Web3 can compete in the computer chip race.

The shortage of graphics processing units (GPUs) will drive the widespread adoption of Web3 infrastructure. High-end GPUs like NVIDIA A100s and H100s are necessary to train artificial intelligence (AI) models. However, these GPUs are expensive, have a limited supply, and are only needed for a specific period of time for each model. As a result, many startups cannot afford to own them.

Amy James and Devon James are co-executive directors of the Web3 Working Group.

There are also concerns about AI safety, including the potential harm of deep fake images and the fear of a malevolent AGI (artificial general intelligence) similar to the Terminator. Decentralized physical infrastructure networks, known as DePIN, provide a solution and will make AI and Web3 more accessible and safer.

Despite existing for almost a decade, Web3 has only been adopted by a small group of crypto enthusiasts. This is surprising considering that the original web had to start from scratch, while Web3 has the advantage of building upon the existing web. The reason for this limited adoption is that mainstream users do not have a strong need for decentralized social networks or protocols like decentralized file storage.

• Tokenized assets are safer during a banking crisis.
• Australia’s CBDC lead is not focused on addressing current issues.
• CEO predicts a brighter future for all as LayerZero’s market undergoes a shift.

The emergence of AI will change this situation and lead to the mainstream adoption of Web3 infrastructure. DePIN protocols offer solutions to the GPU shortage and data challenges faced by AI startups. AI startup founders have the technical ability and motivation to overcome the barriers that have hindered mainstream adoption so far, such as the complicated process of buying tokens. Protocols like Akash provide decentralized marketplaces for GPU owners to rent their resources, democratizing AI innovation and allowing hardware owners to generate passive income.

In addition, Web3 protocols offer features that are not available on the traditional web, attracting AI startup founders to stay on these protocols even after they have obtained the necessary GPUs. Protocols like Arweave disrupt the data oligopoly by providing permanent data storage with a one-time payment model, which enhances trust in AI models.

Currently, AI products from large companies like OpenAI and Google are leading the way, but open source versions have a competitive advantage. Open, publicly-available training data for AI models is not only more trustworthy but also of better quality. Open models can exchange information and improve their learning, resulting in better versions of the models.

Furthermore, open source frameworks like zkML (zero-knowledge machine learning) offer cryptographic proofs for AI models, verifying that they were trained on specific datasets and enhancing the credibility of their output.

While some fear that AI will lead to a dystopian future, DePIN provides the solution by offering trustworthy code and decentralized physical infrastructure networks. Building apps on infrastructure controlled by internet giants like Amazon Web Services (AWS) not only makes it difficult for startups to compete, but also exposes them to counterparty risk.

If a Web3 app is successful it could be seen as competition by the Amazons and Googles on which they depend, and therefore might be cut off from the infrastructure they need. DePIN offers founders infrastructure that is more reliable, resilient to attack and much less expensive.

And the benefits of DePIN for AI startups extend beyond training data and hardware access. As deep fakes become increasingly common and erode public trust in online content – such as the “Pentagon explosion” photos, which caused a brief $500 billion market reaction – Web3 protocols provide the solution via public key cryptography.

To those of us in the crypto space the benefits of cryptographic signatures seem obvious, but AI doomers have yet to discover this solution. Public-key cryptography provides unique, unforgeable signatures which can be used to establish provenance and evaluate the trustworthiness of content.

See also: 10 Ways Crypto and AI Can Make Each Other Better

The GPU shortage could end up motivating startup founders to use DePIN protocols, but the features these protocols unlock will help the apps built on them win against Web2 incumbents. AI apps built with decentralized infrastructure will have the advantages of optimized job routing to reduce latency, faster and lower-cost commerce via micropayments, all kinds of smart contract functionality and automatic filtering of data to keep deep fakes and spam out of our feeds.

DePIN offers better access and features at lower costs to AI startups, creating fertile ground for development and competition. Users of these emergent AI applications will become users of Web3 protocols, often without even realizing it.

This behind-the-scenes integration will introduce users to the benefits of the decentralized web, providing a seamless experience that helps to overcome the traditional hurdles of Web3 adoption. AI and Web3 are not in competition, they need each other like peanut butter and jelly.