One day it was very hot…and Ant said he was as hot as a boiled egg…but Bee said Ant would feel hotter if he lived in /…
More Ant and Bee, Angela Banner, 2013
LLM and Hornet (after Angela Banner), 2026, pages 1-3. Made in the style of Angela Banner’s Ant and Bee series – created to accompany this post, but perhaps a work-in-progress?
I am screaming inside about the now-accepted and increasingly embedded phrase AI. I have banged on about this before but forgive me… it’s driving me nuts. I’m aware I may be in danger of appearing just as maniacal the hyper-positive or most doomerist voices on social media (well, to be honest, as quite a lot of groups and individuals on social media appear about many topics – vicious, rabid and obsessed). But I wish to God people would stop using the term and start being more specific when describing the tool they are discussing. In my eye-twitching state, I have begun to notice and admire writers or commentators who avoid the phrase altogether, employing use-specific phrasing instead; and I find myself not trusting those who acquiesce and continue speaking about so-called artificial intelligence. I am fully, fully aware this probably sounds irrational, if not barking mad. Plenty of serious people who stick with the letters ‘AI’ are writing serious articles. But I can’t help it. It’s as if the more embedded it becomes, the more childlike and bewildered the term seems. With our spleens, would the machines not all be laughing themselves mortal at our ridiculous nomenclature?
Recently, Mamdani, the Mayor of New York City (NYC) – whose general direction of travel I love – has issued a moratorium on student-facing ‘generative AI’ in schools, with varying levels of prohibition across age groups (New York City Mayor’s Office, 2026) and recommended screen time limits too for younger children. So-called AI is banned. But screen time use is left to parents’ and teachers’ discretion. (Why this distinction?) The message is more specific than seen elsewhere – this is generative rather than blanket AI – but still, there’s a fuzziness around those letters that risks encouraging anti-AI doomerists to celebrate on social media, without reading the whole article. The same goes for the blind promoters, only the other way around. In truth, certain tools that may be described as AI will be accepted for limited classes (5% of students) and pilot classes. I may be wrong but I imagine teachers won’t be prevented from using operational tools that are just as likely to circumvent ‘human connection in order to learn and grow’. All technology, the NYC website announces, must be ‘mission-critical to learning’. What the NYC announcement doesn’t say is who decided on the accepted brands for the pilots: Quill, Edia, Brisk Teaching, Playlab, Intel AI-Ready Schools. Who chose these particular tools? There is no published criteria, no threshold, no timetable, no named decision-maker – although they do say ‘The pilots were selected based on rigorous safety and data privacy standards, their focus on teacher-guided instruction and their commitment to keeping students as the primary thinker in the classroom.’ NYC is not promising a forest school by any means, lest anyone get caught out thinking as much.
What frustrates me is that plenty of educational systems, for example, ClassDojo (which faces both children and staff) have relied on algorithms, feedback loops, and dynamic digitisation for over a decade; in the case of ClassDojo to gamify behaviour control in classroom. Punished by Rewards author Alfie Kohn (publication 1993, interview 1995) long ago argued such systems whether administered by a computer or a person ‘destroy the potential for real learning’, and despite his message, products like ClassDojo have proliferated in the decades since. Far from being considered destructive, ClassDojo is generally seen as ‘cute’ and imperative. Alongside that, procurement systems have been narrowing what and how teachers may teach, here as much as in the States, since long before ChatGPT was released to the public in 2022. None of that is what anyone means when they say AI (generative or otherwise) – and all the vitriol aimed at AI ignores automation systems that came before while attacking anything worthwhile that has emerged since. The way the term AI is used seems to exempt everything that arrived before 2022, including processes that favour automation and data collection, and which do as much, if not more, harm than constructive use of a large language model (LLM) might. (Leave LLMs out of English classes if you will. But also analyse how the technology has triggered interesting changes, such as the growth of oral examinations.)
The critical learning sessions the policy proposes are an important step, but what is proposed is limited to two 45-minute modules, twice yearly, high school only. We certainly need more of that across the board in terms of digital culture if we are taking this seriously. Will those 90 minutes address how the letters ‘AI’ sucks in so much unthinking vitriol, including anything potentially useful? Not all of these processes are the work of the devil! Away from the emotionally wrought topic of how we educate our children, The Guardian (Wint, 2026) reports the UK is being invaded by the yellow-legged (Asian) hornet, Vespa velutina. The species is a serious threat to national biodiversity and to humans. (As someone who suffers a nasty reaction to bees and wasps, I am not keen on news that these hornets are on the rise at all). If members of the public discover a hive, says the article, they are reporting them to the British Beekeepers Association (BBKA) (Wint, 2026), although the official channels are usually the Asian Hornet Watch app and the Non-Native Species Secretariat. Many hives are being missed, however, and an automated system, which uses dynamic deep learning processes called VespAI (of course!), is therefore being tested. According to the company’s website, the algorithmic system detects, classifies and notifies. They say it is ‘a passive system that does not harm or kill visiting insects’ (VespAI, 2026) and is ‘a robust early warning system to prevent ingressions into new regions’ (O’Shea-Wheller et al., 2024: 1). I do not know if it is the holy grail in identifying species which threaten to obliterate important parts of our ecosystem. I am also aware the hornets are probably arriving here due climate change which was triggered by industrialisation and exacerbated by digitisation. However, if testing proves effective, how does the generalised appalling PR emanating from those two letters, AI, help or hinder its place in the world?
I am glad that NYC is looking at ways for schools to work with technology. It’s important and necessary. But I do fear our commitment to the letters AI to describe a multitude of processes and systems that have grown out of digitisation, leaves us making arbitrary decisions about how we think about and deal with many of them. I have been as guilty as anyone when it comes to conflating. We’re all rushing about, trying to navigate a paradigm that is changing in significant ways; it is almost, if not entirely, impossible to keep up. Perhaps making us all more than a little psychotic at times. We are understandably worried about how our children will learn when there is a machine that learns certain previously considered human-only tasks faster and more comprehensively than we ever could. We are horrified by the energy resources. We are scared about job losses. There is so much to be concerned about. However, there are also, as much as anyone wishes this not to be the case, elements and outcomes to feel positive about. Finding accurate names for the very many tools that are emerging so that we can talk about them with precision is surely a worthwhile endeavour.
Bibliography
Banner, A. (2013) More Ant and Bee. New and updated edn. London: Egmont.
Brandt, R. (1995) ‘Punished by Rewards?: A Conversation with Alfie Kohn’, Alfie Kohn. Available at: https://www.alfiekohn.org/article/punished-rewards-article/ (Accessed: 6 September 2026).
Kohn, A. (1993) Punished by Rewards: the trouble with gold stars, incentive plans, A’s, praise, and other bribes. Boston, MA: Houghton Mifflin.
NNSS (2026). Non-native Species Secretariat (NNSS) Available at: https://www.nonnativespecies.org/ (Accessed: 6 September 2026).
New York City Mayor’s Office. (2026) Mayor Mamdani and Chancellor Samuels Put Students First with Nation’s Broadest Generative AI Moratorium in Schools (2026) The official website of the City of New York. Available at: https://www.nyc.gov/mayors-office/news/2026/09/mayor-mamdani-and-chancellor-samuels-put-students-first-with-nat (Accessed: 6 September 2026).
O’Shea-Wheller, T.A. et al. (2024) ‘VespAI: a deep learning-based system for the detection of invasive hornets’, Communications Biology, 7(1), p. 354. Available at: https://doi.org/10.1038/s42003-024-05979-z.
VespAI (2026) VespAI. Available at: https://www.vespai.co.uk (Accessed: 6 September 2026).
Wint, N. (2026) ‘Beekeepers warn of Asian hornet threat and call for UK awareness campaign’, The Guardian, 5 September. Available at: https://www.theguardian.com/environment/2026/sep/05/beekeepers-asian-hornet-threat-uk-awareness-campaign (Accessed: 6 September 2026).
Edits
06/09/2026: Edited after posting to add a quote from the New York City Mayor’s Office about how certain tools were chosen, to clarify and simplify some sentences.
Having written this post yesterday, perhaps it is a little bit of synchronicity to find a related article in The Guardian this morning by Daniel Susskind. Yes, yes, yes he uses the term I find irritating.
A different approach in China: ‘Schools in Beijing will introduce AI courses and teaching methods into the primary and secondary school curriculum starting September, to nurture young talent and boost growth in the advancing sector.’ https://www.channelnewsasia.com/east-asia/china-schools-ai-courses-4990306