What happens when AI summaries replace hyperlinks? [View all]

[Celerity]

A linkless internet

In creating anonymous summaries, AI flattens out all the fascinating architecture of thought that makes the internet hum

https://aeon.co/essays/when-ai-summaries-replace-hyperlinks-thought-itself-is-flattened





In the late 1990s, Google demolished the competition of other search engines because of an extraordinary innovation developed by Sergey Brin and Larry Page: the PageRank algorithm. While older search engines, such as AltaVista, Yahoo and (who can forget) Ask Jeeves, relied primarily on matching the terms of the user’s query to the frequency of the same and similar terms on webpages, Google found more useful websites by tracking which pages had the highest quantity and quality of incoming links. The foundational idea behind PageRank is that a ‘web page is important if it is pointed to by other important pages’. Brin and Page realised that the web is not just a lexical environment but a social one, in which links correspond to prestige, and the sites at the centre of the networks tended to be the most reliable ones. Page later remarked that other search engines ‘were looking only at text and not considering this other signal.’

The idea behind PageRank wasn’t new. The sociologist John R Seeley wrote in 1949 that ‘a person is prestigious if he is endorsed by prestigious people.’ In 1976, Gabriel Pinski and Francis Narin applied a similar approach to bibliometrics, claiming that: ‘A journal is influential if it is cited by other influential journals.’ What was new about PageRank, however, was its application to the web. PageRank succeeded because it recognised that language isn’t made in a vacuum. Patterns of words depend upon other forms of affiliation: the social and physical connections that make the web a representation of the real world.

The first page of Tim Berners-Lee’s proposal for the World Wide Web in March 1989. Courtesy CERN

One reason hyperlinks work like they do – why they index other kinds of affiliation – is that they were first devised to exhibit the connections researchers made among different sources as they developed new ideas. Early plans for what became the hypertext protocol of Tim Berners-Lee’s World Wide Web were presented as tools for documenting how human minds tend to move from idea to idea, connecting external stimuli and internal reflections. Links treat creativity as the work of remediating and remaking, which is foregrounded in the slogan for Google Scholar: ‘Stand on the shoulders of giants.’

But now Google and other websites are moving away from relying on links in favour of artificial intelligence chatbots. Considered as preserved trails of connected ideas, links make sense as early victims of the AI revolution since large language models (LLMs) such as ChatGPT, Google’s Gemini and others abstract the information represented online and present it in source-less summaries. We are at a moment in the history of the web in which the link itself – the countless connections made by website creators, the endless tapestry of ideas woven together throughout the web – is in danger of going extinct. So it’s pertinent to ask: how did links come to represent information in the first place? And what’s at stake in the movement away from links toward AI chat interfaces?

snip