08/04/2025
This is the NASDAQ vs GOLD. So the perceived value of technology measured in the oldest most stable measure of value that we have.
The .com hype was huge and computers/the internet were responsible for a big uptick in productivity and efficiency, but there was a hardware constraint. everyone needed to buy a computer and needed to migrate from armies of filing cabinets. that took until about 2010, which is also when smartphones come out and the internet got way better (also infrastructure).
since then the NASDAQ has gained against gold, which essentially means it's gains haven't just been the illusion of the dollar denominator falling.
AI capabilities have been accelerating at an absolutely fu***ng astounding pace. Using AI to train AI is fu***ng insane, you have no idea. You would think that AI is already over-hyped, but I absolutely guarantee you it is well under-hyped. The productivity gains from AI have been concentrated on a few industries, like software development- because the developers are naturally going to make tools for themselves. The productivity gains are hard to quantify because they are still accelerating so rapidly, they actually sound like crypto gains. imagine boosting your productivity at work by 5-10-20x, not just 5-10-20%.
AI compute is scaling in data centres, which is a critical factor. but we already have decent internet and devices, so the rails for the AI to be useful to us already exist to a large extent. This means we almost instantly get to benefit from advancements in the tech.
A big reason for the US imposing tariffs, reducing taxes, minimising red tape, and trying to keep innovation/manufacturing in the US is that they see this boom coming. A massive disadvantage of having research/innovation in the US and then outsourcing all the manufacturing to China, is that China gets all of the innovation for free. In return the US gets cheaper products momentarily, but then very quickly cheaper competitor products spin up out of nowhere. They also sacrifice US jobs and allow a weakening economy to subsist on cheaper imported products momentarily.
It might be a somewhat controversial statement but scaling AI as quickly as possible, potentially causing some extra carbon emissions in the very short term (within reason), will result in a way fu***ng faster transition to green energy and net zero. It will also drive the cost of energy and manufacturing way down.
long story short, I think tech will see a huge boom like in the 90s and we will start seeing it if not this year, in early 2027.