We see an opportunity at the intersection of ChatGPT and Web3 technology. The buzz around generative AI and technologies like ChatGPT and GPT-4 has become familiar, and Web3 is no exception. AI-related crypto assets are hitting historic highs, and there is even a new venture fund investing at the intersection of generative AI and Web3.
While exciting the possibilities of combining technologies like ChatGPT with Web3 infrastructure, the Web3 community must face the reality that most of the value of generative AI is captured in traditional Web2 infrastructure.
Taking that idea a little further, a controversial hypothesis emerges that should be considered. That said, ChatGPT’s momentum could have negative, long-term implications for Web3.
The core idea behind the negative impact of generative AI on Web3 is relatively simple. In short, generative AI has the potential to transform every aspect of software and content development and consumption, from infrastructure to applications.
These days, every major technology and content provider is building generative AI into their platforms. If the core of the revolution happens outside of Web3, it will affect the innovation, talent and funding gap between Web2 and Web3. Moreover, the gap will continue to widen at an exponential growth rate unless urgent action is taken. The solution to this problem is certainly not an easy one. But there are some basic ideas you can consider to get started.
Given that Web3 hasn’t produced any meaningful infrastructure or technology to support machine learning (ML) in the last decade, it’s no surprise that the generative AI movement is happening in Web2.
Web3 technologies have evolved around fundamentals such as distributed computing, storage, identity and messaging, with little attention paid to machine learning. As a result, as a matter of course, machine learning models and the like are not related to blockchain or Web3 infrastructure.
With the release of ChatGPT, GPT-4, and the image generation system Stable Diffusion, proving that generative AI can grow exponentially, Web3 has the necessary foundations to support the evolution of generative AI. I noticed that it didn’t. The problem is even more acute given the speed at which generative AI technology is evolving.
Exponential Growth and Big Tech Gap
The gap between generative AI capabilities in Web3 and Web2 is widening rapidly. Cloud and mobile evolve linearly or quadratically, with new technologies improving existing technologies. Generative AI, on the other hand, is growing exponentially.
ChatGPT, GPT-4 and others use data and infrastructure as their foundation, which is difficult for startups. In addition, generative AI grows exponentially more powerful as more people use it, allowing it to gather more data to train future versions. At this point the gap can grow unbridled.
Currently, the Web3 infrastructure does not have the necessary computing power, data, and data science framework foundations to tackle generative AI. Decentralized apps (Dapps) can certainly take advantage of generative AI capabilities by interacting with them through Web2’s APIs, but Web3 native generative AI seems a bit tight right now. As generative AI continues to evolve rapidly, the challenges facing Web3 will become apparent on multiple dimensions.
Consider a few.
Cloud platforms such as AWS (Amazon Web Services), Microsoft’s Azure (Azure) and Google Cloud (Google Cloud) are rapidly embracing generative AI capabilities in areas such as natural language, images and video. The computing power and data requirements for generative AI currently appear to exceed the capabilities of the Web3 infrastructure.
As a result, the new generation of generative AI applications will essentially run on the Web2 cloud platform and will have little to do with the Web3 infrastructure. If generative AI can live up to the expectations of many, the Web3 platform will fall far behind in terms of adoption.
Once the Web2 platform comes equipped with generative AI capabilities, there will be a new generation of applications that push generative AI to the forefront and appeal. Most of these new generation applications will work on Web2, as Web3 technology does not have the ability to enhance generative AI capabilities. While Dapps will likely incorporate generative AI capabilities, such capabilities will obviously be completely off-chain.
next generation fintech
For years, cryptocurrencies and Web3 technology have been considered the next big trend in fintech. But the focus has definitely shifted to generative AI. Most fintech platforms are more concerned with not being disrupted by new rivals powered by ChatGPT, etc. than with cryptocurrencies.
The level of innovation around generative AI technology and the popularity of technologies like ChatGPT is steadily increasing, attracting developers looking to build the next generation of apps. The explosive growth of generative AI technology coincides with the “crypto asset winter.” Combined, these two may put Web3 at risk of exuding developers to generative AI.
VC (venture capital) investment is also likely to shift from Web3 to generative AI. The 2021 bull market saw record VC investment in Web3 companies, and movements like DeFi (decentralized finance) and NFTs finally demonstrated the utility of Web3. Due to the market slump in 2022 and the explosive growth of generative AI, the flow of VC funds is shifting to generative AI. This has also led to an influx of talented people from the IT industry into generative AI.
Light of hope
The lack of a strong foundation for machine learning prevented Web3 from riding the first wave of generative AI innovation. But it’s not too late yet. Given the current state of the technology and challenges, there are two distinct areas where generative AI can benefit from the native capabilities of the Web3 architecture.
- Distributed Generative AI: Concerns around centralizing knowledge and controlling large-scale generative AI models are great, creating opportunities for decentralized alternatives. Even if the decentralized AI trend hasn’t caught on enough, it’s resurfacing the debate about the value decentralization offers, such as control, bias, and fairness in generative AI.
- Proof of Knowledge: The biggest backlash against generative AI is the potential for generating harmful, racist and bigoted content, and the risk of halcination, i.e. creating plausible content that is not factual. From that perspective, incorporating verifiable traceability mechanisms in pre-training, fine-tuning, and utilization of generative AI will be a crucial factor for its use in mission-critical scenarios. Blockchain is very well suited to bring accountability to generative AI.
These scenarios combine the strengths of Web3 and generative AI. Web3 wasn’t ready to embrace the first wave of the generative AI revolution, but it can still make meaningful contributions to its future.
The emergence of generative AI models such as ChatGPT will undoubtedly be a wake-up call for the Web3 community. Decentralization alone is not enough, we need to build a technological foundation to incorporate future innovations.
Mr. Jesus Rodriguez: CEO of IntoTheBlock, a market information platform for crypto assets.
｜Translation and editing: Akiko Yamaguchi, Takayuki Masuda
｜Original: The Next ChatGPT Won’t Be in Web3 Unless Some Things Change
The post Generation AI, negative impact on Web3 as it is[Column]| coindesk JAPAN | Coindesk Japan appeared first on Our Bitcoin News.