AI
Latest
Japan's space agency considers using rockets with artificial intelligence
The keyword here is obviously "considers," but it looks like Japan's space agency, JAXA, is indeed seriously thinking about using artificial intelligence to improve their rocket launches. As JAXA scientist Yasuhiro Morita explains, as opposed to simply being "automatic" as rockets are today, an "artificially intelligent" rocket would be able to keep watch on its condition, determine the cause of any malfunction, and potentially even fix it itself. According to JAXA, that would not only make rocket launches more efficient, but more cost-effective as well given the reduced manpower needs. That's not the only new measure being explored to cut costs, though -- as Space.com reports, JAXA's new Epsilon launch vehicle is also being built using fewer, but more advanced components, which promises to let it be moved to the launch pad nearly fully assembled. It's currently set to launch sometime in 2013, although it's not yet clear how much it will actually be relying on AI if such a system is put in place.
Massively's exclusive TERA lore: Day in the Life - Kumas
Yesterday we got a glimpse of two new BAMs courtesy of En Masse Entertainment. Intrepid explorers ranging far and wide across the vast reaches of Arborea will no doubt run across a Naga or a Kumas in their travels, and today En Masse brings us a new Day in the Life lore snippet focused on the latter. Penned by Stacey Jannsen, Day in the Life -- Kumas takes you inside the one-track mind of the titular creature as he attempts to make a meal out of a wayward Popori and finds himself face to face with another (equally hungry) BAM with a similar goal. Head past the cut for more, and don't forget the three new screenshots in our TERA gallery. %Gallery-88481%
Darkfall blog details PvE tweaks, hints at big PvP changes
Exciting things are afoot in the world of Agon, and Aventurine's weekly Darkfall activity report is here to fill you in on all the grisly details. Whether you're keen on seeing one of your personal weapon designs imported into the game or curious about the ongoing PvE and AI updates, this week's entry makes for interesting reading. To kick things off, Aventurine's Tasos Flambouras gives us the low-down on grungrocs, and these gnarly boar-like bipeds are bent on ruining every adventurer's day. There's also talk of boss fight tweaks, and Flambouras writes that there are "significant changes" coming to high-end PvE content including damage cap adjustments, fight tuning, and loot table changes. Darkfall's PvP is getting some love as well, and new land-based objectives (similar to the game's sea fortresses) are in the works. Flambouras also hints at a new "Arena" project and "a lot more we're doing with PvP" that will be revealed at a later date. Finally, all Darkfall accounts have been gifted with 25,000 meditation points (the game's offline leveling currency) to celebrate the title's second anniversary. Check out the full presentation on the Darkfall Epic Blog.
IBM demonstrates Watson supercomputer in Jeopardy practice match
We're at IBM's HQ in upstate NY, where IBM will pit its monstrous Watson project (in the middle buzzer spot) against two Jeopardy greats, Ken Jennings and Brad Rutter. Watson has been in development for four years, and this is its first big public practice match before it goes on national TV in February for three matches against these giants of trivia. Unlike IBM's Deep Blue chess project in the 90s, which was pretty much pure math, Watson has to deal with the natural language and punny nature of real Jeopardy questions. IBM, ever the salesman, has thrown gobs of its fancy server hardware at the project, with 10 racks full of IBM Power 750 servers, stuffed with 15 terabytes of RAM and 2,880 processor operating at a collective 80 teraflops. IBM says it would take one CPU over two hours to answer a typical question, so this massive parallel processing is naturally key -- hopefully fast enough to buzz in before Ken and Brad catch on to the human-oriented questioning. We'll update this post as the match begins, and we'll have some video for you later in the day.
Quadrocopter plays the piano, wishes us a happy and complacent holiday (video)
Our worst frienemies, the quadrocopters, have decided to act cute for the holidays and play us a merry little jingle. Yes, the guys and gals behind the Flying Machine Arena have put together an airborne robot sophisticated enough to lay down a few seasonal notes on a Yamaha electronic keyboard. And we're still sitting around debating inconsequential topics like net neutrality -- all of human civilization is at stake here, people! Be a good citizen and watch the video after the break to scout out any weak points to this most imminent threat to humanity's survival.
Robo-nurse gives gentle bed baths, keeps its laser eye on you (video)
When they're not too busy building creepy little humanoids or lizard-like sand swimmers, researchers at the Georgia Institute of Technology like to concern themselves with helping make healthcare easier. To that end, they've constructed the Cody robot you see above, which has recently been demonstrated successfully wiping away "debris" from a human subject. The goal is simple enough to understand -- aiding the elderly and infirm in keeping up their personal hygiene -- but we'd still struggle to hand over responsibility for granny's care to an autonomous machine equipped with a camera and laser in the place where a head might, or ought to, be. See Cody cleaning up its designer's extremities after the break.
Dota 2 FAQ has Icefrog talking AI bots, replays and more
Icefrog himself has answered a set of frequently asked questions on Dota 2's official website, clarifying and detailing a few concerns players have had since the game's official announcement. Much of the questions revolve around players leaving games -- Valve has not only implemented an AI system to take over for missing players, but players will be able to have a "friend of similar skill level" jump in and take over if necessary, or simply jump into current games that are missing players. The bots will also be customizable, so if you want to play against an AI that tests a specific part of your game, like "denying, last hitting, and harassing," there will be specific configurations to challenge those. Replays are also mentioned -- players will be able to examine games in progress from all sorts of angles, watching charts and graphs in real time as the game goes on. There will also be "commentating and editing" features for replays, and the game will make extensive use of the Steam Cloud, keeping keyboard preferences, replay files and other settings available from any computer. Icefrog also says that there won't be any changes for the sake of change in the actual gameplay, and that he will continue development on the original Warcraft 3 mod "for as long as the community wants," but that Dota 2 "represents the long term future for the game." In that sense, Dota 2 sounds like a very direct translation of the original game, though Valve is sparing no expense in adding metagame features to make it a smoother experience for everyone.
NELL machine learning system could easily beat you at Trivial Pursuit
If fifteen years ago you would have told us that some day, deep in the bowels of Carnegie Mellon University, a supercomputer cluster would scan hundreds of millions of Web pages, examine text patterns, and teach itself about the Ramones, we might have believed you -- we were into some far-out stuff back then. But this project is about more than the make of Johnny's guitar (Mosrite) or the name of the original drummer (Tommy). NELL, or Never-Ending Language Learning system, constantly surfs the Web and classifies everything it scans into specific categories (such as cities, universities, and musicians) and relations. One example The New York Times cites: Peyton Manning is a football player (category). The Indianapolis Colts is a football team (category). By scanning text patterns, NELL can infer with a high probability that Peyton Manning plays for the Indianapolis Colts - even if it has never read that Mr. Manning plays for the Colts. But sports and music factoids aside, the system is not without its flaws. For instance, when Internet cookies were categorized as baked goods, "[i]t started this whole avalanche of mistakes," according to researcher Tom M. Mitchell. Apparently, NELL soon "learned" that one could delete pastries (the mere thought of which is sure to give us night terrors for quite some time). Luckily, human operators stepped in and corrected the thing, and now it's back on course, accumulating data and giving researchers insights that might someday lead to a true semantic web.
Google and TU Braunschweig independently develop self-driving cars (video)
There's a Toyota Prius in California, and a VW Passat halfway around the globe -- each equipped with bucket-shaped contraptions that let the cars drive themselves. Following their research on autonomous autos in the DARPA Urban Challenge, a team at Germany's TU Braunschweig let the above GPS, laser and sensor-guided Volkswagen wander down the streets of Brunswick unassisted late last week, and today Google revealed that it's secretly tested seven similar vehicles by the folks who won that same competition. CMU and Stanford engineers have designed a programmable package that can drive at the speed limit on regular streets and merge into highway traffic, stop at red lights and stop signs and automatically react to hazards -- much like the German vehicle -- except Google says its seven autos have already gone 1,000 unassisted miles. That's still a drop in the bucket, of course, compared to the efforts it will take to bring the technology home -- Google estimates self-driving vehicles are at least eight years down the road. Watch the TU Braunschweig vehicle in action after the break. Update: Though Google's cars have driven 1,000 miles fully autonomously, that's a small fraction of the time they've spent steering for themselves. We've learned the vehicles have gone 140,000 miles with occasional human interventions, which were often a matter of procedure rather than a compelling need for their human drivers to take control. [Thanks, el3ktro]
Fujitsu's social robot bear is the supertoy of Kubrick's dreams, almost (video)
Ah, the Fujitsu bear cub social robot. What child or elderly person should go without a "people-friendly terminal" with snuggly-soft fur and a camera in lieu of the ever-popular button nose? The little guy made the rounds briefly earlier this year, but this is our first time making acquaintances. The duo waved at us, waved with us, laughed with (at?) us, and then at some point, decided to arbitrarily fall asleep and ignore us -- just like our actual friends! The representative told us this prototype -- with a reported 300 actions from 12 actuators (three face, three neck, and six in the body), 13 skin touch sensors, hand grip sensors, a tilt sensor, and a microphone -- is a ways off from hitting the childcare / nursing home market. For now, enjoy the pictures below and footage after the break... and if all this seems eerily familiar, hey, we're with you. A wink and a nod in that direction can be found via the second video. %Gallery-104247%
Aventurine expanding Darkfall, development team
As much as many in the MMO community love to hate on Darkfall, the game soldiers on, and if recent reports coming out of Aventurine are any indication, it's actually growing a little bit. Producer Tasos Flambouras recently posted a lengthy development update on the game's official forums. In it, he discusses the ongoing server issues before moving on to more interesting news bits like the upcoming expansion, lore updates, and tweaks to various smaller aspects of the game's operation such as the online help and download sections of the website. Aventurine's main task these last few weeks has been polishing the new expansion, and Flambouras has lots to say on the matter. The team has redesigned many of the new dungeons, added new quests, monsters, and rewards, and continued working on terrain improvements. "I have to repeat what a monumental task this has been: the entire world has been repainted, all textures have been redone from scratch, there are no texture seams anymore or stretching," Flambouras writes. Finally, Aventurine looks to be expanding not just the world of Agon but the team constructing it. In addition to welcoming back a former member of the world-building team, Aventurine has hired two new programmers and one designer, and is currently interviewing more programmers in order to bring its future Darkfall plans to fruition.
Bungie's Halo AI secrets exposed in technical review
If you're looking for information about a hot new gun that will be shootable in the almost-here Halo: Reach, you might want to flip to one of the many other stories we've written about the game. If you want super esoteric insight into the high-functioning AI design of the series -- that is to say, the enemies' actual programmed intelligence, and not the game's fictional AI constructs -- we suggest checking out AI Game Dev's 42-point analysis on Bungie's tricks and innovations for making those pesky Elites so darn pesky. Of course, these tips will really only benefit you if you happen to be a game designer or programmer, though we suppose you could hypothetically use the information to impress your friends while playing Reach next week. "Look at the way Bungie implemented those perceptual information objects," you might say to your awestruck cohorts. "Man, that really must have cut down on some memory fragmentation and recycling problems."
Swiss researchers show off brain-controlled, AI-augmented wheelchair
They're far from the first to try their hand at a brain-controlled wheelchair, but some researchers at the École polytechnique fédérale de Lausanne (or EPFL) in Switzerland seem to have pulled off a few new tricks with their latest project. Like some similar systems, this one relies on EEG readings to detect specific brain patterns, but it backs that up with some artificial intelligence that the researchers say allows for "shared control" of the wheelchair. That latter component is aided by a pair of cameras and some image processing software that allows the wheelchair to avoid obstacles, but it doesn't stop there -- the software is also able to distinguish between different types of objects. According to the researchers, that could let it go around a cabinet but pull up underneath a desk, for instance, or potentially even recognize the person's own desk and avoid others. Head on past the break to check it out in action.
Super Scribblenauts features a reprogrammed Maxwell
Upon playing Super Scribblenauts, it becomes readily apparent why you'd want to use the D-pad to manipulate the game's hero, Maxwell. So why wasn't that an option in the first game? Joystiq asked 5TH Cell Creative Director Jeremiah Slaczka, and were told that the original touch-screen input didn't draw any complaints -- at least not from casual players. "We haven't gotten any emails from casual users," he told us this week, "that are like, 'Hey, I don't like the controls.' It's the hardcore users that are like, 'I play Mario all of the time, and I'm a hardcore gamer, and I'm used to these kinds of controls, and what's what I want." Though controlling Maxwell directly with the D-pad may seem an obvious choice, it wasn't the first time around. "Maxwell's actually an AI," Slaczka said. "So it wasn't just like throw the D-pad controls in." Maxwell was programmed to respond to the rest of the game's systems rather than just follow button directions, and so it didn't occur to the developers to control him directly. "You'd have to overwrite all of the code that we'd built up for him. So in the second one, we basically did that." The team "stripped out" all of the behaviors and responses that had been coded, and created the option for "one-to-one player control." For his own part, Slaczka says he doesn't care which option players use, but he'll stick with the stylus. "This isn't a platforming game," he said. "This is a puzzle game. Unfortunately, it has a platforming-type feel to it. So we said for the second one, that's fine, we'll address it, we'll give you both. And we actually fixed up the stylus controls, too, so Maxwell doesn't run away as you tap -- when you let go he stops. So we made it way better."
Prototype of robot that develops emotions on interacting with humans officially complete
The first prototype of a Nao robot that can develop emotions as it interacts with a human caregiver has been completed. A team across Europe was led by Dr. Lola Cañamero of the University of Herefordshire in the UK to develop the bot, which differs in several significant ways from those that came before it. These robots develop over time in much the way that a child does, learning to interact with and respond to the human beings around them. Modeled after human and chimpanzee childhood development paths, they are programmed to be highly adaptable to the people around them, and to become attached to whatever person is most suited to its needs and 'personality' profile. Over time, the more they interact, the more they learn and bond to the human being. These little ones, moreover, are capable of expressing a wide range of emotions, including anger, frustration, fear and happiness. The next steps are to research the bots' emotional and non-linguistic behavior, and to move toward combining linguistic and non-linguistic communication to become further attached and adapted to them. Yes, we want one.
Driverless vans set off on intercontinental trek from Italy to China (video)
You might not have expected the future to look like your granddad's groovy camper van, but take a closer look here and you'll find that this is indeed nothing like your forefather's people carrier. The VisLab team from the University of Parma have taken a fleet of Piaggio Porter Electric vehicles, strapped them with an array of cameras, lasers and other sensors, and topped them off with solar panels to keep the electronics powered. Oh, and lest we forgot to mention: the vans are (mostly) autonomous. VIAC (or VisLab Intercontinental Autonomous Challenge) is the grand name given to their big demonstration: an 8,000-mile, 3-month tour that will ultimately find them arriving in Shanghai, China, having set off from Milan this Tuesday. You can follow the day-by-day development on the blog below, though we're still being told that practical driverless road cars are a measure of decades, not years, away.
Darkfall details expansion mob AI
One of the things that makes Darkfall unique is its mob and NPC AI. Unlike traditional MMORPGs, the creatures in the world of Agon use the same skills that players use and also drop appropriate loot (for example, killing a goblin will likely net you his sword and armor, and boars don't drop coin). With the upcoming free expansion, the developers at Aventurine are further tweaking the game's PvE elements by adding additional monster skills to specific creatures. Laenih, Aventurine's community liaison, has all the details in a new post on the game's official forums. "Some existing monsters as well as new ones have already received their own abilities which we'll leave for you to discover. Additionally there will be more strategic interaction between monsters that spawn together," he writes. You can get a glimpse of one of the new monster teams in the form of a video that showcases the Gorra Dar and Gorra Slaves. Check it out after the cut or on YouTube.
Intel Connected Cars will record your bad driving for posterity, take over if you're really screwing up
Intel's latest Research Day has sprung up a new vision for "smart" vehicles; a vision that frankly chills us to our very geeky core. Cameras and sensors attached to an Intel Connected Car will record data about your speed, steering and braking, and upon the event of an accident, forward those bits and bytes along to the police and your insurance company. Just makes you feel all warm and fuzzy inside, doesn't it? Don't get us wrong, the tech foundation here is good -- having cars permanently hooked up to the ether can generally be considered a good thing -- but what's being envisioned is as obtrusive as it is irritating. Oh, didn't we mention that the cars can become self-aware and overrule you if you try to bend the rules of the road? Because they can.
Muon the humanoid robot is our ideal best friend
We don't speak German, and machine translation continues to be an intermittent and annoying bundle of failure, so bear with us on this one as we try to cobble together what exactly is going on here. This is Muon, the humanoid robot who is apparently being developed in Berlin by Frackenpohl Poulheim at the ALEAR Laboratory of Neuro Robotics at the Humboldt University in Berlin. Like other humanoid bots, Muon is about the size of an eight year old child so as not to creep out his human companions by being too threatening, and his design, while reminiscent of previous robots we've seen, is pretty original. It's actually hard to tell what stage of development Muon is in -- certainly many of the photos we have spied were concepts -- but we're going to keep our eyes peeled for him moving into the future. If you hit up the source link, you can check out a video of Muon's development. There's one more amazing shot after the break.
IBM's Watson is really smart, will try to prove it on Jeopardy! this fall (video)
As much as we love our Google homepage, computer search remains a pretty rudimentary affair. You punch in keywords and you get only indirect answers in the form of relevant web results. IBM doesn't seem to be too happy with this situation and has been working for the past three years on perfecting its Watson supercomputer: an array of server racks that's been endowed with linguistic algorithms allowing it to not only recognize oddly phrased or implicative questions, but to answer them in kind, with direct and accurate responses. Stuffed with encyclopedic knowledge of the world around it, it answers on the basis of information stored within its data banks, though obviously you won't be able to tap into it any time soon for help with your homework. The latest word is that Watson's lab tests have impressed the producers of Jeopardy! enough to have the bot participate in a televised episode of the show. That could happen as early as this fall, which fits right in line with our scheduled doom at robots' hands by the end of 2012. Ah well, might as well get our popcorn and enjoy the show.
