Brain activity drops when using AI chatbots, says MIT! Are you learning less? #AIChatbots #BrainHealth #AIScience
Explanation in video
Hey everyone, John here! We’re always chatting about the incredible new things AI can do – it feels like every week there’s a new breakthrough! From helping doctors to creating amazing art, AI is definitely changing our world. But have you ever stopped to think about how using these super-smart tools might be changing… well, us? Specifically, our brains?
And as always, my wonderful assistant Lila is here to help us dive into these topics and make sure we all understand what’s going on.
Lila: “Hi John! This sounds really interesting. I use AI tools sometimes, and I’ve definitely wondered if they’re helping me learn or just doing the work for me. So, what’s the latest news on this?”
John: “Great question, Lila! It’s a hot topic, and some researchers at the very famous Massachusetts Institute of Technology, or MIT, have just released some early findings that are making a lot of people think. It seems that when we use certain AI tools, our brains might be taking a bit of a vacation!”
What Exactly Did the MIT Researchers Look Into?
John: “Alright, so these scientists wanted to see what happens in our brains when we use AI chatbots to help with tasks that usually require a good bit of thinking, like writing an essay. They set up an experiment to compare people doing these tasks with AI help versus doing them entirely on their own.”
Lila: “Hold on, John. The article mentioned this was a ‘preprint study.’ What does that mean?”
John: “Excellent point, Lila! A preprint study is like an early sneak peek of scientific research. Scientists write up their findings and share them with the community before the study has gone through a full, formal review process by other experts (that’s called peer review). It’s a way to get new information out quickly, but it also means the findings are preliminary and might be refined later. So, we should keep that in mind – it’s interesting, but not the final, final word just yet.”
Lila: “Okay, got it! So, how did they see what was happening in people’s brains? That sounds like something out of a sci-fi movie!”
John: “Haha, not quite sci-fi, but very clever! They used a technique called EEG.”
Lila: “EEG? That sounds super technical!”
John: “It does, doesn’t it? EEG stands for Electroencephalogram. It’s a bit of a mouthful! But what it does is pretty straightforward. Imagine tiny sensors, like little listeners, are placed on someone’s scalp. These sensors can pick up the very faint electrical activity that our brains produce naturally when we’re thinking, feeling, or doing anything, really. More activity often means the brain is working harder. Think of it like a fitness tracker, but instead of tracking your steps, it’s tracking how much ‘effort’ your brain is putting in.”
Lila: “Wow, that’s cool! So they had people writing essays, some with AI like ChatGPT, and some without, and they were monitoring their brain activity with EEG?”
John: “Exactly, Lila! You’ve nailed it. ChatGPT (that’s one of those AI programs where you can type in a question or ask it to write something for you, and it gives you a surprisingly human-like response) was one of the tools they looked at. So, one group used AI to help with their writing tasks, and the other group just used their own brainpower.”
So, What Did They Find? Is AI Helping or Hindering Our Brains?
John: “This is where it gets really fascinating. The researchers found that when people used AI chatbots to help them with writing tasks, their brain activity, as measured by the EEG, was noticeably lower compared to when they did the same tasks without any AI assistance.”
Lila: “Lower brain activity? So, like, our brains are being a bit lazy when AI is involved?”
John: “That’s a good way to put it, Lila! It’s as if the brain says, ‘Oh, the AI is doing a lot of the heavy lifting here, so I can just chill out a bit.’ It’s not necessarily ‘lazy’ in a bad way all the time, but it does mean less mental effort is being exerted. Think about using a calculator for a simple math problem versus doing it in your head. Your brain works harder when you do it yourself.”
Okay, Less Brain Activity… But Does It Affect Learning?
John: “That’s the million-dollar question, isn’t it? It’s one thing if our brains work a bit less, but if we’re still learning effectively, maybe it’s not a big deal. However, the MIT study didn’t just stop at brain activity. They also looked at how well the participants remembered information related to the tasks they performed. This is often done using something called recall tests – basically, seeing what you can remember after a certain period.”
Lila: “And what did these recall tests show? Did using AI help people remember better because the task was easier?”
John: “Actually, it seems to be the opposite, Lila. The study suggested that the people who relied on AI chatbots not only showed less brain activity during the task but also had poorer fact retention afterwards. In simpler terms, they didn’t seem to learn or remember the material as well as those who tackled the tasks on their own.”
Lila: “Oh, wow! So even if the AI helped them produce a good essay, for example, they might not have learned as much from the process of creating it?”
John: “Precisely! It’s a bit like if someone gives you a beautifully written summary of a book. You get the information quickly, but you probably won’t remember the details or understand the nuances as well as if you’d read and processed the entire book yourself. The struggle, the effort of thinking, connecting ideas, and putting things into your own words – that’s a huge part of how we learn and remember things. If AI shortcuts that process, it seems it might also shortcut some of the learning.”
This is often talked about as ‘cognitive offloading’.”
Lila: “Cognitive offloading? What’s that, John?”
John: “Good question! ‘Cognitive’ refers to thinking, and ‘offloading’ means passing a task or burden onto something or someone else. So, cognitive offloading is when we use external tools – like a calculator, a GPS, or in this case, an AI chatbot – to do some of the thinking work for us. For example, instead of memorizing a phone number, you save it in your phone. You’ve offloaded the memory task to your device. It’s super convenient, but it might mean you don’t actually know the number by heart.”
What Does This Mean for Us and How We Use AI?
John: “This is a really important takeaway. This study doesn’t mean AI is ‘bad’ or that we should stop using it. AI tools are incredibly powerful and can be amazing assistants. They can help us be more productive, brainstorm ideas, and access information quickly.”
However, these findings suggest we need to be mindful about how and when we use AI, especially when it comes to learning new things. Here are a few thoughts:
- AI as a Helper, Not a Replacement: Think of AI as an assistant that can help with research or first drafts, but the core thinking, understanding, and learning should still be driven by you.
- Active vs. Passive Learning: Grappling with information, rephrasing it in your own words, and connecting it to what you already know – these are active learning processes. If we let AI do all that for us, our learning might become more passive, and less sticky.
- The ‘Desirable Difficulty’ Principle: Sometimes, a bit of a struggle is good for learning. Psychologists call this ‘desirable difficulty.’ If something is too easy, we don’t engage our brains enough to form strong memories. If AI makes everything too easy, we might miss out on this beneficial struggle.
- Critical Thinking is Key: Even when using AI, we need to engage our critical thinking skills. Is the AI’s output accurate? Is it complete? Does it make sense? Don’t just accept what it gives you without thinking.
Lila: “So, it’s about balance then? Using AI smartly, but not letting it do all our thinking for us, especially if we actually want to learn something properly?”
John: “You’ve hit the nail on the head, Lila! It’s all about finding that balance. AI can be a fantastic tool to augment our abilities, but we shouldn’t let it diminish them by over-relying on it for tasks that help our brains grow and learn.”
A Few Final Thoughts
John: “For me, this study is a fascinating, if preliminary, reminder that our brains are like muscles – they need exercise to stay strong. If we constantly outsource the ‘heavy lifting’ of thinking and learning to AI, we might find our own cognitive abilities aren’t getting the workout they need. It’s not about avoiding AI, but about using it wisely to support, not supplant, our own intellectual efforts.”
Lila: “That makes a lot of sense, John! It’s a bit like using a map versus just following a GPS. When I use a map, I tend to learn the area better. If I just follow the GPS voice, I get where I’m going, but I often have no idea how I got there or how to get back without it! I’ll definitely be more thoughtful about how I use AI for learning from now on, making sure I’m still doing the core thinking myself.”
John: “Well said, Lila! It’s all about active engagement. So, while we continue to be amazed by what AI can do, let’s also remember the incredible power and potential of our own minds.”
This article is based on the following original source, summarized from the author’s perspective:
Brain activity much lower when using AI chatbots, MIT
boffins find