If you place AI on humanity's historical timeline, we are in the bronze age. The stone age was that whole long stretch when nobody put real money into AI and most of it was not in actual use, only in certain use cases. Then bronze arrived, and after that there was nobody who didn't know about it.
The stones
The stone age is by far the longest of all the ages. It was the same with AI. Spam filters, recommendations, image recognition, things few people even called AI while they were using them. Simple tools that each did one thing. Compared to today's systems they were stones.
Then came bronze. Copper and tin melted together into something neither one was on its own. Today's systems work the same way. A powerful model, integrations and tools around it, put together into a whole. The material can be used with almost anything for almost anything. It is expensive, and it still changes the whole world fast.
The bronze started being cast in 2017 with Google's paper "Attention Is All You Need", but it only broke out properly with ChatGPT 3.5. Before that most people didn't know they were using AI. After it, it was impossible not to know.
The questions nobody has time to ask
In the middle of a gold rush, some questions never get asked at all. Vicki Boykis wrote about local models in June and asked, roughly: if performance and price actually constrain us, what architectural tradeoffs do we need to make? (Boykis 2026) It is a question almost nobody has asked during the token gold rush. Bronze was expensive for a reason. Tin was not everywhere, it needed long trade routes, and that is what set the limit in the end.
When it has to pay off
Around the 2010s, companies started to seriously use their technology and market position to make money. That would explain why many people feel technology after 2015 improved more slowly, and why everything filled up with ads and microtransactions. Now everyone is asking when the AI investments are going to pay off. The difference is that last time it was mostly already rich investors' money on the line. This time the whole world goes down with them if it doesn't work out.
The iron
During the bronze age there were probably not many who thought an even more effective material would show up in such a short time. The bronze age turned out much shorter than the stone age, and the next step came faster than expected.
Iron isn't really better than bronze. Pure iron is softer. What made the difference was availability. Iron ore was almost everywhere, tin was not. So iron reached out into societies in a way bronze never did.
That is how I would guess the AI iron looks. Everyone has access, it isn't that expensive anymore, and it can still be used for a lot of different things. But the value is in AI being implemented the right way into everything we already have. Not dumb chatbots, but specialised tools that actually work and make life easier for the person using them.
We don't know what that looks like yet. That is exactly why I think we are still in the bronze age.
— Niklas