I have seen a theory before, three stages of innovative technology: creation of new technology - elite services - popularization
Three-stage theoretical model of technological diffusion
- 1. Innovation Monopoly Period (elite creation stage)
In the early stages of the technological revolution, innovation activities relied heavily on knowledge-intensive investment. The development of AI presents the characteristics of "hierarchical technology", and the breakthrough in core algorithms is still led by top scientific research teams. - 2. Value capture period (service elite stage)
In the process of commercialization of new technologies, beneficiaries have shown an "inverted pyramid structure". AI is used first for a few people, such as the current KOLs, etc. - 3. Inclusive spread period (popularization stage)
When the technical infrastructure matures, more people will use it. Some typical features include improved capabilities, reduced costs, and ecological openness. DeepSeek is all available.
At present, it is mainly in the second stage, developing rapidly towards popularization.
What should I do at this time? One end of the supply chain has matured and the future direction has been clear. More applications are needed to dive into more scenarios.
The characteristics of this revolution start from the supply side, with "shells" in the middle, one after another, gradually plunging into the user scenario.
Instead of the opposite direction. The improvement of AI, especially the huge improvement of AI Coding now, is not an obvious demand-driven.
On the surface, it seems that innovative technology has created "new demands", which are essentially existing demands and have a more efficient way to implement them.
What you need at this time is not to tell users how to solve this problem and that problem using AI? The user's perception of the replacement upgrade brought by this underlying technology is not direct. The best way is to try faster with the help of improving application development efficiency. Directly display the solution effect to the user, let the user feel it by themselves, and then Wow.
I have used Lovable a lot recently, and it is an AI application development platform. The scenario I use the most is as a tool to communicate with the business teacher. It greatly mentioned efficiency: whether you want to make a prototype quickly, where you meet and where you don’t, it is clear at a glance. One is better than a thousand words, and one quickly builds an AI prototype that is better than ten thousand words.
Regarding where the application opportunities are, I think the core is to penetrate into the scene. In the past, due to supply technology issues, we had to reduce the edges of demand and vaguely classify demand, so as to provide an application and build a website.
But now there is no problem with the technology side. You can enter the demand scenario deeper and restore the demand edges that have been reduced by twice the reduction. It may be called long-tail demands that are not suitable. In essence, times have changed, and every demand is worth meeting.
For example, Lovable and Monica are both products that can be split into many scenarios to achieve the ultimate goal, even if there are only 1,000 users, they can survive.
In the AI era, every need will be met.