The present global economy is, overall, exploding. Naturally, with a statistic like this, we get different numbers. Some sources place it around $110 trillion. Our World in Data places it at $166 trillion as of 2023 (using 2021 dollars). We will use the latter for this article in an effort to maintain consistency. The raw numbers do not matter as much as the percentages.
Here is what the chart looks like:
In 1990, the total was $59.68T. Over the past 30+ years, we saw more than a 2.5x of the global output. This coincides with the explosion in technology. We saw the digital world explode in the last 40 years. It is showing up in the numbers.
That said, there was a slowing over the last few decades. The move from 1960 to 1990 was more than a 3x. This is consistent what other studies found.
Many are forecasting that the emergence of AI, along with the feedback loops, will not only bring the growth rates back to historical norms, it will obliterate them.
EpochAI is doing its best to model where things are going. It is working on a number of different scenarios that could be unfolding. The results we will use here stem from looking at compute, investment, and automation trends. These are tracked individually and then merged into economic forecasts.
Before digging into what they are proposing, let us answer what is EpochAI? According to Grok:
Epoch AI is a research institute focused on studying the trajectory and impact of artificial intelligence (AI). It investigates key trends and questions related to AI development, such as advancements in computing power, data usage, and algorithms, to forecast how AI might shape economies and societies in the future. The organization aims to make its findings accessible through reports, models, and visualizations, providing an empirical foundation for discussions about AI. Founded in 2022, Epoch AI operates as a nonprofit with a multidisciplinary team dedicated to AI governance and forecasting.
Anyone who went through the site sees the research they are undertaking along with the results. This is quickly becoming one of the cited works for people who follow the world of AI and its progress.
We discussed the potential of the economic singularity, which I defined as 30% annual growth for the global economy. If EpochAI is anywhere near accurate, this is a minimum guarantee. They focus heavily on the automation process, something that will make sense as we delve deeper.
One of the keys to their analysis is the rate of automation. This is contained in another chart, one that is interactive. To view that, click here.
As we can see from that chart, it has automation being a small percentage of the total economy. The rate is about 10% into 2026. From there, however, the expectation is that it will explode.
By 2027, the forecast is 27% on the conservative side with 32% on the default. The 2030 reading is 51% and 69% respectively.
Obviously, this results in a major difference. Here is the forecast charted out.
Like all exponential charts, things start off slowly. The early years appear flatlining. Things change as the compounding impact takes hold.
What does these numbers amount to?
If we look at the 30% line, we see that we reach 10^16 in 2041. That is $10 quadrillion. Anyone who follows the power of compounding understand this.
Looking at the other lines (light blue and pink) we can see both the full automation pace for both the default and conservative. The latter lags the 30% growth rate into the 2040s before passing it. Default follows fairly close to the 30% trendline before exploding in the mid 2030s.
This chart shows why companies are investing tens of billions into AI and associated technologies, including robotics. The projected economic productivity is enormous.
EpochAI predicts that we will see full automation due to AI within two decades. Here is the definition it uses:
The GATE model translates a given stock of effective compute into a degree of AI automation. This occurs by expanding the fraction of tasks that can be automated and by increasing the effective runtime compute on tasks that are already be automated. Widespread automation often happens within two decades from the start of the simulation.
This becomes a feedback loop.
As more tasks are automated, the amount of data generated increases. This is fed into more powerful compute system, which processes data at a faster pace. Improvements in data labeling, algorithms and stack efficiencies also play into this.
When that happens, the pace of output advances. Naturally, this has an effect on the next generation of tasks, which are eventually automated. For the moment many frame this in terms of computers. However, we also have to consider this will be the "brains" for robots.
Embedded AI is predicted to be a massive data generator. This is where AI actually moves in the real world. Packaged with sensors throughout, every move will be captured by the system. Here again, we see the feedback loop, generating ever more data.
Rinse and repeat.
The Age of Abundance is on its way. It is something that was evident to me for the past 7 years or so. Advancement is happening at a pace that is unrecognizable to us.
For this reason, most everything we know has to be thrown out the window. I try to stress this focusing upon the economics of what we are dealing with. We have no clue what the economic system will even look like by 2040. The entire structure of the economy is being altered.
How does this affect politics? International relations? Work? Recreation? Money? Wealth?
All of this has a large degree of uncertainty. This breeds a lot of fear, egged on by the media and posts on Facebook and X. People love to espouse the negative.
There is a potentiality of things going bad. We cannot deny that. Humans have a way of messing things up. For the last 80 years, the world lived under the threat of nuclear obliteration. Fortunately, nobody took that step (yet).
Over the last 30 years, biological threats have been another constant threat. something that was driven home over the last half decade.
Hopefully, humanity can resist the temptation to destroy itself. If that can happen, we will see economic output skyrocketing. A workforce of a billion robots alone will add enormous growth. By 2040, that will be a very small number.
Get ready for the economic singularity. It is coming.
