This is an article that was stimulated by an interview with Marc Andreeson. He is one of the leading technological minds in the country, having invested in many of the largest startups over the last 3 decades.
His conclusion is that, in economic terms, we havent really seen a massive step forward technologically. As a percentage of the overall output, his claim is that we have been rather flat. Certainly there was innovation but it lags where it should be based upon population growth and the corresponding growth in GDP.
With the population increase evidently ending, this will cause a contraction in the economy. Of course, the way to offset this is to increase the amount of technology.
Could that be lining up? I take a look at this.
AI: Offset Population Decline or Produce Job Loss
I am in the camp that AI will cause massive job loss. On this, I agree more with Emad Mostague as opposed to Andreeson. Both are very smart individuals with long histories in the tech sector.
So what is the thesis put forth by Andreeson.
His view is that AI is coming at the perfect time. In fact, without it, we would be in fear of economic collapse. With population decline almost certain over the next century, economies would be wiped out. A depression the likes of the 1930s would be the result.
Most of the largest economies face population decline. Even the African nations are seeing a decline in the birthrates, albeit still above replacement level in most instances. If those areas progress economically, it is likely that we see those countries replicate the West and China.
This is a logical conclusion. If we are seeing a massive population decline, automation, especially robots, can do the things we simply do not have enough people for. It will replace the labor that is basically lost through the aging process.
It is one that is hard to argue, at least on the surface. The only variable is whether the population will decline, which is almost a guarantee.
So how could anyone give any credence to what Mostaque proposes?
The Pace of Change
One of the keys in economics is pace of change. It is one thing to look at raw numbers. Unfortunately, that often tells only part of the story.
I wrote an article that discussed Disney stock being down 33% over the past 5 years. In an absolute lens, this is bad. It is horrific when we look at it on a relative basis. Over the same period, the overall market is up 86%.
Within this discussion is the question of pace. Which is going to progress faster.
Here is where I think Andreeson's thesis falls about: He mentioned population decline over the next century, something I agree with. However, the pace of AI is in years with robotics likely having a major impact over the next decade.
In other words, technology is moving much faster than population decline. This makes sense when we think about how long it takes for a generation to age out. We are looking at 25-30 years before one generation is lost. Since the major drop in the fertility rates were in the last 3 decades, it will take 60 years or so before this full weight of this is felt.
Anyone care to guess where we will stand with automation in the 2080s? How about the 2040s?
It is almost certain we will have AGI (whatever that is) by then. At the same time, it is probable that some type of "superintelligence" is reached.
This will be far advanced from what we are dealing with today. Hence, the reason why I am in the Mostaque camp as compared to Andreeson is the forecast of 900 days. This is what he put forth as his guess when AI will be able to replace the vast percentage of knowledge workers.
Since this was made around September, we are now down to 800 days.
Even if this is off by 500 days, the population decline will not be felt by then. Even South Korea is okay for the next decade or so.