In the recent past, we’ve heard a lot about how AI is going to replace every worker. The problem is, that’s a lie.
To put this into perspective, we’ll start off with a thought experiment. Picture a fruit seller in the year 1700. He has a horse drawn cart and picks his own fruit. Now let’s say he takes a nap for 100 years and wakes up in 1800. Not much has changed for him, same old cart, same old dirt road. Now the fruit seller goes back to sleep until 1940 and wakes up. Everything around him has changed.
He has a van, paved roads, he can talk to someone thousands of miles away in real time, not over a period of months. He could bring light into a room with the flick of a switch. He– you get the point. Now, the fruit seller goes back to sleep and wakes up in 2019. Apart from the internet, nothing else has changed. The only new thing he’d have in his house would be a microwave. That’s it.
This monumental change between the 19th and 20th century was driven by two new technologies: electricity, and the internal combustion engine. The magic of these two technologies was that they were adaptable. Electricity can light up a room, heat a pan, power a computer; the engine could be used for deliveries, for transport, for assembly lines. The versatility of these technologies allowed them to become a part of everyday life.
But these technologies created a ridiculous increase in productivity (something economists fawn over) that was long lasting. The only other time we’ve seen a similar spike in productivity was when computers were introduced. Since then, productivity growth has largely stagnated, despite the increasing attention to AI.
This is the case made by Robert Gordon: AI isn’t the next big thing, according to the numbers. In fact, Robert Gordon goes on to argue that we need more change, not less in his book The Rise and Fall of American Growth. There is still reason for worry as the book goes on to warn about the results of an aging population, rising inequality and the level of debt borne by college graduates. Gordon also seems to imply that we may be reaching a ceiling in terms of productivity as previously difficult tasks which involved manual labour were easy to solve. However, most work is in the service sector and is unlikely to be easily improved. In fact, some technological advances may dramatically change the way we live our lives, but may have little or no impact on our work, as is the case of the iPhone as mentioned in a Planet Money interview.
But the change that may significantly increase productivity is likely to be something small. For example, air conditioning allowed workers, who were often stuck in an office during summer to focus more on their work. In a similar vein, plumbing and increased sanitation provided unlikely benefits. We may not need to reinvent the wheel, we just have to find one small thing that hinders most of a population and fix that.
To conclude, AI is not going to be the next big thing for one reason: to truly revolutionise the world, it has to be as commonplace as electricity. We may see this in a few decades, according to conservative estimates, but until then, we don’t have to worry. To further reassure you that machines and robots will not be taking away jobs, you might want to try and play a game made by Gordon: “Spot the Robot”. Next time you’re out and about, count how many robots you can see. Chances are, that number will be small.
This article was also published on forefrontmedia.net
Interesting. I was in a similar stance when I first thought about AI. Then I began reading into the subject and exploring the various possible routes it can take. And with the current pace, there really is just one – and it’s a dangerous one.
You used an intelligence analogy at the beginning about how shifting the same guy through the centuries has a ‘surprise factor’. So a 1500 guy would NOT be blown away by the development of 1700 but a 1700 guy would probably die from surprise if you put him into 2000. What we’re actually noticing from these shifts relative to the year is that there’s an exponential rate of progress in intelligence and therefore human development.
If you imagine this scale being stretched out. It sky rockets upwards. But, there’s a certain point in time where we are intelligent enough to give birth to a vastly higher intelligence species. In our case, it’s AI.
AI has 3 levels:
1) ANI – Artificial Narrow Intelligence (can perform simple and specific tasks)
2) AGI – Artificial General Intelligence (human level intelligence)
3) ASI – Artificial Super Intelligence (vastly more intelligent than a human)
Within the next 100 years, we’re looking at an almost complete transition to ASI. For intuition, humans have an IQ ranging from 85-135 whilst these ASI’s have the IQ of about 12,000. Yes! And, what we’ve noticed from ourselves is that more intelligence = more power. So, as we currently are the most ‘intelligent’ species, we have the power. For example, if you saw an ant and it was obstructing you in even the slightest amount from achieving your task, you’d instantly kill it. Or at least, you have the power to harm it. Just imagine what a vastly more intelligence species would do if it saw that we were trouble(which is almost always likely). For all we know, we could go extinct!
This is exactly why famous AI thinkers including James Barrat, Ray Kurweil, Nick Bostrom and many more are exploring the topic to raise awareness to perhaps slow down the rate of development of AI to think about its possible danger(s).
This may be a slightly pessimistic viewpoint however, I do believe that we need to raise awareness that AI is actually the most important debate now and we need to shift the spotlight to it. And by assuring people that it’s not a danger, we are not giving them full context. Hence, I disagree with you and I believe that we SHOULD be worried about it. Forget jobs, our species could be at stake here.
I have condensed my thoughts into a comment. I have a lot more thought relating to the topic so if you have any queries, fire them at me!
I’ll admit that I don’t know as much about AI as you clearly do, but as of yet we haven’t seen a massive shift in the way we live. It’s likely that the power of AI will grow exponentially but as of yet we’re in a weird lull between life changing innovations. I do agree that not doing something about AI is shooting yourself in the foot then the head.
Excellent article! AI’s always been regarded as a leviathan turning point in society – I’ve never thought about it like this.
At first, reading the heading I thought it’s weird, but you know what, you are right!
Gave me different perspective! Well done!