Cover image: ‘Print killed the priest and AI brought the priest back’. A reader’s arrow goes to an AI answer standing inside a church window, while the sources behind it stay unopened.

Print Killed the Priest, So Why Are We Letting Someone Read for Us Again?

In 2007, Google estimated that fewer than 1% of searches used the ‘I’m Feeling Lucky’ feature, which directs users to a single page. Currently, one answer is default. When Google displays an AI-generated summary, users click on a search result in only 8% of visits, compared with 15% when no summary appears. Additionally, they click on a source within the summary just 1% of the time.

Bar chart of Google search visits in March 2025. Without an AI summary, 15% of visits led to a click on a result and 16% ended the session. With an AI summary, 8% led to a click, 26% ended the session and 1% included a click on a link inside the summary
When the answer is already on the page, fewer readers reach the source.

Does format matter if the facts remain the same? According to McLuhan, “the medium is the message.” For example, ten links send you to read, while a single paragraph reads for you. Innis showed how the medieval Church had a “monopoly over religious information” through hand-copied scripture until the advent of print allowed believers to read without relying on “second-hand sources such as clergymen”. Search engines became our printing press, and AI answers are bringing back the role of the priest.

Innis also argued that empires survive by combining media that endure with media that travel. AI does both, freezing the web into models in a few companies’ data centres and delivering it instantly to every phone.

McLuhan’s tetrad for the AI answer: it enhances instant synthesis, obsolesces the link and the click, retrieves the priest and the oracle who read the sources for you, and reverses into model collapse.
McLuhan’s tetrad applied to the AI answer

The open web once exchanged content for traffic, but AI now takes content while keeping the traffic. The New York Times has sued OpenAI and Microsoft for using its articles in AI training, Cloudflare now blocks AI crawlers by default, allowing websites to charge for each crawl. Peters calls media ‘elemental’, and this cloud runs on electricity, water and land. The International Energy Agency warns that data centre electricity consumption could more than double by 2030, and in Ireland, data centres already use 23% of metered electricity usage, surpassing all urban homes combined.

Whoever writes the answer holds power, so states compete to write it. For example, China requires generative AI to adhere “core socialist values”. U.S. federal agencies must buy only models built on ‘truth-seeking’ and ‘ideological neutrality’ as outlined in Executive Order 14319, while users, including many Democrats, see nearly all models as left-leaning. The instinct is old, in 1559, Rome responded to the printing press with the Index of Prohibited Books, which banned Bibles in common languages.

Timeline titled ‘Who reads for you?’. Before 1450, manuscripts: the priest reads for you. In the 1450s, print: you read for yourself, and in 1559 Rome responds with the Index of Prohibited Books. From the 1990s to the 2010s, search: you choose from ten links. In the 2020s, AI answers: the model reads for you, and China (2023) and the US (2025) respond with rules for what AI may say.
Each new medium breaks one monopoly of knowledge, and power moves to control the next

Whether this medium changes our brains is unclear: the ‘Google effect’ on memory failed to replicate, and the viral MIT EEG study of ChatGPT users remains a preprint. Its economic and political effects are already measurable.

Williams would label my argument as technologically deterministic. Choosing one answer instead of multiple links is a business decision that keeps users on Google, not a matter of destiny. McLuhan might call this phenomenon retribalisation, but Towns argues that McLuhan’s idea of ‘tribal’ is rooted in colonial psychiatry concerning Africans, making ‘recentralisation’ a more accurate term. Park and Curran would question whose answer it truly is, considering that Llama 2 was trained on text that was nearly 90% English.

My previous post suggested that AI could eliminate the first step of the two-step flow of communication. Medium theory shows who benefits from this change. When was the last time you scrolled past an AI-generated response to consult an original source?

Reference list

Camerer, C.F. et al. (2018) ‘Evaluating the replicability of social science experiments in Nature and Science between 2010 and 2015’, Nature Human Behaviour, 2(9), pp. 637–644. Available at: https://doi.org/10.1038/s41562-018-0399-z.

Curran, I. (2026) ‘Data centres account for almost a quarter of Irish electricity usage in 2025’, The Irish Times, 7 July. Available at: https://www.irishtimes.com/business/2026/07/07/data-centres-account-for-almost-one-quarter-of-irish-electricity-usage-in-2025/

Cyberspace Administration of China (2023) Interim measures for the management of generative artificial intelligence services. Translated by China Law Translate. Available at: https://www.chinalawtranslate.com/en/generative-ai-interim/

Encyclopaedia Britannica (2026) Index Librorum Prohibitorum. Available at: https://www.britannica.com/topic/Index-Librorum-Prohibitorum

Executive Order 14319 (2025) Preventing woke AI in the federal government. Washington, DC: The White House. Available at: https://www.whitehouse.gov/presidential-actions/2025/07/preventing-woke-ai-in-the-federal-government/

Grynbaum, M.M. and Mac, R. (2023) ‘The Times sues OpenAI and Microsoft over A.I. use of copyrighted work’, The New York Times, 27 December. Available at: https://www.nytimes.com/2023/12/27/business/media/new-york-times-open-ai-microsoft-lawsuit.html

Hall, P. (2025) ‘Cloudflare will now, by default, block AI bots from crawling its clients’ websites’, MIT Technology Review, 1 July. Available at: https://www.technologyreview.com/2025/07/01/1119498/cloudflare-will-now-by-default-block-ai-bots-from-crawling-its-clients-websites/

International Energy Agency (2025) Energy and AI. Paris: IEA. Available at: https://www.iea.org/reports/energy-and-ai

Kosmyna, N. et al. (2025) ‘Your brain on ChatGPT: accumulation of cognitive debt when using an AI assistant for essay writing task’, arXiv [Preprint]. Available at: https://arxiv.org/abs/2506.08872

Laughey, D. (2007) Key themes in media theory. Maidenhead: Open University Press.

McLuhan, M. and McLuhan, E. (1988) Laws of media: the new science. Toronto: University of Toronto Press.

Newnam, B.F. (2007) ‘Are you feeling lucky? Google is’, Marketplace, 19 November. Available at: https://www.marketplace.org/2007/11/19/world/are-you-feeling-lucky-google

Park, M.-J. and Curran, J. (eds) (2000) De-westernizing media studies. London: Routledge.

Peters, J.D. (2015) The marvelous clouds: toward a philosophy of elemental media. Chicago: University of Chicago Press.

Pew Research Center (2025) Google users are less likely to click on links when an AI summary appears in the results. Available at: https://www.pewresearch.org/short-reads/2025/07/22/google-users-are-less-likely-to-click-on-links-when-an-ai-summary-appears-in-the-results/

Shumailov, I. et al. (2024) ‘AI models collapse when trained on recursively generated data’, Nature, 631, pp. 755–759. Available at: https://doi.org/10.1038/s41586-024-07566-y.

Touvron, H. et al. (2023) ‘Llama 2: open foundation and fine-tuned chat models’, arXiv [Preprint]. Available at: https://arxiv.org/abs/2307.09288

Towns, A.R. (2020) ‘Toward a Black media philosophy’, Cultural Studies, 34(6), pp. 851–873.

Westwood, S.J., Grimmer, J. and Hall, A.B. (2025) Measuring perceived slant in large language models through user evaluations. Working Paper 4262. Stanford Graduate School of Business. Available at: https://www.gsb.stanford.edu/faculty-research/working-papers/measuring-perceived-slant-large-language-models-through-user

1 thought on “Print Killed the Priest, So Why Are We Letting Someone Read for Us Again?”

  1. I really liked your post, I found the comparison between the printing press and AI very interesting, especially the idea that AI could bring back a form of dependence on intermediaries for accessing information. I agree that the way information is presented can be just as important as the information itself, many people now rely on AI-generated summaries because they are quicker and more convenient, but this can also discourage them from questioning information or checking original sources. I also think your point about who controls these technologies is important, as it raises questions about bias, power and whose perspectives are represented. It makes me wonder whether we are becoming more informed or simply more dependent on the platforms that provide our answers. Great analysis!

Leave a Reply