Artificial Intelligence and the Human Mind – Wai H. Tsang
Artificial Intelligence (AI) is so much in the news today. It is the Holy Grail of modern times. The creation of True AI will fundamentally transform the world, our lives, society and political economy. From reading the news headlines, we seem to be on the verge of this great transformation. But if we see beyond the hype then we realize that the recent ‘breakthroughs’ in AI are actually rather limited in potential, narrow in scope and not very ‘intelligent’. We’ve seen AIs that can play mindless computer games with superhuman ability but fail miserably at even simple games like ‘Pac Man’, which require just a little planning ahead. We now have self driving cars that navigate highways competently 99% of the time but constantly require human intervention without which they even end up crashing into parked fire engines. We now have AIs that can defeat the best humans at TV quiz games and the ancient board game ‘Go’, but can’t play hangman or tic tac toe, even at a small child’s level. The goal of creating true, general and strong AI is as elusive as ever.
What then is the problem with existing AI? Why is there this big discrepancy between on the one hand the headlines and the claims of some AI experts versus on the other hand the actual state of affairs. The most popular AI algorithm today, which is used in the so called deep learning variety of AI and which has been instrumental in much of the recent progress, has been touted by some as the path to human level intelligence. But even its main originator and AI leading light, Jeffrey Hinton, has said with respect to this technology that, "My view is to throw it all away and start again,", “It works better than anything else at present but that doesn’t mean it’s right”. This point of view is growing in the AI community. In the technological hype cycle of AI, we are now entering the ‘trough of disillusionment’. Are we heading for another so called ‘AI Winter’ where interest in AI and funding for it, will fade away over time? The answer is no, because the biggest corporations and national governments of the world have too much focus and money invested already. All eyes are on the prize, and the prize is very great indeed.
Artificial Intelligence (AI) is so much in the news today. It is the Holy Grail of modern times. The creation of True AI will fundamentally transform the world, our lives, society and political economy. From reading the news headlines, we seem to be on the verge of this great transformation. But if we see beyond the hype then we realize that the recent ‘breakthroughs’ in AI are actually rather limited in potential, narrow in scope and not very ‘intelligent’. We’ve seen AIs that can play mindless computer games with superhuman ability but fail miserably at even simple games like ‘Pac Man’, which require just a little planning ahead. We now have self driving cars that navigate highways competently 99% of the time but constantly require human intervention without which they even end up crashing into parked fire engines. We now have AIs that can defeat the best humans at TV quiz games and the ancient board game ‘Go’, but can’t play hangman or tic tac toe, even at a small child’s level. The goal of creating true, general and strong AI is as elusive as ever.
What then is the problem with existing AI? Why is there this big discrepancy between on the one hand the headlines and the claims of some AI experts versus on the other hand the actual state of affairs. The most popular AI algorithm today, which is used in the so called deep learning variety of AI and which has been instrumental in much of the recent progress, has been touted by some as the path to human level intelligence. But even its main originator and AI leading light, Jeffrey Hinton, has said with respect to this technology that, “My view is to throw it all away and start again,”, “It works better than anything else at present but that doesn’t mean it’s right”. This point of view is growing in the AI community. In the technological hype cycle of AI, we are now entering the ‘trough of disillusionment’. Are we heading for another so called ‘AI Winter’ where interest in AI and funding for it, will fade away over time? The answer is no, because the biggest corporations and national governments of the world have too much focus and money invested already. All eyes are on the prize, and the prize is very great indeed.
