November 30, 2022.
To me, this is the start date of AI for the average person. It was when GPT3 was released by OpenAI. On that date, a switch was flipped where AI went from something discussed in the tech community to something the regular individual could use.
It also showcased the potential of where things are going.
Fast forward 38 months and we see an entirely different discussion. The leading thinkers in the field are now looking at things in a much more advanced way. We see the bar being raised as early stage utility is realized.
Much of what is taking place is still happening behind the scenes. The mainstream media talks about the lack of real use cases. This can be misleading. AI is being used in many fields, creating advanced outcomes as compared to humans.
Since this is being applied in areas such as drug discovery, the general public is not aware of it.
What will become evident is when the solutions to a problem (such as disease) is realized.
So what are we looking at?
In this article I will discuss the concept of hyper-civilization, something we are woefully ill-prepared for.
Here is a video worth watching that was given at a TED Talk. It goes into some of the concepts that I will be bringing up.
Dyson Swarms? Taking planets apart.
On this one, we will pass for now. That is outside the scope of what we are discussing.
However, what happens if AI is able to provide the cure for 5,000 diseases? How do governments handle that? It is impossible to spin up that many trials in a short period of time yet the saving of millions of lives could be the outcome.
We often touched upon the topic of money. In an age of abundance, what does money really mean? On the scale, we are moving away from the traditional model we operate under now towards the Star Trek Replicator. The latter is a long way off in my opinion but we will be closer on the spectrum to that as compared to where we are now.
The Internet overwhelmed people with information. This is certainly still the case.
Our next challenge is already upon us. We are swimming in AI and it will only get worse. For the moment, AI is still resident mostly in data centers. That will not remain the case for long.
The cost of compute is dropping fast. It is a fact that changes the entire economic equation completely. We never encountered something of this magnitude.
Timelines are speeding up.
Forecasts that are decades into the future might only be a few years off (or at most a decade).
The Experience Curve was mentioned in the video. This is something that many technology (AI) companies are looking at. It is a way to gauge the change in AI similar to what Moore's Law did for computing.
OpenAI concludes that we are seeing a 40x reduction each year in the cost of intelligence per dollar. As stated, even if that is off by 4, a 10x, we would see a 1000x improvement in 2 years.
That means, by early 2028, we could be looking at something 1,000 times more powerful for the same money. How does that alter society?
Of course, the world of atoms can be slow to build out. It is much easier to deploy things in the digital world. When entering the physical world, the challenges mount. This is why I think some of the predictions in this video might be a bit aggressive.
Nevertheless, it is insightful into the pace which things are happening.
Over the next decade, we will likely reach hyper-civilization. That means society will be changing at paces we cannot comprehend. This could result in unemployment rates of 40% or higher as jobs are eliminated.
Deflation, not inflation, will be the norm. The amount of "money" required will be unimaginable. Central banks are going to have to "print" like never before simply to combat the massive downward pressures created by technology.
Of course, this will be a positive. The fact is that many things will be getting cheaper as technological deflation eats up more of the economy.
Just consider this idea:
TSMC chips have a 50% profit margin; Nvidia 80%. Elon Musk is looking at building his own chips for the multitude of companies he runs. Think of what that saves on the cost of compute.
Solving bottlenecks such as energy will also be necessary. This is why the discussion of data centers in space, something unheard of 6 months ago, is now in the conversation. The merger of xAI and SpaceX shows this is where some are looking.
The mobile phone changes society in many ways. It did this over the course of a decade. Now we are looking at things happening in just a few years.
By 2035, my guess is everything is completely different.
