Artificial intelligence (AI) has advanced in almost every part of society, but what people can do with it is currently limited.
Artificial Narrow Intelligence (ANI) is technology that uses algorithms and programming to solve problems in a limited set of tasks at a capacity exceeding human potential.
Examples of ANI include “recommended” sections on Netflix that are based on previous searches, or self-driving cars that are expected to be on the market by 2018and reportedly run on data stored on an internal map, various algorithms, and cameras and sensors that anticipate obstacles to choose the fastest route.
Moore’s Law – a term coined by Intel co-founder Gordon Moore in 1965 – predicts that the processing power of computers doubles every two years.
While University of Newcastle Laureate Professor Jon Borwein and research fellowDavid Bailey wrote in The Conversation that progress might be stalling in a processor’s clock speed, it’s widely agreed that technology is developing at an exponential rate.
But humans have an advantage over ANI, because they can think laterally and learn autonomously.
Ben Goertzel’s coffee test found while ANI enables a robot to use a coffee machine, it can’t find the coffee beans if it’s not programmed to, highlighting the disparity in intelligence between AI and humans.
But leading professors specialising in artificial intelligence argue that AI’s inability to perform these tasks is only temporary.
Artificial General Intelligence (AGI) – also known as High Level Machine Intelligence (HLMI) – is the next progression for robots to perform the same tasks and hold similar problem solving abilities as humans.
Moore’s Law suggests the transition from ANI to AGI is more a question of “when” not “if”.
In a questionnaire provided to 170 AI experts at the Philosophy and Theory of AI conference held on November 2012, Swedish philosopher Nick Bostrom asked: “In what year would you see a (10 per cent/ 50 per cent / 90 per cent) probability for such High Level Machine Intelligence (HLMI) to exist?”
The media response from academics was that there was a 50 per cent chance AI would reach human-level intelligence capacity in 25 years.
The survey also showed experts believed it was 24 per cent likely we will create machine simulation of human thought in the next 26 to 50 years, while 22 per cent believed we will create machine simulation of human thought in the next 26 to 50 years.
In ground breaking progress, researchers at the US Ransselaer Polytechnic Institute – a technological research university – in July programmed robots to answer questions on a logic puzzle, testing their self-awareness.
Two robots took “dumbing pills” that silenced them, and the other robot took a placebo. When asked which robot had taken the placebo, one of the robots that had responded: “I don’t know”. But the robot changed its answer after the other robots didn’t respond.
According to Business Insider, this is the first time a robot has shown self-awareness.
“To be able to claim that the robot is exhibiting ‘self-awareness’, the robot must have understood the rules, recognised its own voice and been aware of the fact that it is a separate entity from the other robots,” Business Insider journalist Celena Chong said.
According to Mr Bostrom, Artificial Super Intelligence (ASI) is the stage after AGI where robots are “much smarter than the best human brains in practically every field, including scientific creativity, general wisdom and social skills”.
Moore’s Law anticipates an incredibly rapid transition from AGI to ASI. When AI reaches human capacity, it’s expected to keep doubling in its processing power every two years until it far exceeds human intelligence.
But because it’s based on future predictions, experts aren’t unanimous in deciding what impacts this will have on humans.
Mr Bostrom said ASI might lead to either, immortality and solutions to global problems, including global warming, poverty, and disease, or the destruction of humans.
In the two-part series The AI revolution on Wait But Why – a New York based blog that tackles a different topic in depth every few weeks – Tim Urban explained how these two possibilities might occur.
Mr Urban wrote that ASI could be used to solve famine by using nanotech – technology that manipulates atoms and molecules – “to build meat from scratch”.
He also uses a hypothetical to show how ASI could destroy humanity. An AI would be programmed to fulfil an activity, but the biggest threat to this activity may be mankind’s ability to destroy the robot.
Through a combination of nanotechnology – which allows a robot to build things – and Internet connection, the robot may choose to eliminate the population to prevent this risk from occurring.
He wrote that “an ASI system will be both amoral and obsessed with fulfilling its original programmed goal. This is where AI danger stems from. Because a rational agent will pursue its goal through the most efficient means, unless it has a reason not to.”
Google’s director of engineering Ray Kurzweil predicted that by 2045 technology’s “progress [will be] so rapid it outstrips humans’ ability to comprehend it”.
But he believes that – through ethical constraints and guidelines – it will likely present positive outcomes for humans.
“We have the opportunity in the decades ahead to make major strides in addressing the grand challenges of humanity. AI will be the pivotal technology in achieving this progress. We have a moral imperative to realize this promise while controlling the peril. It won’t be the first time we’ve succeeded in doing this,” he wrote.
Mojonews.com.au will publish a three-part series that unpacks a more imminent threat posed by ANI and AGI than the destruction of humanity – the possible destruction of the job market.
The debate on how the development of AI may affect the job market polarises.
Manufacturing jobs have been in decline since industrialisation, and jobs where customer service is a key skill are among the latest causalities to robots, such as self-service checkouts at supermarkets.
Some predict that it will be a massive interference, taking all of the jobs away from humans, leaving the world in a post-work state.
Derek Thompson speculates on what this dystopia might look like in The Atlantic.
Experts currently predict that about 50 per cent of jobs will be made redundant and filled by artificial technology by 2025, according to a joint report from realty consulting firm CBRE and property development company Genesis.
But many believe AI eliminates jobs we don’t want, and creates new jobs that previously didn’t exist. It even might lead to a society where technology is productive enough to leave humans with more leisure time