And on top of that the blog is not ephemeral like a conversation. Write it up as it was fresh in your head, and someone across the world might find it interesting 2 years after the fact. They might even return to it multiple times. It is almost like google scholar in that regard, a store of findings that might be useful for future doers, covering life's challenges and pursuits.
The fact that we are losing this capability of basic experience and information sharing vs. expanding it to more people is quite sad for us I think.
This is a commonly used meaning of independent study. But it isn't quite the same as Curie's independent study.
If Stackoverflow is experiencing a steep dropoff it suggests that people are more satisfied with AI. Presumably they are still learning independently with the help of web-based AI.
You should see the damage AI is doing to classroom instruction. People who are trying to learn can benefit from AI just like they could from the massive human effort of Wikipedia. People who are trying to dodge can hurt themselves with AI in the same way people hurt themselves with Wikipedia.
None of that means the web is dying.
The way marketing works is you will always see technology develop towards commercializable technology vs just genuinely good ideas. There is no great lobby out there pushing for technical blogs for example, despite their high signal to noise ratio and utility. That is done solely from the goodness of the authors heart for the most part, a cost to them in website maintenance not a profit center. You do see AI companies lobbying every government on earth right now, on the other hand, because they are working hard to entrench their tooling into the mindshare of as many people they can. It is pretty dystopian how the incentives are so aligned towards a few people who are cracking the deals. Even on the microscale they align like this: you see people on HN complain their employer bought them a copilot license they don't use because of the problems with it, but no doubt whoever secured that contract for the business probably looks great in front of the shareholders who are more concerned with keeping up with the Joneses than if the tool actually works as it is pitched. Seems we are far more concerned with the next quarter than the next ten years.