Author: Meng Yan, co-founder of Solv Protocol
Editor's note: Eric Schmidt served as CEO of Google from 2001 to 2011. Recently, he was invited to the Stanford University School of Computer Science conference and gave a speech titled The Age of AI. Since he did not know in advance that the conference was broadcast live, Eric Schmidt's speech can be said to be "letting go", exposing many inside truths of the AI industry. At present, the speech video has been deleted and removed from the shelves, but the full text backup can still be found on Github. Solv Protocol co-founder Meng Yan interpreted and analyzed Eric Schmidt's speech on X from the perspective of blockchain development.
These days everyone is talking about Eric Schmidt's leaked speech at Stanford. I think the most thought-provoking part of the entire speech is the last part, which compares artificial intelligence with electrification.
Schmidt mentioned that after the advent of the electric motor, it took people a full thirty years to realize the fundamental changes brought about by this technology. Electric motors of various sizes can be made and placed in various places, thus decentralizing power.
Because he didn't elaborate on it in his speech, I went to look up the relevant background. It's like this: In the steam engine era, a factory usually had only one centralized steam engine power workshop to provide all the power. In order to transmit power to each workshop and adapt to the different power requirements of different processes, the factory usually installed a set of sky axis power transmission system. The sky axis is usually suspended under the ceiling of the factory, so it is called the sky axis. It is driven by the centralized steam engine power room and rotates above the machine. The machine under the sky axis transmits power to the machine through gears and belts, as shown in the figure below.
Schmidt said that when electric motors first appeared, people simply replaced the original steam engine power center with an electric motor power center and used electric motors to drive the axis. In other words, they only changed the performance and efficiency of the power transmission system, but did not change its structure. Thirty years later, people gradually realized that electric motors can be made into various sizes and powers, and placed in machines and equipment nearby, so that electricity can run instead of mechanical power. This is the correct way to use electric power. Schmidt believes that the distribution of electric power has triggered important organizational innovations and changed the relationship between various components, which has truly changed the world.
At this point, Schmidt seems to have summarized a process law of technological innovation triggering technological and economic changes. First, there is pure efficiency innovation, replacing key components without modifying the structure. Then structural innovation begins, from centralization to decentralization, from centralization to decentralization. Then this structural innovation triggers organizational innovation, bringing about a huge increase in productivity. We might as well call this process the Schmidt process.
According to the Schmidt process, AI is still in its early stages and is still very centralized. In the second half of the Schmidt process, AI applications will be decentralized, just like electrification. AI models are widely distributed in every corner of computing, and AI collects data, makes decisions, and executes data nearby. Only at this stage, how long will it take to reach this process? It may not take 30 years, but it may take more than 10 years.
I can’t help but think that if Schmidt is right, then investors who are investing in AI now are really Lei Feng.
So what about blockchain?
I read Schmidt's speech and felt that it could provide four inspirations for the blockchain industry.
First, according to Schmidt’s thinking, blockchain and Web3 should represent the right direction.
Essentially, blockchain is the decentralization of "autonomous computing" and "trustworthy computing". Autonomous computing is the ability to fully ensure control over one's own digital resources, including identity, data, assets, and computing processes. Trustworthy computing is the ability to guarantee to users that the results of the calculations are fair, reliable, and trustworthy, and will not be maliciously tampered with or erased. With these two things, we can disperse the key calculations involving monetary value that were originally concentrated in centralized institutions such as banks, third-party payment platforms, and social networks into smart contracts or ZK programs. In abstract terms, this process is like dispersing the power engine from a centralized power plant to various locations and devices during the electrification stage. It can be seen that blockchain is fully in line with the Schmidt process and should represent the right direction.
Second, even if it represents the right direction, it will take time to achieve real success. If Schmidt is right, the application of blockchain and Web3 should explode earlier than AI.
Third, blockchain innovation should still start from solving user problems.
Since the Ethereum killer narrative was popular with capital in 2017, the most highly valued and most watched innovative projects in the blockchain industry have basically started from solving the problems of blockchain professionals themselves, and formed a rather dogmatic concept, which has influenced the valuation of the primary and secondary markets. Everyone is talking about the big story of infrastructure without considering users. The more such projects are, the more they are popular in the primary and secondary markets. Some projects that start from the user's perspective are ignored and have no reason to complain. It's like you brag every day about how awesome the electric motor you can make is, how much the stock price should be worth for such an awesome electric motor, etc., but you never talk to us about whether this electric motor is used to drive a car, drill a hole, or drive a hard disk after it turns.
The worst consequence is that after a full decade, no real user group has been cultivated. Most of the participants in this industry are speculators, not real users. Without users, there is no motivation and direction for innovation. This is the main reason why blockchain and Web3 are currently in an innovation dilemma. To get rid of this dilemma, acquiring and cultivating users is the top priority. Everyone should think about: What problems are users willing to spend money to solve, but the existing Internet and social networks cannot solve or solve them well, and need to be solved with the help of blockchain? I feel that few people are thinking about such problems now, and most projects are going around in circles around some concepts and dogmas all day long.
Fourth, the final outcome is still the token economy. Schmidt emphasized that organizational innovation is what will ultimately drive productivity change. The token economy is organizational innovation, a reconstruction of the relationship between people, and a new collaboration mechanism. It should be said that the token economy is directly aimed at the core issue. What will happen when Web3 reaches its final outcome? If a more convenient and freer payment and financial network is created, it is certainly amazing. As Musk said, blockchain is already very useful in solving payment problems. However, I think this is just a foundation, not the most powerful part of blockchain. When blockchain-based payment and financial networks become popular, the mode of collaboration between people, people and AI, people and machines, the structure of digital economic organizations, and even the social structure of the real world will undergo fundamental changes. This is actually the token economy, and this is the final outcome of blockchain.
