The father of the web asks that AI preserve the “original values” of the internet

The prevalence “of the person, of the individual” was at the core of the creation of the internet and should also apply to AI, he told AFP in an interview on the sidelines of the SXSW technology festival in London.

LOOK: China creates a unique digital ‘ID’ for humanoid robots to ensure their traceability

This British physicist turned computer scientist conceived of the World Wide Web (WWW) while working at the European particle laboratory CERN in 1989 as a communication mechanism between scientists.

“It is important that people use this technology (AI) to make sure that their clients, their citizens, have control over their own data,” he defended.

“AI models are a different layer” within the internet, “they take advantage of the fact that the web contains so much data to train themselves,” he said.

He praised that technology as an “exciting” advance, although he believes it would benefit from regulation.

In particular, he noted that AI does not yet have anything similar to the World Wide Web Consortium (W3C), the international organization he founded that sets standards for the Internet.

As a result, AI pioneers are not “benefiting from the collaboration that they would get if they had something like that,” he said.

The web, free for everyone

Berners-Lee originally proposed his invention, which would later change the world, as a way for researchers around the world to share information about their findings.

He called this new network the World Wide Web and, in 1990, he joined forces with the Belgian Robert Cailliau to develop it.

It was based on two pillars: the HTML language, a code that allows the creation of a website, and the protocol for the exchange of hypertext HTTP, the system that allows the user to request and then receive a web page.

Determined to make the web freely accessible to all, he did not patent his program, which ensured that it took off and spread quickly.

LOOK: Microsoft announces its first artificial intelligence model: MAI-Thinking-1

As the use of personal data by AI models worries authorities, especially in Europe and the United States, Berners-Lee has made its protection his main cause in recent years, particularly through the startup Inrupt, which he co-founded and launched in 2018.

“Without data, they (AI models) cannot exist. And now they have had unrestricted access to everyone’s data, and if we are not careful, we are going to get into a really serious situation,” warned the company’s other founder, John Bruce.

Inrupt relies on secure data wallets that remain in the hands of users.

He is also working on creating an AI assistant called Charlie that will be able to filter user requests to tools like ChatGPT or Claude.

“When you ask a question (…) analyze what the question is (…) and decide what information to send” to the AI ​​tool, Berners-Lee said.

If there is personal information there, it will “modify” it so that the AI ​​tool “gets an idea (…) but then can’t really use it to identify you,” he said.

“Charlie is about preserving the original values ​​of the web,” he added.