What do you think is the future of AI?
View all responses
Most forecasts about artificial intelligence are forecasts about the machine. How powerful will it be in ten years? Which jobs will it take? Will it one day think like us?
My forecast is about the person on the other side of the screen, and I found it in an unlikely place: a Tamil grammar1 written more than eight hundred years ago.
Here is my answer to today’s question. AI will not create a new kind of human being. It will make the old kinds more extreme.
1. The machine is the easy part to predict.
AI will almost certainly become faster, cheaper and more capable. That forecast takes no courage. The harder question is what all that capability will do to the habits of the people who use it, because habits, not hardware, decide what a tool is worth.
2. An old classroom has already sorted us.
Nannool, a grammar by the Jain monk Pavanandi Munivar, contains a short verse that ranks students using animals and household objects. The best are like the swan, which is said to drink the milk and leave the water, and the cow, which grazes first and chews the cud later. In the middle are the parrot, which repeats only what it was taught, and the soil, which gives back only as much as the farmer puts in. At the bottom are the leaky pot, which loses whatever is poured into it, and the ghee-strainer, which lets the good pass through and keeps the dregs.
3. Another tradition reached the same kitchen.
A thousand years earlier and thousands of kilometres away, the Jewish sages of the Mishnah described four kinds of people who sit before the wise: the sponge, the funnel, the strainer and the sieve.2 The sponge soaks up everything, clear or murky. The funnel lets everything run straight through. The strainer lets the wine flow away and keeps the sediment. The sieve holds back the fine flour. Two traditions that never met reached for the same kitchen utensils, and agreed on what the poorest learner keeps: the dregs.
4. AI makes the lower ranks effortless.
This is the heart of my forecast. AI does not invent these types. It removes the effort that once kept them apart. Being a parrot used to require memorising; now the machine repeats for us. Being a funnel used to waste a teacher’s afternoon; now answers pour through us at no cost and leave nothing behind. And because AI can sound just as confident when it is wrong as when it is right, the strainer’s danger grows. We may keep the fluent error and let the careful truth flow away.
5. The higher ranks still cost what they always cost.
No machine can do the swan’s work for it. To separate milk from water, you have to taste both. No machine can chew the cud for the cow; understanding settles only when we return to a thing and turn it over. We learn to reason by reasoning, to write by writing, and to judge by weighing the evidence ourselves. This year I have been learning a few terminal commands, and an assistant can hand me any of them in a second. Copying is faster. Asking why the command works is slower. Only the slower way leaves something behind.
6. Granted, every new tool has faced this charge.
In Plato’s Phaedrus, Socrates retells an Egyptian story in which a king warns that writing will breed forgetful learners who seem wise without being wise. We know of that warning only because Plato wrote it down. Writing did not turn humanity into funnels; it gave us libraries. AI may follow the same path, and it can serve the swan well by explaining a hard idea, offering the counterargument or testing a draft. Nor does everything deserve rumination. The cow chews what nourishes it, not everything in the field. Hand the trivial to the machine; keep the formative for yourself.
7. So the gap will widen.
If I am right, the future of AI is not one future but two. Those who use it to extend their thinking will grow sharper, faster than any generation before them. Those who use it to avoid thinking will grow emptier just as fast. Institutions will tilt the balance. A school that grades only the finished essay rewards the parrot. A school that asks students to defend their reasoning aloud rewards the swan.
So what do I think is the future of AI? I think it will be remarkably powerful, and that its most important effects will be decided in small, private moments: each time an answer arrives instantly, and we choose whether to taste it, chew on it, or simply let it pour through.
The machine will not tell us what kind of learners we must become. That decision remains ours.
When you use AI, which of the old classroom’s creatures do you recognise in yourself, and which would you rather become?
Keep learning, keep showing up, and keep striving to elevate in life.
Notes
1 Nannool (Naṉṉūl, “the good treatise”) is a Tamil grammar composed around 1200 CE by Pavanandi Munivar, a Jain ascetic, and is regarded as the most important Tamil grammar after the Tolkappiyam. The verse ranking students as first, middle and last (sutra 38 in the common numbering) appears in its general preface; the images above follow the traditional commentaries.
2 Pirkei Avot (“Chapters of the Fathers”) is a tractate of the Mishnah, the foundational compilation of Jewish oral teaching completed around 200 CE. The passage cited is Pirkei Avot 5:15 (5:18 in some editions).
John Britto Kurusumuthu
Rise & Inspire · riseandinspire.co.in
In response to the WordPress Daily Writing Prompt, 24 October 2026: What do you think is the future of AI?
© 2026 Rise & Inspire. All Rights Reserved.