By Jake Brukhman

Compiled by: TechFlow

Photo credit: Sofia Vaschetto

Since we first wrote about the intersection of Web3 and AI in 2022, this rapidly evolving technology trend has become a major feature of crypto conferences this year. However, it’s not often that AI events emphasize decentralization, and it’s even rarer that the two sides are so closely integrated at a single event.

SuperAI was the flagship event of Singapore AI Week in June this year, and it was probably the first conference I’ve attended that started to bridge the cultural gap between traditional Web2 founders and projects, and developers breaking new ground at the intersection of Web3 and AI.

The conference was well-organized, with a large global attendance, and felt similar to the Token2049 crypto events held twice a year in Dubai and Singapore. People flocked to the main stage to hear investor and former Coinbase CTO Balaji Srinvasan's speech on the complementarity of Web3 and AI. This was followed by a robot demonstration by Boston Dynamics founder Marc Raibert, followed by former StabilityAI CEO Emad Mostaque announcing his new project Schelling AI, an AI-powered cryptocurrency. Attendees from all over the world - mainly from Asia, the Middle East, India, Australia, as well as many Western countries - gathered to absorb the colorful AI and crypto knowledge.

While I talked to some of the most cutting-edge teams working to bring AI production processes to the blockchain - such as Kiva, Pax, Ritual, Sapien and Schelling AI, I also talked to veteran AI entrepreneurs who have long been pioneers in machine learning and AI at traditional companies such as Meta and Microsoft. SuperAI also held the Genesis Startup Competition, where early AI companies presented their products on stage to investors and competed for awards. Among 700 applicants worldwide, OpenOrigins, a blockchain trust layer project focused on media provenance, won first place.

Theory in practice

SuperAI proves that our 2022 theory is not only possible, but will provide an alternative, more democratic foundation for building this technology. Many startups have already begun building AI infrastructure using Web3 primitives, and funding momentum in the space is increasing.

At CoinFund, we often say that AI needs Web3, not the other way around. However, the success of Web3’s role in AI will depend on how these two perspectives are culturally combined. At this gathering of some of the brightest minds in cognitive computing, we’ll hopefully see the two groups move closer to each other, acknowledging their ability to build more open, more democratic, and ultimately more secure AI technology by bringing the two fields together.

For developers, Web3 hopes to democratize resources and allow more people to participate in AI research. It will also ensure that open source AI continues to be the main source of innovation in terms of capabilities, explainability, and consistency. For users, Web3 data can ensure that agents have self-sovereignty and ownership of their digital identities. In terms of intellectual property, it will prove that users and copyright owners are protected and appropriately compensated for their creations.

This trip was a necessity for me as someone who has been investing in AI in the Web3 space since 2020. I’m excited to bring a crypto-native perspective to the conference and work with Web2 AI founders to bridge the cultural gap within the ecosystem.

reward

While a lot of innovation has occurred in decentralized AI over the past two years, it is still early days. Most of the progress to date has been in laying the infrastructure foundation for decentralized AI, while the application layer is still in its early stages. On the infrastructure side, the most concrete capability is GPU compute aggregation. For example, io.net, a company that creates instant access to GPUs, sponsored SuperAI. Decentralized training is still in its infancy and remains a challenging and controversial topic, although several companies, including GenSyn and Pax, have made important contributions in this area.

Asia continues to play an increasingly important role in Web3 innovation and adoption. Many technology projects are based in Singapore and Hong Kong, and strong machine learning talent from around the world is gathered here. CoinFund has expanded its team in Asia over the past two years, and about half of our portfolio is based outside the United States. We are excited that more builders are joining the global crypto ecosystem every day.

More viable infrastructure will help people build applications that generate new capabilities, pushing AI development to new heights. Leaders like OpenAI and Anthropic often talk about general artificial intelligence (AGI), however, the real practical "danger" of AI in the short term may be that it will lead to the automation of some, say, 20% of jobs. More broadly, most people I talk to are less exaggerated about the dangers of AI. I have spoken to leading figures in the field who give less than a 10% chance that AI will pose an existential risk in the short term.

Reflecting on my week in Singapore, it’s clear that the intersection of Web3 and AI is here to stay, and innovation will continue to grow as both industries work together to form fruitful projects and advance the state-of-the-art in decentralization.