AI to pass Turing test?

I am somewhat doubtful whether this project will succeed, but let’s wait and see what it can do …

PORTLAND, Ore. — Passing the Turing test–the holy grail of artificial intelligence (AI), whereby a human conversing with a computer can’t tell it’s not human–may now be possible in a limited way with the world’s fastest supercomputer (IBM’s Blue Gene), according to AI experts at Rensselaer Polytechnic Institute. RPI is aiming to pass AI’s final exam this fall, by pairing the most powerful university-based supercomputing system in the world with a new multimedia group designing a holodeck, a la Star Trek.

http://www.eetimes.com/showArticle.jhtml?articleID=206903246

Controlling computers by thoughts

http://news.bbc.co.uk/2/hi/technology/4074869.stm

A team of US researchers has shown that controlling devices with the brain is a step closer.

Four people, two of them partly paralysed wheelchair users, successfully moved a computer cursor while wearing a cap with 64 electrodes.

Previous research has shown that monkeys can control a computer with electrodes implanted into their brain.

The New York team reported their findings in the Proceedings of the National Academy of Sciences.

“The results show that people can learn to use scalp-recorded electroencephalogram rhythms to control rapid and accurate movement of a cursor in two directions,” said Jonathan Wolpaw and Dennis McFarlane.

Moth brain in robot

moth robot

The moth, immobilized inside a plastic tube, was mounted on a 6-inch-tall wheeled robot. When the moth moved its eyes to the right, the robot turned in that direction.

The research, presented at the Society for Neuroscience’s annual meeting in San Diego this week, marked the latest advance by scientists who are trying to perfect the link between brain and machine.

The technology might one day help people who are paralyzed or have lost their limbs regain the ability to move. Scientists have developed experimental devices that allow those patients to move cursors on a screen or crude robotic hands using only their thoughts.

http://www.latimes.com/news/science/la-sci-robomoth7nov07,0,3564441.story

Virtual robots duped by illusions help to explain human vision

From http://www.ucl.ac.uk/media/library/robotillusions

28 September 2007

A study by researchers at UCL (University College London) explains why humans see illusions by showing that virtual robots trained to ‘see’ correctly also – as a consequence – make the same visual mistakes that we do. The study, published in the latest edition of PLoS Computational Biology, shows that illusions are an inevitable consequence of evolving useful behaviour in a complex world.

Dr. Beau Lotto, UCL Institute of Ophthalmology, said: “Sometimes the best way to understand how the visual brain works is to understand why sometimes it does not. Thus lightness illusions have been the focus of scientists, philosophers and artists interested in how the mind works for centuries. And yet why we see them is still unclear.”

To address the question of why humans see illusions, researchers at the UCL Institute of Ophthalmology used artificial neural networks, effectively virtual toy robots with miniature virtual brains, to model, not human vision as such, but human visual ecology. Dr David Corney in Dr. Lotto’s lab trained the virtual robots to predict the reflectance (shades of grey) of surfaces in different 3D scenes not unlike those found in nature. Although the robots could interpret most of the scenes effectively, and differentiate between surfaces correctly, they also – as a consequence – exhibited the same lightness illusions that humans see.

larynx silent-speech control wheelchair

From http://www.engadget.com/2007/09/10/mind-controlled-motorized-wheelchair-demonstrated/

Ambient, in partnership with the Rehabilitation Institute of Chicago, has developed a wheelchair that can be instructed to move when the driver thinks about certain words. The key component is a larynx control system called the Audeo, developed by the founders of Ambient, Michael Callahan and Thomas Coleman.

http://technology.newscientist.com/article.ns?id=dn12602&feedId=online-news_rss20

Using AI to run scientific experiments

Nature 427, 181 (15 January 2004); doi:10.1038/427181ba British team unveils an automated system that “originates hypotheses to explain observations, devises experiments to test these hypotheses, physically runs the experiments using a laboratory robot, interprets the results to falsify hypotheses inconsistent with the data, and then repeats the cycle”.

What’s more, when set loose on experiments to investigate the genetic control of important metabolic pathways in yeast, it performs more cost effectively than scientifically educated human volunteers.

Braingate implant reads brain signals

Nov 2005A few years ago scientists managed to wire a monkey’s brain to a robotic arm. The monkey learned to manipulate the arm simply by thinking. This year, John Donoghue at Brown University has managed to do the same – with a human.

Matthew Nagel, a C4 paralytic – from the neck down – has been playing pong, drawing circles and manipulating a computer with thoughts alone via the BrainGate. This technology promises to have profound changes on the lives of millions of paralyzed and handicapped individuals, but implications run much deeper and may one day affect us all.

Epileptic 14-year old plays space invaders with brain signals

http://news-info.wustl.edu/news/page/normal/7800.html

Oct. 9, 2006

A St. Louis-area teenage boy and a computer game have gone hands-off, thanks to a unique experiment conducted by a team of neurosurgeons, neurologists, and engineers at Washington University in St. Louis. The boy, a 14-year-old who suffers from epilepsy, is the first teenager to play a two-dimensional video game, Space Invaders, using only the signals from his brain to make movements.

The teenager had a grid atop his brain to record brain surface signals, a brain-machine interface technique that uses electrocorticographic (ECoG) activity – data taken invasively right from the brain surface. It is an alternative to a frequently used technique to study humans called electroencephalographic activity (EEG) – data taken non-invasively by electrodes outside the brain on the scalp. Engineers programmed the Atari software to interface with the brain-machine interface system.