Many are discussing the prospect of us achieving artificial general intelligence (AGI) and how soon it will come. While estimates are still varied, one thing is noticeable. In surveys among those involved in the field, the time of consensus keeps moving closer.

It was not long ago where the majority thought it would be around 2100.

I have not seen anything lately but some prominent figures are starting we will get there in either 2026 or 2027. Of course, we have the issue with getting a concise definition of what AGI is. Basically, this appears to be when computers are better than the top expert in all fields.

Even if we do not achieve that, even 95% there is enormous.

So what will the world look like at that time? Let us dive in and see.

Image generated by Ideogram

The Path To AGI And What It Means For Crypto

Sam Altman laid out certain fact in his most recent blog post. He is one who believes it is much closer than the masses are expecting.

There are three trends that are holding true even as the amount of compute keeps expanding. They are:

1. The intelligence of an AI model roughly equals the log of the resources used to train and run it.

1. The cost to use a given level of AI falls about 10x every 12 months, and lower prices lead to much more use.

1. The socioeconomic value of linearly increasing intelligence is super-exponential in nature.

What is written here is important because it provides a recipe along with the impact of what is output. Since we know this, we only have to monitor whether this is still holding true. That becomes the only question. If there is a change in one of the trends, that we would have to adjust expectations.

The first tells us that money in = units of cognition out.

There is a correlation between the amount of money put into AI systems and the scaling of it. This is actually a declining correlation in that the return for each dollar is diminishing. By this, a 10x increase in spending does not equate to a 10x in model advancement. Eventually, this might end but it is not forecast anytime soon.

One thing to keep in mind money does not necessarily mean compute. That is the bulk of it, both for training and inference. However, we are also talking about personnel and advancement in algoirthms or AI architecture. Deepseek is an example of that.

Our second point is that we seeing a massively deflationary cost curve. What this tells us is that a model that is a year old will cost 1/10th to train today. So while Big Tech races ahead to be the most advanced, a year behind can still yield powerful systems.

Finally, the impact is seen across the board. Even with linear improvement from a socioeconomic standpoint, the impact is super-exponential. I think that will become evident when we start to see AI agents built by the millions.

Post Monetary Society

Does AGI mean we are in a post-monetary society? In my view, the answer is no. It does mean, however, that crypto is crucial.

To start, AGI deals with cognitive ability. For us to truly realize an autonomous economy, we have to have both the digital and physical world covered. This means billions of robots handling real world activities. That is something that is at least 10 years out. Hence, we have a gap between the two.

For this reason, money will not change. We are still dealing with a medium of exchange that will occur, probably with greater frequency. That said, there will come a time when this is in question.

Most realize, by now, that I do not think most crypto is currency. Outside of stablecoins we are really dealing with crypto-assets. This is the key to understanding what the future looks like.

AI agents will utility stablecoins to transact. Some of these will achieve such denominations that can allow for micropayments. These will occur by the trillions. The rest of crypto does not serve this purpose.

The impact here is in value distribution of the wealth generated by AGI. We are going to see massive impact economically. After all, it is exponential. When robotics reaches the point of fully automated labor, we could be facing an economic singularity. That means we have to question who owns the means of production?

Crypto-assets solve this. When we attain AGI, anyone will be able to build AI services on a network which he or she has stake. This is the major opportunity. In an era where "local" community is, once again, the key barometer, building financial value within that is crucial.

Having Stake

Predicting the future is rather difficult. At best, it is a fool's game.

What we do here is to find the trends, then extrapolate them out. There is the ability to influence the future by our present actions. If we have some idea where things are heading, at least with a high degree of probability, then we can align said actions with the path we are heading down.

My conclusion from this is that we are seeing labor capital being threatened. This means incomes will dry up. The classic economics model of labor and capital is becoming one sided. As Altman shows, more capital equals greater cognitive output. This is rather disruptive to the economy as we know it.

Hence, people need a new outlook on things. Having stake in these systems will be crucial. What this means will differ amont individuals. One could own stock in a Meta or Google. That might help provide exposure to the growth that is expected.

Or, perhaps the focus shifts to Web 3.0 and people achieve stake in permissionless networks which will capture the value. This is, in my opinion, the most likely path for the masses.

Either way, having some holding in assets is crucial. Capturing the value generated by systems in an AGI world is vital. Even more important, is the distribution of it.