The future is very difficult to predict.
It is said the prediction game is one of simply being the least wrong. None of us have a crystal ball so being clear about how things will unfold is impossible.
That said, we can look to trends, hone in on the pace of advancement, and isolate our focus to that which is starting to have an impact to have some idea where things are going.
Of course, when stated in the moment, these ideas often seem absurd, especially to the general public.
Here is an example from Reddit where a comment was downvoted and the one that basically laughed at it was upvoted.
Naturally we are not dealing with huge numbers yet it is indicative of how things often go. The masses laugh at the idea as being absurd.
By the way, in this instance, the original poster was absolutely correct. Today, we have the ability to code photorealistic video with a sentence or two. Over the next year, we will likely see a large jump in both quality and length of these videos.
There are many reasons for this including demographics. Those who were reared in the era where change was slow find it more difficult to adapt to the new pace that was set. Even the early years of the Internet seem slow compared to right now.
Technology is accelerating. We are in a period where the digital is affecting the physical. This means that we can see entire industries flipped in just a few years.
Video conferencing is doing a number of commercial real estate. Have you seem what the values are in major cities? This is due to occupancy rates of roughly 50%. The idea of being present in person is no longer guaranteed.
Technological progress advances at a pace faster than human embracing of said technology. That, however, is condensing. People are more accepting of newer technologies, with adoption rates speeding up.
Ultimately, this entire conversation boils down to exponentiality. humans are not wired to think in this manner. Our natural tendency is to look at things in a linear fashion. We can grasp 2% growth in our economy, and extrapolate that out 25 or 30 years.
When we make a statement that the economy could grow at 30%-50% annually, people's minds are blown. They simply do not believe it is possible.
Nevertheless, if we look at exponential impacts, this rate of growth, called the economic singularity, is not only possible, it is likely at some point. If I had to guess, I would say this will be towards the end of the 2030s.
Do you have the inclination to state this idea is foolish. Be careful as we know some things do not age well on the Internet.
The one posting about video was certain he/she was right that we might see that in our great-grandkids times. I guess that family can pop out a generation each year.
The digital world is taking over and it is one of compounding.
Whatever we focus upon within this world, we see expansion. Compute, images, video, memory, and bandwidth all keep growing. The Internet is the world's largest copy machine.
Coding is the same way.
The hardest development is the first line. Once the first feature is built, others can work off this. Over time, companies increase their output on their platforms simply due to the fact other code is already in place.
When it comes to AI, this is placed in the express lane. The capabilities of the systems are impacted by the data, algorithms, and compute. A single change can have significant effects. That said, when there is improvement in each area, we have a compounding effect.
In general, infrastructure is the hardest job to code.
To start, it is mundane with few rewards. The masses are not excited about infrastructure. At the same time, it is a smaller part of the developer field since it normally entails certain abilities the average coder doesn't have. This means working in advanced languages which are not commonly used for, say, website development.
Once, however, this is complete, then others can jump on top and start building. Here is where the numbers flip and many more can join in.
The result here is a pace that far exceeds the physical world. This is also changing with the potential of robotics.
When looking at what is taking place, the moral of the story is two years is a lifetime.
The pace of change, especially regarding AI, is such that we will not be in the say place 2 years from now. The capabilities of these models will be such that society will be in the process of being completely transformed.
Again, this might sound absurd but then we have documented evident of how this was viewed 3 years ago. Is what I am stating here any crazier than the assertion put forth then?
As mentioned, ChatGPT was released in November 2022. Look at the advancement in under two years. The difference at this moment is we are operating from a much higher baseline.
What preceded is not lost.
That means, even advancement that mirrors the last 22 months will put us much further ahead as we start to close out 2026.
What happens if there is acceleration in many areas?
We have all seen exponential growth curves. Here is a basic chart of what it looks like.
When it comes to generative Ai, we are at the early part of the curve. We have not hit what is commonly called the "knee", the point where the curve starts to go parabolic.
My view is we will hit that with the next generation of models.
The point here is to start trying to think in these terms. It is unnatural and difficult but it is the only way to stay abreast of what is going on. We are looking at a rapidly changing world due to technology.
And it is not slowing down.
The masses are going to miss what is happening and join in only when it is too obvious to be ignored. That is how it works.
They miss most of the advancement that is taking place right in front of them.
While none of us will be accurate on everything, we are in a much better place if we simply stay ahead of the curve in our understanding of where things have a high probability of heading.
