The elephaint is not in the room: staying human in the tech bros’ world

By Jeremy Harmer

When it comes to AI, that old idiom ‘the elephant in the room’ (i.e. something we all know is there but we all avoid talking about) is massively out of date. The elephaint©! is no longer in the room but right out there in the open, and now, it seems, we can talk about almost nothing else. Some see what the tech bros (Musk, Altman, Zuckenberg etc) have done to us as ‘enshitification’ (we are dumb obedients being led by profit-chasing cynics who really couldn’t care less) (Doctorow 2025).

In January 2026, a documentary premiered at the Sundance Film Festival called ‘Are you an Apocaloptimist?’ – a question which provokes, in me, the same kind of response as Nicky Hockly, who said, in her 2026 IATEFL talk, how her feelings about AI vacillated dramatically from excitement to fear on a daily basis (Hockly 2026). She sounds just like me. I could go on and talk about the scandal round The New Atlantic’s awarding of the Commonwealth writer’s prize to a Trinidadian writer’s short story now widely assumed to be, perhaps wrongly, an AI-written piece (Vara 2026), or the continuing fears in the academic community that students are turning in AI-generated work which their lecturers can’t spot. A recent article in the New York Times bemoaned silent mathematics classrooms where students were working with AI, thus avoiding both the mental struggle (which many see as a vital component of learning and development) and the human interaction of a helpful teacher (Polakow-Suransky 2026).

And what about us, EFL (and ESOL) teachers? What match are we for the multitudinous universe of AI with its ability to offer planning and practice, instant translation, learnt empathy, and individually-tailored tutoring? Is our profession, which I have been part of for more years than I can count, doomed to extinction?

Frankly, I have no idea.

But right now, and as far into the future as I can predict, I don’t think we are going anywhere. The ELT world may have changed in many ways (more English in mainstream schools, fewer ‘general’ English classes; more online teaching from IH schools, for example, and even, in some countries, a move away from English to other languages), but the value of the human teacher is as vital, as heartwarming, and as messy as it ever was, and those of us who stand on the shoulders of giants (and in particular, in the International House world, the enthusiasm of IH founder John Haycraft) need to trumpet these qualities from the turrets and battlements as we tell AI to stick to its realm and do what it does best while we take care of the important stuff.

“I was at a company,” Steve Wozniak, co-founder of Apple said in his address at the Grand Valley University Graduation ceremony in Michigan recently, “where the engineers worked out how long it takes to make a brain,” he told the graduands, “Yeah, it takes nine months,” His audience loved him – in stark contrast to other speakers enthusing about AI in ceremonies across the campuses of the USA who have been jeered and booed by a new student population.

There is hope. There is a future!

In the light of all this, I thought I would show some old – and some some less old – examples of what being human means in ELT and how it beats AI hands down.

1. The human touch

In her recent book (Hasper 2026) on dealing with students who have experienced trauma (aka all of us, in a way) Anna Hasper mentions a boy who would not stop being noisy (talking, singing etc) when the class was supposed to be working silently. It was only after gentle human interaction with him that she discovered that silence unnerved him because at home silence meant conflict was coming. Human teachers are sensitive beings – AI only pretends to be.

2. Challenging AI

How good, really, is this Artificial Intelligence, anyway? Here’s a human task for B2+ students to find out.

Choose a poem (I have used Everyone Sang by Siegfried Sassoon – an ambiguous celebration (?) of the end of a war). Have students learn it by heart in a designated time. I have done this many times. Even though it is a struggle (see above) students respond very positively to the experience. Have them write down, in pairs or groups, the themes they think emerge in this poem. Now each group challenges AI to write an original poem (not just a ‘scraped’ or stolen one) on these themes (they should specify length, etc). The students then say whether they think the AI has written a better poem than Everyone Sang. They can be asked to justify their opinion and then, perhaps, they can try and write a poem of their own.

3. Making AI work for us

Teachers are getting AI to write lesson plans all the time. Hopefully, however, they bring their own human insights and evaluative powers to assess how good – and, especially, how appropriate – the AI plans actually are. Here’s a student activity for A2 and above. It has five stages and comes with a musical health warning! First, organise a whole-class discussion about what makes a good song. Perhaps the main requirement is melody. Or perhaps it’s the rhythm, or the lyrics? Does it need to have a hook? Etc. The class then decides what today’s song is going to be all about. They then get into groups. Each group discusses what kind of song they want to have and then writes prompts for AI to write ‘their’ song for the topic.

Now the class listens to the songs from each group. They vote on which one works best and give reasons for their choice.

Now we move to stage 5, which turns out to be a challenge to the whole nature of AI itself. Students discuss the following three questions: (1) How original are the songs that AI has provided? Why do you think this? (2) Which would you prefer? A song written by a living, breathing human, or a song created by AI? Why? (3) Do you think record companies, film-makers, etc should release AI-generated music or not? Why?

4. Stopping at the traffic lights

The Armenian teacher Margarít Grevórgyan suggested a traffic light system for AI (i.e. green = use it, amber = maybe use it, red = definitely do NOT use it) in a session in Yerevan earlier this year. I saw something similar when I observed a lesson in London recently where the students were told: “AI is not your friend for this activity”.

Students stop at a red light when we ask them to do something genuinely creative by themselves. A favourite (old) example of mine is the ‘you’ poem where students have to think about someone they love and then write them a poem. Each line of their poem starts with the word You and then describes the person as (if they were) something else, e.g. a type of food (You are creamy garlic cheese), an article of clothing (You are a soft silky sari) or an animal (You are a long-haired monkey swinging through the trees). All the examples above were written by nine-year-olds by the way. Students pass their poems around and then someone reads out the poem in front of them (not their own) so that the class can hear what the others have written.

Among other red light activities students choose their favourite sentence in a text before practising it and then reading it out loud or, in groups, they act out a poem they have been given in front of their classmates after they have had time to work out how to perform it.

5. Critical thinking

In conversation recently, a 7-year-old granddaughter of mine mentioned that “we do critical thinking at school,” and when questioned she explained with great patience (!) that it means always asking “Is it real? Is it true?” Which seemed like a pretty good starting point. Her school was following a trend to try and get children to interrogate the information they are given, from any source.

I thought about this the other day when someone was claiming on TV to speak for ‘the ordinary Joe’, by which he meant ‘everyman’. He was offering the opinion of all of us though it was, as it happens, an opinion I disagreed with profoundly, so he obviously wasn’t speaking for me. “We’ve had enough of this,” he was saying and I was suddenly reminded of an ELT activity by Mario Rinvolucri and Paul Davis from more than thirty years ago in which the teacher dictates ambiguous statements such as We like ice cream or We are losing the battle against pollution and students have to decide who ‘we’ are. It occurred to me, then, that critical thinking is not a new phenomenon or a ‘twenty-first century life skill’ or whatever it is called. On the contrary, there has always been an obligation to question what we are told rather than following the crowd blindly.

And that led me to a further, and perhaps more fundamental, conclusion: we should stop being spooked by AI. It is just another manifestation of that usual mixture of human progress and carelessness and it is our job, being human, as we are, to use it well and wisely and not let it use us.

(Oh, and an admission: I couldn’t remember where the ‘we’ activity came from, so I asked Gemini – which told me straight away!)

Bibliography

Doctorrow, C (2025) Enshitification. Why everything suddenly got worse and what to do about it. Verso

Hockly, N (2026) Cruel optimism? ELT writers in the age of generative AI. Paper delivered at the IATEFL conference. Brighton 2026.

Polakow-Suransky, S (2026) The high cost of silent classrooms. New York Times, June 8. https://www.nytimes.com/2026/06/08/opinion/ai-classroom-silence-reform.html?unlocked_article_code=1.olA.9fSG.uv2Zt7za3fKm&smid=nytcore-android-share&utm_source=follow.it&utm_medium=email

Rinvolucri, M & Davis P (1995) More Grammar games: cognitive, affective and movement activities for EFL students. Cambridge University Press

Vara, V (2026) This literary AI scandal changes everything. The New Atlantic May 31 ((https://www.theatlantic.com/books/2026/05/granta-ai-fiction-book-scandal-changes-everything/687243/)

Biography