Author : admin | Thursday, 25 July 2019
Author : admin | Thursday, 25 July 2019
Artificial Intelligence (AI) - the automated and technological imitation of human and animal intelligence - is a principle almost as old as civilization itself. In Homer’s Iliad, the Greek God of fire and forges, Hephaestus, was a founder of robots. From golden tripods that served food at banquets to fully discovered simulacrums of humanity suitable of conversation and storytelling.
The machines may have been divine in origin, but were also automated in nature, and betray the first flickerings of human attraction with recreating, from metal and stone, the most enigmatic and powerful element of humanity: our intelligence.
In her book, Machines Who Think, Pamela McCorduck notes that, historically, human attitudes towards AI diverged in antiquity into two camps: the Hellenic and the Hebraic. The Hellenic (Greek) tradition, as in the Iliad, expressed human wonder at the idea of thinking machines, and treated them as something to be strived for.
The latter school of thought, educated by the Second Commandment ("Thou shalt not make unto thee any graven image" - fundamentally a non-compete clause stating humans shouldn’t have any other gods or undertake any of the activities that fell within God’s purview) accomplished the idea of making something akin to life in any way but the biblical sense with fear and revulsion. Turning to his creator, the monster says: “I ought to be thy Adam; but I am rather the fallen angel.” Frankenstein’s monster perfectly embodies the Judeo-Christian fear of the thinking machine. Fast forward to the 1980’s, and you have The Terminator; captivation and fear have always surrounded the idea of AI.
Evidently, the Second Commandment didn’t stand in the way of human inventiveness for long. Today, AI drives business analysis and decision making around the world. Accenture forecasted in 2017 that by 2035, the US market for AI would attain a value of $8.3trn per year. Constantly improving in terms of power and computational capability, AI is perhaps the most impressive advancement of the past century, competing with polio vaccines, the internet and manned space travel. But there’s a problem...
The limitations of specialized AI
An AI can play chess better than a grandmaster, read an x-ray with more preciseness than a head of surgery and fly a fighter jet with more flair than Will Smith in Independence Day. Nevertheless, all these mind-boggling feats need to be practiced by many AI systems, and each one demands significant training and preparation.
“We’re quite far from having machines that can learn the most basic things about the world in the way humans and animals can do,” said Facebook’s head of AI, Yann LeCun in an interview with The Verge. “In certain areas gadgets have superhuman performance, but in terms of general intelligence we’re not even close to a rat.”
The future is adjustable
The trick to the next step towards the holy grail of an adjustable, usually intelligent AI (an AGI) may presently lie within the walls of an office on 18th Street in San Francisco. Started in 2015 by a cohort of Silicon Valley luminaries (including Elon Musk) OpenAI’s 100-strong team has been striving to break the chasm between the present state of the art and the probably limitless applications of the next step in creating a true “thinking machine.”
According to the OpenAI team, “AI system building today includes a lot of manual system for each well-defined task. In contrasting, an AGI will be a system capable of learning a field of study to the world-expert level, and mastering more fields than any one human — like a tool which integrates the skills of Curie, Turing, and Bach. An AGI working on a challenge would be able to see relationships across disciplines that no human could. We want AGI to operate with people to solve currently intractable multi-disciplinary problems, including global challenges such as climate change, affordable and high-quality healthcare, and personalized education. We think its impact should be to give everyone economic freedom to pursue what they find most rewarding, creating new opportunities for all of our wellbeing that are unimaginable today.”
So far, it’s done quite well. The OpenAI has displayed a various range of abilities, breaking records for dexterity, creating persuasive song lyrics and short stories, and taking on human champions at Dota 2. The OpenAI project was formerly intended to be purely non-profit, but has since announced that it is starting itself up to exterior investment.
The software giant announced yesterday that it will be spending $1bn into OpenAI, in order to further drive the company’s advancement. According to the OpenAI team, “we’re pairing to develop a hardware and software platform within Microsoft Azure which will scale to AGI. We’ll collectively create new Azure AI supercomputing technologies, and Microsoft will turn out to be our exclusive cloud provider—so we’ll be working hard with each other to further extend Microsoft Azure’s capabilities in large-scale AI systems.”
“The creation of AGI will be the most significant complex development in human history, with the potential to shape the trajectory of mankind,” said Sam Altman, CEO, OpenAI. “Our mission is to guarantee that AGI technology benefits all of humanity, and we’re working with Microsoft to build the supercomputing foundation on which we’ll build AGI. We feel it’s crucial that AGI is integrated safely and securely and that its economic results are widely distributed. We are excited about how deeply Microsoft shares this vision.”
This article is originally posted on manufacturingtomorrow.com