Google has sent internet into ‘spiral of decline’, claims DeepMind co-founder

The British entrepreneur says search results are plagued with ‘clickbait’

Google has plunged the internet into a “spiral of decline”, the co-founder of the company’s artificial intelligence (AI) lab has claimed.

Mustafa Suleyman, the British entrepreneur who co-founded DeepMind, said: “The business model that Google had broke the internet.”

He said search results had become plagued with “clickbait” to keep people “addicted and absorbed on the page as long as possible”.

Information online is “buried at the bottom of a lot of verbiage and guff”, Mr Suleyman argued, so websites can “sell more adverts”, fuelled by Google’s technology.

Mr Suleyman was one of three people who set up pioneering AI lab DeepMind in London 2010. The company was bought by Google for £400m and it has become the cornerstone of the search giant’s AI operations.

Mr Suleyman, 39, quit Google 18 months ago and has since set up a rival venture, Inflection AI. The company is developing a conversational chatbot, similar to ChatGPT, amid a race by AI companies to usurp Google’s dominance of the web.

The entrepreneur has developed a chatbot called Pi, which he says can act as a kind of AI confidante or coach. He has raised more than $1.5bn for the new technology.

The criticism of his former employer came as Mr Suleyman told the Telegraph about plans for a new international body to monitor AI threats.

Mr Suleyman, along with billionaire former Google chief executive Eric Schmidt, plan to present proposals for an International Panel on AI Safety at Prime Minister Rishi Sunak’s global summit on the technology next month. 

The DeepMind co-founder said the panel could be “modelled on the IPCC” – the Intergovernmental Panel on Climate Change – to “establish the scientific consensus around the current capabilities” of AI.

Mr Suleyman said the IPCC, which was first set up in 1988, was a “good inspiration” for establishing a “rigorous body” for making predictions about AI risks. Other backers of the plan include Reid Hoffman, the billionaire LinkedIn founder, and Florentino Cuéllar, president of the Carnegie think tank.

The AI panel would provide governments with regular assessments on the level of danger posed by the technology.

The UK’s AI Safety Summit is due to take place at Bletchley Park and is expected to gather world leaders and tech entrepreneurs to address the challenges of “frontier AI” that might cause “significant harm, including the loss of life”.

The two-day summit on Nov 1 and 2 is expected to be attended by top lobbyists from the likes of Meta and Google. Kamala Harris, the US vice president, is expected to attend, while a Chinese delegation has been invited. 

The leaders will try to find common ground on tackling AI risks. Officials are also understood to be considering setting up an international institute for AI safety.

Michelle Donelan, the technology secretary, told the Telegraph the conference would “consider the biggest risks and biggest opportunities that come from frontier AI”, bringing together “companies, countries and also experts”. 

Concern about the technology has been sparked by the overnight success of ChatGPT, which was seen as a wake-up call for world leaders about the speed at which the technology was being developed. 

A new wave of chatbots, built on so-called “large language models”, can answer questions and have online conversations in an almost uncannily human way. They can write emails, essays, poetry or music, prompting concerns they could create upheaval in the jobs market.

Writing in his recent book, The Coming Wave, Mr Suleyman called for a “containment” of high-risk AI advances, so governments can get ahead of developments that may threaten jobs, elections and human life. 

Mr Suleyman said chatbots could “take on Google” by providing more accurate information than internet search.

However, despite recent advances, many chatbots suffer from errors and in some cases may make up false information, creating a hurdle to taking on search engines. 

Google is also developing a rival AI chatbot project, called Bard. 

Google was approached for comment. 

Meanwhile on Saturday, the Government confirmed 100 winners of £5m in grant funding for AI start-ups, including a business tackling clothing waste and a company finding a use for the technology in vineyards. 

Ms Donelan also announced a further £32m in grant funding for high-growth businesses using AI, to be confirmed after the UK summit.

Source