← AI for Educators Daily with Dan Fitzpatrick
Loading summary
Transcript1 lines
- [00:00]
A
If this episode makes you think, please let us know in the comments and support us by subscribing and leaving a review. Thank you. Today we are exploring a BBC Science Focus article by Ian Taylor called We ranked the most environmentally damaging things you can do online. AI didn't top the list. It is a short but very useful piece asking a question a lot of people are suddenly asking how bad is AI for the planet compared with other things we do online like streaming, social media and gaming? And for educators, I think it opens up a much bigger conversation about digital literacy, sustainability and how we help young people think clearly when a new technology becomes the target of public anxiety. Now, the first thing the article does well is acknowledge that AI absolutely does have an environmental cost. According to the piece, one 2026 estimates suggested that global AI use, and specifically the data centers powering it, emits as much carbon dioxide in a year as New York City. It also cites an estimate that every 5 to 50 ChatGPT queries may use about half a liter of water for server cooling, and notes warnings from advisors to the UK government that AI data centers could threaten water security. So this is not an AI is harmless story. Not at all. But the article is trying to do something more careful than that. It asks whether AI is actually the worst online habit environmentally and and its answer is basically not necessarily. According to the piece. It is hard to compare technologies cleanly, but some analysts have tried. So one example it gives is a calculation from tech writer Andy Masley suggesting that an average ChatGPT query emits about 0.28 grams of CO2, which he says is roughly equivalent to streaming a video for 35 seconds, uploading nine photos to social media, or using a laptop for one minute. And this is where it gets really interesting for schools. Because public conversations around AI often become strangely symbolic. AI starts to stand for everything we are worried about all at once. Cheating, laziness, job loss, surveillance, environmental damage, cognitive decline. And once that happens, it becomes very easy to stop thinking clearly. We stop asking compared with what? We stop asking under what conditions. We stop asking what is the actual scale here that matters in education? Because one of our jobs is to teach students how to reason proportionately. Not just how to have opinions, but how to weigh claims, how to compare impacts, how to notice when a debate has become more emotional than analytical? The article makes a very simple but important point on page four. A better way to think about the environmental cost of online behavior, it suggests, is to look at the amount of data being used at a given time. Reading text posts on LinkedIn has a lower impact than watching videos on TikTok. A text query to an AI chatbot uses much less data than an AI text to video prompt. And cloud gaming may be one of the worst online offenders because it requires gaming servers to run continuously. That is such a useful framing. Not is AI bad? But what kind of digital activity are we talking about? Text is not the same as video. A quick chatbot prompt is not the same as generating media rich outputs. Passive scrolling is not the same as reading. Streaming is not the same as messaging. And in schools that kind of distinction really matters because too many technology conversations collapse everything into one blurry category called screen use or AI use. But digital life is not one thing. It contains very different behaviors with very different consequences. The article then adds another twist I found really helpful. It says that a 2025 Greenlee report suggested physical video games may actually be around 100 times more carbon intensive than online streaming once you factor in manufacturing, packaging, distribution and disposal. In other words, the environmental story is not only about what happens on screen, it is also about the physical systems behind the thing. And that again, feels like a brilliant teaching point, because one of the big mistakes we make with students is talking about digital as though it is somehow weightless. Cloud, online, Virtual, Frictionless. But of course it is not. Every supposedly invisible system sits on physical infrastructure. Data centers, cooling systems, devices, supply chains, electricity grids, rare earth materials, manufacturing, waste disposal. The cloud is always someone else's warehouse full of machines. So I think this article gives educators a really nice opening into a more sophisticated kind of sustainability education, one that goes beyond easy slogans like AI is killing the planet or Digital is greener because it's paperless. Both of those are too simple. The truth is messier. Some digital habits are relatively light, some are much heavier. Some physical alternatives are worse. Some are better. You need judgment. And then the article lands its strongest comparison. According to the piece, an annual Netflix subscription with average viewing times produces around 17kg of CO2, roughly equal to a 60 mile petrol car journey. A single economy flight from London to berlin emits around 10 times that amount, and one sirloin steak may carry 20 to 30 kilograms of carbon, more than a year's worth of binging Bridgeton. Honestly, that is the line that stayed with me, because it forces a kind of perspective that is often missing in AI debates. If your goal is genuinely to reduce carbon impact, then yes, it makes sense to think about digital habits. But according to this article, what you buy, how you eat, and how you travel may still have much bigger effects than your chatbot use. Now why does that matter for education? Because schools are one of the places where young people learn what counts as a serious issue. And if we over focus on the newest, loudest technology without helping them compare it to broader systems of consumption, then we risk teaching distorted environmental literacy. Students need to understand not only that AI has costs, but that sustainability is about trade offs, scale and context in it is about whole lifestyles and systems, not just the newest tool in the headlines. This also connects to how schools themselves make decisions. I can absolutely imagine a school becoming very animated about the environmental ethics of AI while saying very little about device replacement cycles, energy use, food procurement, travel policies, printing habits or procurement more broadly. And that would be a mistake. Not because AI does not matter, but because selective concern can become performative very quickly. The better approach is a whole school lens. Where does digital use genuinely save resources? Where does it quietly increase them? When is AI a smart substitution and when is it simply additional load? What tasks are worth the cost because they create real educational value? What habits are just wasteful convenience? That is a much more mature conversation. And I think there is a final educational angle here that is easy to miss. The article is really about comparison, not panic. It is modeling a habit of mind schools should care deeply about. Don't fixate on the thing everyone is blaming until you understand the wider system. That is a lesson students need not just for sustainability, but for media literacy, politics, economics and AI itself. Because the pattern repeats everywhere a new tool arrives, public fear spikes, headlines get dramatic, social media reduces the argument to a slogan, and suddenly everyone is certain about something very complex. What education should do at its best is slow that process down, ask better questions, compare cases, note a scale, resist easy villains. So where does this leave us? For me, the educational takeaway is not don't worry about AI's environmental impact. We absolutely should worry about it. We should ask better questions about data centers, energy, water, and what kinds of AI use are actually justified. But we should do that without losing perspective. A text chatbot prompt is not the same as a video generator. AI use is not the whole of digital life, and digital life is not the whole of environmental impact. In other words, this is a great reminder to teach students and ourselves not just to react to the newest technology, but to think systemically about it. That is the real skill. That's all for today. Thanks for listening.