The creation of artificial general intelligence would be one of the biggest breakthroughs in the field of AI for years. Has it already happened or are the claims of Kimera Systems just hot air?
The field of artificial intelligence has spawned a vast range of subset fields and terms: machine learning, neural networks, deep learning and cognitive computing, to name but a few.
However here we will turn our attention to the specific term 'artificial general intelligence', thanks to the Portland-based AI company Kimera Systems' (momentous) claim to have launched the world's first ever example, called Nigel.
What is AGI?
The AGI Society defines artificial general intelligence as "an emerging field aiming at the building of “thinking machines”; that is general-purpose systems with intelligence comparable to that of the human mind (and perhaps ultimately well beyond human general intelligence)".
AGI would, in theory, be able to perform any intellectual feat a human can. It is also referred to as 'strong AI' or 'full AI'.
You can now perhaps see why a claim to have launched the world's first ever AGI might be a tad ambitious, to say the least. It certainly merits a lot of interrogation.
However Kimera Systems insists that is precisely what it has done. It says 'Nigel' is a "breakthrough achievement in artificial intelligence coming decades before most experts predict such AI technology could be ready for commercial deployment".
It would be an enormous advance within AI and indeed technology as a whole.
"The field of “artificial general intelligence” or AGI has made no progress whatever during the entire six decades of its existence", according to Oxford University physicist David Deutsch in a wide-ranging essay on the topic. AI experts predicted AGI would take many more decades even from now to become a reality. So can AGI really already exist now, in 2016?
Is 'Nigel' really AGI?
Kimera Systems insists it is. "Nigel completely qualifies with those definitions [above]", a spokesman told Techworld.
"I do think “performing any intellectual feat a human can” may be overstating where Kimera is now. But Nigel observes, learns and understands without any human intervention (no guiding hand, no special programming to point it toward understanding),” he says.
Its cofounder Mounir Shita started working on the AGI technology in 2005 but unveiled it to the world on 15 August 2016. It has been launched in private with a public beta promised "soon" in the press release.
We shouldn't necessarily write 'Nigel' off just because of the enormity of the breakthrough it would represent. However it is impossible to judge until its algorithm is peer reviewed and properly tested, according to Matt Jones, head of analytics at data science consultancy Tessella.
In theory, AGI doesn't need to be trained in order to learn, as opposed to other established cognitive platforms like IBM's Watson, which must ingest training data in order to make sense of tasks (for example it needs a lot of oncology data to effectively diagnose cancer), he explains.
"In general terms AIs perform best at large numbers of repeatable tasks using large data sets. Humans tackle novel situations where there's little prior information or random context. Kimera claims that Nigel operates in human space," Jones says.
In short, it seems too early to tell whether or not 'Nigel' is true AGI, although it would be a huge breakthrough if so. We'll be following this one closely and will update with any and all developments, reaction and commentary. Watch this space.
Original source: TechWorld