The Internet computer A blockchain project that has tried to differentiate itself from the rivals, is being doubled in its tone as the network to go to artificial intelligence in the chain (AI).
This could be the beginning of a new technological pile, one in which AI, not humans, becomes the main application developer, according to Dominic Williams, founder of the Internet computers developer Dfinity.
Williams argued that, although cryptography prices remain largely driven by market mechanics (treasure operations, liquidity games and speculation, underlying technology will eventually force a calculation in an interview with COINDESK.
“In the long run, markets begin to reflect realities in the field,” he said. “But so far you are not seeing what is happening with the Internet computer reflected in the price of ICP.”
Executing AI in the chain
The Internet computer first demonstrated neuronal networks that are executed as intelligent contracts in April last year, beginning with the classification of images and posterior facial recognition, said Williams.
While these were relatively simple models compared to large language models, of the type that feed ia tools such as Chatgpt and Gemini, they were concept tests: that AI can be executed natively in a block chain. No other network has achieved this, Williams said, despite the talk about “decentralized.”
Where others trust infrastructure outside the chain as Amazon Web Services, ICP seeks to integrate the complete development and execution pile in the chain. Williams describes this as “a self -enfatable internet”, a system where users describe what they want, and an AI gives it as a job application, hosted directly to the Internet computer.
The greatest idea, said Williams, is that the AI ​​itself will replace much of today’s developer’s workflow. Instead that humans write code, configure databases and maintain servers, an AI could turn the applications instantly, update them continuously and guarantee resilience through blocks based on blockchain.
This rethinks the block chain not only as a settlement layer for tokens, but as the optimal environment for applications generated by AI. The ICP design, with characteristics such as “reverse gas”, the model where developers pay the computer costs of their applications, instead of demanding the end users to pay a transaction rate, eliminates the need for firewalls or migrations of databases that affect the traditional infrastructure.
“IA is developing these applications hundreds of times faster than humans,” Williams said. “And because there are no standing system administrators, it needs the railings that only blockchain can provide.”
Williams pointed to the first Hackathons, where common people used AI in ICP to build applications: from a CrowdSourced potholes mapping platform to a tool to generate health and health directives.
The vision is that such tools could proliferate in millions. Entrepreneurs, small businesses and even NGOs could create personalized applications without technical experience, paying the use of Fiat, while cryptographic tokens underpin the system behind the scene.
Price action still delays
Despite these developments, Token ICP has not seen a sustained impulse. It was briefly recovered when AI Integrations was announced last year, but since then it has negotiated more in line with a broader feeling of the market than with the adoption of the user.
Williams accepts this disconnection, but predicts that markets can catch up very soon.
“This could be the first time that web3 really exceeds the web2 technologically, without an incentive in sight,” Williams said. “The shock will be when people realized that they can talk with an AI, and a blockchain application appears in a URL.”
