of a looker -- feel free to insert your own beer goggles joke here -- Basil the Robot is an experiment by Jim and Louise Gunderson to train an AI to identify its surroundings symbolically. That way, the couple hopes, he can react dynamically in new environments. Basil was intended to be shown off at a Cafe Scientifique meeting by having him go to the bar and order drinks for his creators, but that proved unsuccessful when Basil's battery died. The couple did videotape a successful trial run from the night before, which you can see after the break. Basil will next receive a microphone for voice commands and be upgraded from his current sonar navigation to a video sight system. The eventual goal is to teach Basil to go to the local brewery and pick up beer. Sure, we're still terrified of the robot revolution, but let's get serious here -- who are we to deny our mechanical overlords if they keep proffering us six-packs? Hit up the read link for the full story.[Via Metafilter]
MMOGology: Braindead
You tiredly trudge up the cliff-side path, the rhythmic squish of your soaked boots beating a slow dirge. Cresting the ridge of the trail you see the full moon shining over Bloodstone's harbor where several galleons rot in their moorings. It almost looks like a peaceful town from up above the sweat and fish-soaked stench of the city. You sigh. It's been an exhausting trip dodging banshees and hollow men through the slime of Wraithmarsh. All you can think about is a mug of ale, a warm bed in the inn and perhaps a little company to take your mind off things. Bloodstone is known for its "hospitable" women after all. With a renewed sense of purpose you pat your faithful dog on the head and take the sloping trail down into town. As you reach the outskirts of Bloodstone the rancid smell of a fish merchant's stall nearly slaps your nose off your face. You vainly wave your hand to clear the air."Wot's a matta gov?" the merchant asks, sneering. "Can't 'andle a little fresh fish?""That fish is as fresh as my feet," you reply. The short tempered merchant draws a rusty cutlass and grimaces. Several ruffians milling about sense an impending fight and begin to circle you."If ya don't like it, you and that mangy dog can bugger off!" he responds, kicking your dog in his hinder. Your dog whimpers and sits next to you, tail between his legs. The bolder members of the encircling crowd brandish knives and mock you and your dog. They see a tired traveler and easy pickings. As tired as you are, you can't help but laugh to yourself and shake your head. Your eyes begin to glow a soft white hue. The humid air of Bloodstone begins to crackle. Time to teach these lowlifes a lesson.
DARPA enlists IBM to build computer brain as smart as a cat
Researchers have long been trying to model actual brains in order to build a better computer "brain," and it looks like IBM is now getting a helping hand from none other than DARPA in its attempt to create one that it hopes will one day have the intelligence level of a cat. To that somewhat unnerving end, DARPA is pouring $4.9 million into a project that'll include five universities and scientists of all stripes, who will work together to create an artificial brain that behaves like a real one right down to the neuron level. As the BBC reports, the researchers are describing this latest initiative as a "180 degree shift in perspective" from previous efforts, as they're now seeking an algorithm first and problems second, as opposed to starting with an objective and devising an algorithm to achieve it. As for DARPA's ultimate goal, well, that's still a bit of a mystery, though let's just say we won't be surprised if future robots start to become very easily distracted. [Via Daily Tech, image courtesy Mack J, Truth and Beauty Bombs]
The Army's Artificial Intelligence invades WoW
Joe Martin at bit-tech.net picked up an article on Gizmodo talking about the coming invasion of Army Artificial Intelligences masquerading as real players in World of Warcraft. According to Dr. John Parmentola, the plan is to test the AI's ability to be a "fake" human by letting it interact with real humans in a virtual world.My first reaction was, "Whoah, cool. All your base are belong to us." But after a moment's thought, this might not actually be such a great idea. Given the communication skills of some players (especially in the battlegrounds), I'm not seeing this as a litmus test of what in-game speech can pass for spoken by real people. While I'm pretty sure the AI won't communicate like a roleplayer, the AI could probably get by with a series of "lol" and "kek" typed out in rapid succession.This isn't the first time we've heard about the military using WoW (or WoW-like systems) for training purposes, which is the nominal purpose of this new AI research. Maybe it won't be too long before we're logging in to have a Gnome Rogue named Joshua quietly whisper us, "Shall we play a game?"
A closer look at Elbot's Turing test conversation
Earlier this week, Elbot made a fairly impressive showing (comparatively speaking, at least) when fooling three judges into thinking it was human; had it fooled one more on the dozen deep panel, it would have successful passed the famed Turing test. Auntie Beeb now has a report on what exactly Elbot said when asked a litany of questions away from the competition, and there's also a video with the related experts dissecting its performance. To be totally honest, its responses weren't too far from being completely passable as ones from a tired, potentially inebriated Earthling (in our humble opinion), but we'll leave the final determination to you. Touch the read link for a one-on-one with ones and zeros.
New round of Turing test fails to crown a winner
While some folks are considering taking the Turing test one step further and applying it to military robots, a group of researchers in the UK led by none other than would-be cyborg Kevin Warwick are doing their best to keep things as Turing intended and simply trying to fool some humans into thinking that the robot they're taking to is actually a person. Fortunately for us on the human side of the equation, they weren't quite successful, though one "robot" known as Elbot did get relatively close to the goal, fooling 25% of its human interrogators, which is just 5% off the mark set by Alan Turing. Each of the four other "artificial conversational entities" also managed to fool at least one of their questioners, though they eventually showed their true colors with random answers like "soup" when pressed as to what their job was.
