> ## Content Index > Fetch the complete content index at: https://www.thelivelihood.org/llms.txt > Use this file to discover other available public pages before exploring further. # Thinking outside the Bot - URL: https://www.thelivelihood.org/thinking-outside-the-bot/ - Published: 2026-10-12T09:03:00.000Z - Updated: 2026-10-12T09:02:59.000Z - Author: Catherine Pope - Tags: Tech, Books An IPSOS poll this year found that around two-thirds of workers expect AI to eliminate jobs, but only 6% to 7% believe AI will replace their job. Amid all the panic and denial, advice usually suggests doubling down on being human, pursuing roles that require empathy and other “soft” skills. According to economist Daniel Susskind, even these jobs aren’t safe. In [*What Should My Children Do?*](https://uk.bookshop.org/a/2760/9780241791172?ref=thelivelihood.org), he urges us to accept that anything we can do, AI can do just as well or even better. Humans are overrated. Consequently, schoolchildren should be taught to routinely use chatbots in the classroom. Rather than struggle with difficult texts, AI can explain it to them. Instead of borrowing a book from the library, the can ask AI to write a story that makes them “laugh and shudder” (p.9) Susskind describes AI as “an increasingly impressive tutor,” claiming it can help students solve problems and do so “with sophistication and care.” (p. 88) I agree that platforms like the Khan Academy have been an absolute boon for teaching maths. But I’m flummoxed by the idea that they do this with “care”. A chatbot doesn’t care. Teachers show “sophistication and care,” at least when they’re given the resources to do so. Why don’t we invest more in teachers to reduce class sizes? Susskind doesn’t cost his proposal, but I’m willing to bet that it’s a lot more expensive than humans. Especially when you factor in the pace of change and inevitable obsolescence of the infrastructure. My visceral dislike of AI teachers is exquisitely expressed by [Katherine Rundell in a recent article for *The Guardian*](https://www.theguardian.com/books/ng-interactive/2026/aug/08/i-hate-what-ai-is-doing-to-the-minds-and-happiness-of-the-young-katherine-rundell-on-the-view-from-the-classroom?ref=thelivelihood.org). We should be engaging with young people, rather than literally leaving them to their own devices. Otherwise, we’re teaching them that human interaction is either unnecessary or an unaffordable luxury. Susskind rejoices in how much time he saves by getting chatbots to write bedtime stories, answer questions, and learn about topics. What does this efficiency free up time for? More work? More online chats? Do these time-savings justify the environmental impact? Indeed, Susskind makes no mention of the environmental impact of his vision.[\[1\]](#fn1) The term “cloud computing” effaces a huge physical infrastructure that consumes vast amounts of finite natural resources. How many more data centres would be needed to service the demand? Are we heading for a system where children interact with chatbots from home, with the data stored in their former school? Is that a more “efficient” use of the space? Yes, the curriculum needs to reflect our increasing reliance on technology. But it must also address the climate crisis. How will children cope with rising temperatures, mass migration, and crop failures? Are chatbots going to provide practical assistance? Or will they just offer a useful distraction? Assuming we still have enough power to run them. We need human connection to confront the climate crisis, not symbiosis with LLMs. Even therapists aren’t safe. Susskind claims that AI already exhibits “impressive soft skills” (p.74) and is better than a human therapist: “you never worry that AI will get tired or bored of you; you never worry it is hungry or distracted or not paying attention; and you never fear it will go home at the end of the day and chat about your misfortunes at the dinner table.” (p. 76) Trained therapists don’t discuss their clients at the dinner table. Apart from the confidentiality breach, they’d much rather talk about something else. And there’s a reason why therapy sessions are rarely more than one hour a week. It’s not healthy to spend hours in therapy, especially not with a chatbot. A trained therapist responds to verbal and visual cues, noticing when you’re stuck on an issue or when you’re in pain. Human limits are important. While chatbots won’t talk about you behind your back, they hold transcripts of your sessions on their servers. Frankly, it’s a lot safer in someone else’s head. Although AI has pervaded many areas of our lives, I don’t agree that its continued domination is as “inevitable” as Susskind suggests. There have been several AI winters, where periods of hype have been followed by a withdrawal of funding. Even some AI enthusiasts acknowledge the expanding bubble. Tech companies can’t [scale infinitely](https://www.thelivelihood.org/scale-by-geoffrey-west/), but that’s what they’re expected to do if they want investors to keep shovelling the money. It’s not a case of either embracing or denouncing AI: we can decide where it’s genuinely useful and also establish some boundaries where it’s likely to be harmful. Do we really want AI to take over creative tasks? Despite claims that it has revolutionised our lives, Claude has never helped me with the laundry or unjammed the printer. We can also remain sceptical of predictions. In 2016, Geoffrey Hinton [famously predicted the demise of radiologists](https://time.com/article/2026/05/12/leaders-need-to-stop-pretending-they-can-predict-the-future/?ref=thelivelihood.org), insisting that we should stop training people in a profession that would be replaced by AI. Now there’s a shortage. Nobody can predict the future, least of all the people who claim to own it. There is so much that troubles me about this book. Above all, though, I think it’s the obsession with efficiency. Learning shouldn’t be efficient. Making mistakes, pushing through the difficulty, and discussing approaches with other humans is what helps us develop. I agree that young people need to understand how to use AI, but it’s more important that they understand other humans. We should be raising citizens, not consumers. Just as the tech companies crave content that hasn’t been tainted by their own slop, we need brains that haven’t been trained on it. People who can still think independently, read deeply and see the bigger picture will become increasingly valuable. In [*Users: How Big Tech Took Control and How to Fight Back*](https://uk.bookshop.org/a/2760/9780753562468?ref=thelivelihood.org), Beeban Kidron quotes a 13-year-old who asks, "What is efficient about making human beings redundant?" (p.290) Quite. Big Tech solutions are seldom cheaper, especially not when you factor in the cost of environmental damage and mass unemployment. We can’t do much about the relentless drive of Big Tech, but we can decide how much we want to engage with its products: what we’re willing to sacrifice for the convenience, how much of our data we’re prepared to share, and what skills we’ll allow to deteriorate. We have a choice. It’s not inevitable. The ability to think outside the bot is the most valuable skill of all.  --- 1. I wanted to double-check this omission. Unfortunately, there’s no index. Incidentally, chatbots are absolutely terrible at indexing books. [↩︎](#fnref1)