The Showdown between AI and Web3?

Since OpenAI launched its chatbot program, ChatGPT, and Microsoft integrated similar functions into its search engine, Bing, investors have been eagerly investing in artificial intelligence technology. Some even predict that funds will flow from the cryptocurrency market to AI projects, further impacting blockchain-related businesses that have recently been hit hard by the US financial sector. As the two revolutionary technologies of our time, is AI really in direct competition with Web3 (which includes blockchain, applications built on it, and cryptocurrencies)?

While many people can talk at length about artificial intelligence, their understanding of blockchain is often limited. The general public’s impression of blockchain is mainly associated with the meteoric rise of Bitcoin and the exorbitant prices paid for NFTs. That is to say, they only see it as a speculative tool. Even many tech industry professionals, including the father of the internet, Tim Berners-Lee, view blockchain as something created out of nothing by a few people, likely just a bubble or even a scam. On the contrary, artificial intelligence is considered the technology that will truly change the fate of humankind, as expressed by Bill Gates and others.

The reaction to the “miracles” created by artificial intelligence in language, art, and music generation is polarized. Some are amazed, excited, and embrace it, while others are skeptical, resistant, and fearful. Both responses have irrational and emotional components. The problem with the former is excessive credulity (believing that AI already possesses consciousness and emotion), and the problem with the latter is exaggeration of facts (that humans will soon be replaced or even destroyed by AI) or underestimation of the facts (that AI’s capabilities are limited and it can only create low-quality, lifeless things that cannot compare to humans).

In addition to its practical applications, AI has also ventured into the realm of artistic creation. Generating works that mimic the styles of masters and producing works that are almost indistinguishable from real ones are now easy tasks for AI. As language generators continue to improve, poetry and even novels can also be created by AI. The main problem is that AI views art simply as a “technology,” and that is all it can be, given its nature. Its method of acquiring technology is through constant imitation and learning, which allows it to reach a high level of proficiency in repeating existing techniques, but it is unable to truly surpass and innovate beyond them.

Of course, it is possible to break down general creativity into some principles and try to apply them to generated programs. However, true artistic creativity is not based on general principles such as “think out of the box”, but is a response to the world that arises from the unique experiences and feelings of the creator in a specific historical and cultural context. Mastering and simulating this concrete state of life is extremely difficult for AI. It first needs to be conscious of the state and more importantly, it must have the source of creativity – the unconscious or instinct. Skipping these, what is generated is just an imitation of creation, and even if it achieves some technical similarity, it is not creation itself.

I do not underestimate the potential of artificial intelligence, nor do I assert that AI cannot achieve the above-mentioned creative work. The real problem is that before artificial intelligence reaches this level, technology may have already replaced art as the criterion for value judgment. This is not a problem with AI, but a problem with humanity itself. It is us who proactively view technology as art, forgetting what art truly looks like. The most likely scenario in the future is not that AI replaces or destroys humanity, but that humanity regards AI as the criterion and AI-izes itself. At this point, humanity will lose its instinct and become a follower of machines, and creativity will also wither away.

Another recent point of controversy is the showdown between the “traditional” search engine, Google, and ChatGPT. Some even claim that the latter will completely crush the former and take its place. Regardless of their search capabilities, the biggest difference between the two is that Google is focused on the “source,” while ChatGPT (or Bing) focuses on the performance of the information in the “results.” Although we usually say “search results,” the search engine actually only helps users find where the search target is located on the internet, which is the source or origin of the data. Moreover, there is not just one search result (source), and users still have to rely on their own judgment to distinguish and choose. In contrast, ChatGPT does not tell users what original data its answers are based on; it only tells you the “results” it has searched and synthesized itself. I don’t know if users have the advanced right to check the source of the information, but even if they do, when we get used to this kind of secretary mode, we will be too lazy to verify the results. (If we need to verify everything, what is the secretary for?) This will lead to the crisis of human dependence on AI and losing the ability to think independently.

Another problem with artificial intelligence is high centralization. Of course, there are many levels and sizes of AI research, but for AGI (general artificial intelligence or strong artificial intelligence) like ChatGPT, only a few countries, companies, and teams have the relevant development capabilities, and the results are often monopolized by large organizations. In the chatbot competition, only a few oligopoly companies such as Microsoft and Google can participate. Although the original intention of OpenAI’s establishment is technical openness, in practical popular applications, large organizations have the advantage of monopoly. The main applications are often provided by a few suppliers, and in some countries, they may even be controlled by the government. The so-called AI controlling humans is not just a technical issue but also a systemic issue. Large AI systems and monopolistic systems have inherent matching conditions. Artificial intelligence is not based on the concept of the individual; it is an integrated system. No matter how many people use it, no matter how users can customize the service, ChatGPT or Bing only have one. Even if Google’s upcoming Bard is added, there are only two. (I don’t know if Baidu’s Wenxin counts.)

What artificial intelligence brings us is efficiency and capability, but we are easily tempted to surrender individual autonomy due to convenience and laziness. In contrast, in an increasingly institutionalized society, decentralized blockchain technology offers us an opportunity to regain autonomy. Unfortunately, for now, blockchain is still dominated by speculation and faces suspicion and repression from existing political and financial systems, and its true potential and meaning have yet to be fully realized. More importantly, Web3 requires individual autonomy and participation, in contrast to AI which encourages passive gain without effort. The former is bottom-up, while the latter is top-down. This distinction is where the true confrontation lies.






Leave a Reply

Your email address will not be published. Required fields are marked *