What is the use of a book without pictures?

A 2025 paper by Olof Sundin worries about people using Google Search uncritically. Specifically, it points out how Google has changed the way it delivers search results over its lifetime. Each change has made the sources of the search results less visible to the user. Overall, the changes have resulted in fewer stages between the user question and providing an answer, which is a good thing, in Internet terms, but a potential hazard, in that the user doesn’t know where the answer came from:

1. In 1998, when Google Search was released, the results highlighted “ten blue links”, the top ten search results, encouraging users to look at the top results
2. In 2012, Google began displaying Knowledge Panels
3. In 2016, Google started to show featured snippets
4. In 2023, Google introduced “AI Overviews”, providing answers to queries without necessarily stating the sources of those answers

Sundin uses fancier language to explain this, for example, describing  attributed sources as “materiality of information”, but we don’t need to make it that complicated, and his article is perfectly intelligible.

I don’t deny that LLMs are “shifting the focus towards searching for content rather than identifying sources”, but in many cases, this is entirely appropriate. Not all searching is for academic purposes. If I want to learn how to change the SSD in my PC, I’m looking for one practical answer, and Google can give me that.

Sundin’s pessimistic conclusion is that “increasing inbuilt AI support for searching and drawing conclusions from information sources may lead to de-skilling of searching for and evaluating information”. This kind of pessimistic comment has existed ever since reference books have been compiled, since an encyclopedia provides an answer without you having to look at the original source. The problem predates Google, in other words.

My real concern is with the presentation by Jenna Hartel, of the University of Toronto, who creates from this paper a multimedia experience for her INFIDEOS series. As if that isn’t enough, Scholarly Kitchen then gives a summary of Hartel’s summary. You could just read Sundin’s paper, or you could see it turned into a cacophony of images and sounds by (Professor) Jenna Hartel. For example, Hartel states she will do a deep dive into the paper’s content – cue for a graphic of deep dive, in case you don’t know what this phrase means, perhaps:

Having read Sundin’s paper, I find it doesn’t present any issues of intelligibility, even if Sundin as an academic wants to talk about the materiality of information. But Hartel’s treatment turns this perfectly readable text into a lot of noise and colour, in fact, literacy reworked for a generation of people who, it is assumed, cannot read text without moving images and sounds alongside. Perhaps I am neurodiverse and I have problems with sounds and animations while I read, but I would be more worried about INFIDEO and what it implies about literacy than anything in Sundin’s paper.

Following the lead of INFIDEOs, I asked Copilot to create a graphic (above) to illustrate that using Chrome without checking the sources may be dangerous. This is provided for those modern readers who don’t read the words, but just look at the pictures. After all, as Alice said, “What is the use of a book without pictures?”.