The story of artificial intelligence

Written by bit-tech Staff

March 19, 2012 | 09:36

Companies: #bit-tech

Perception And Mobility
The field of humanoid robotics is one offshoot of AI that attempts to meet the challenges of perception and mobility head on – a type of research not looked upon favourably by all members of the scientific community. Doctor Marvin Minsky, an important voice of optimism in the early days of AI, was highly critical of such projects. Speaking to Wired magazine in 2003, he was particularly scornful of the field of robotics. "The worst fad has been these stupid little robots," said Minsky. "Graduate students are wasting three years of their lives soldering and repairing robots, instead of making them smart. It's really shocking."

Such comments didn't deter projects like Asimo, Honda's attempt to create a humanoid robot. While its photogenic appearance and ability to walk with relative ease immediately captured media attention, what is most significant about Asimo is its ability to recognise objects, gestures and sounds; Asimo will give a handshake to an outstretched hand, respond to its name when called, and can, most remarkably, work out the identity of an unfamiliar object by comparing it to similar items in its memory banks.

The story of artificial intelligence Artificial intelligence, perception, and achievements
Honda's Asimo

Despite its impressive abilities, Asimo is but a small shuffle forward on the road of AI research; a cutting-edge synthesis of already extant technologies and programs. The fundamental difference between human and programmed intelligence is still science's greatest challenge. While neuroscientists continue to map the human brain, a complete and comprehensive theory of how it actually works has proved tantalisingly out of reach.

Connectionism, a philosophical theory of the mind's workings that originated in the nineteenth century, became popular with scientists in the 80s. According to connectionist thinking, the brain works like an enormously complex computer, with its neurons behaving as individual units interconnected like computers on a network. The attempt to artificially recreate a neural network has been the subject of ongoing research - and constant debate - for decades. Rival schools of thought, for example computationalism, argue that connectionism doesn't accurately describe how the mind works.

In an era that is still struggling to simulate the comparatively simplistic cerebral workings of insects, American neuroscientist Jim Olds has argued that what science needs is an authoritative model for how our minds work. "We need an Einstein of neuroscience," Olds said, "to lay out a fundamental theory of cognition the way Einstein came up with the theory of relativity."

The AI Around Us
Glancing over the history of AI research, it's easy to assume that little has been achieved. The all-knowing machine intelligences promised by writers and distinguished scientists failed to materialise, and the robot servants predicted to appear in every modern home have remained firmly in the realms of science fiction.

Yet AI has yielded a wealth of applications that we now take for granted. Search engine algorithms like Google, which are in constant use by millions of people every day, exist as a result of AI research. Meanwhile, banking systems can recognise unusual patterns in customer spending to detect credit card fraud. Facial recognition systems are now used as a part of airport security. Speech recognition continues to be refined and has been integrated into Windows system software since Vista (and there's Apple's Siri, too). And, of course, AI plays an important role in videogames. It’s inarguable that many of these AI-related innovations would not have existed were it not for government - as opposed to private - funding.
Discuss this in the forums
YouTube logo
MSI MPG Velox 100R Chassis Review

October 14 2021 | 15:04