The old web is being destroyed by AI, while the new web struggles for birth
AI has surpassed the capacity of internet to scale.
In broad strokes, the problem is this. In the past, people used to create things on the internet. Then, they started making homepages, mailing lists and forums. They also made a little money. Then, companies realized they could do better. The companies created feature-rich and attractive platforms, and opened their doors to anyone who wanted to join. We filled in boxes and placed them before the public. Companies chased scale because, once a large number of people gather somewhere, it’s easy to make money. AI challenges these assumptions.
AI systems, especially the generative models that are currently in fashion, scale easily when given money and computing power. They are able to produce a lot of text, images, and music, and will soon be able to do the same for video. They can outcompete or overtake the platforms that we depend on for information, entertainment, and news. These systems are often of poor quality and built in a manner that is parasitic on the internet today. These models are trained using strata of data that were laid down in the previous web age, which they reproduce imperfectly. The companies scrape data from the web to create machine-generated content, which is cheap but not as reliable. The product competes with other platforms and users to get attention. Users and sites are trying to figure out how they can adapt to these changes.
Source:
https://www.theverge.com/2023/6/26/23773914/ai-large-language-models-data-scraping-generation-remaking-web
