Is AI Making Us Dumber? (Can Teachers Save Us?)

[Music] The Nickley. >> Hello everyone. It's Nick and this is the Nick Stanley Show where we explore the future of learning, thinking, and artificial intelligence. Today we're doing something a little different. We're going to answer your questions about AI. If you want to submit your own question for a future episode, just drop it in the comments on YouTube or Spotify or leave it in an Apple podcast review or you can send it directly through our website at nickshow.podbean.com. Okay, let's dive into our first question. Question one comes from Rick. Is AI making us dumber? That is a brutal question and I kind of love it because it captures what so many people quietly fear that this technology is outpacing our capacity to keep up with it.

Here's my take from talking to so many people in the AI space. AI is not making us dumber. It's removing the friction that used to make us smarter. So when you had to draft an essay, you had to wrestle with the ideas, not just the grammar. You'd think, rethink, reorganize. That mental friction builds neural connections. It literally strengthens your brain. Now, Chad GPT can do the wrestling for you. And if you never step back into the ring, so to speak, your cognitive muscles start to atrophy. It's like having a personal trainer do your push-ups for you. Daniel Conorman's work here is insightful.

So he describes two operating systems for the brain. System one and system two. What's the difference? Okay, ask yourself, what's 5* 5? 25. That's system one, right? It's automatic. Notice how this feels different. What is 13 * 17? Really try to figure it out. See how that feels different? That's system two. That's the slow, deliberate, analytical mind struggling with that question. Here's the thing, though. One, you once had to use system 2 to figure out 5* 5. Only with repetition and practice did that become automatic. 13 * 17, you never drilled that one enough to transfer that knowledge to the automatic system one.

Now, Conorman would say learning is the process of transferring new information from system two, the slow, deliberate, analytical mind, to system one, the fast, intuitive one. That transfer only happens through struggle. There is no shortcut. You can't outsource it. You have to get things wrong, fix them, test them, talk about them, wrestle with them. Only then does the knowledge become automatic. So, I think the real question isn't is AI making us dumb. It's will we still choose to struggle? Because if you learn to love the struggle, that cognitive friction, AI can be an amplifier, not a crutch, it can accelerate your path while everyone else is taking shortcuts and quietly eroding their long-term capacity to think for themselves.

And we actually have data on this. MIT and Stanford studies found that people who collaborated with AI, who checked its reasoning, asked it to critique their drafts, used it as feedback, produced work that was more accurate, more creative, more original than both AI alone or humans alone. The key is the loop. Think, prompt, reflect, adjust, repeat. So if AI gives you an answer, your job is to learn the question behind it. Examine the assumptions baked into that output. Identify where it's confident but potentially wrong and figure out what question would have led to a better answer. Pro tip here, ask the AI to help you find the right questions.

Seriously, tell it, "Help me think more clearly about what I'm missing." This is a great prompt. That little shift forces you to engage with the reasoning instead of just ingesting it. And bottom line, the future does not belong to people who know the answers. It belongs to those who know how to question the answers they're given. Thank you for the question. I really like that one. Question two from Laya. Is AI going to replace teachers or finally make education fair for everyone? Interesting. That one hits hard for a lot of people because I think it gets to the soul of what we value in education.

Look, AI can teach or at least it can simulate teaching and it can earn perfect test scores on AP exams. The question though is does that translate to teaching? I'm going to say no. Learning is not access to more information in a faster way as we discussed in question number one there. Not to mention AI doesn't care. It can't look at a student and say, "Hey, I see you're scared, not stupid, not lazy." It can't sense when a student is anxious or ashamed or just having a bad day. It can't look you in the eye and say, "Hey, I know you can do this. It can't understand that your dog was diagnosed with terminal cancer and so you need to take it easy today." That emotional layer, that's what psychologists call relatedness.

It's one of the three pillars of motivation and self-determination theory, which is just a really fancy way of saying most of teaching is the soft skills. It's not a transaction, it's a dance. And history gives us a reality check here as well. So every few decades, a new technology arrives, promising to revolutionize education. Thomas Edison said movies would make textbooks obsolete. And then it was radio. And then it was television and the internet and the mass open online courses. They were all touted to revolutionize education. And while the tools changed, the base experience of teaching stayed the same.

At its core, learning is still wrestling with new knowledge, making connections through reading, writing, talking, and testing. And there is not a shortcut to that process. So yeah, we need teachers for all of that. Take writing as an example. We often write not to communicate what we already think, but to discover what we think. That's why writing and conversation are irreplaceable cognitive tools. They expose the gaps in our own reasoning. So, here's my plea. If you're young right now, and it's never too late, learn to read deeply and write clearly. Learn to tell stories. Learn to speak in front of people.

Those skills are going extinct in some circles, but they're the ones AI can't replicate. the ones that let you direct the machine rather than being directed by it. And as those skills become more rare, they will separate you from your peers in an AI saturated world where others are taking shortcuts. So the future of education, I don't think, is AI or teachers. It really should be AI for teachers. Let AI do the grading, the analytics, the repetitive drill work, the drudgery, so that teachers can do what only humans can. Connect, inspire, interpret. Most education tech has been obsessed with scaling teachers.

But maybe teachers are like plumbers. We don't want to scale one legendary plumber to serve a million homes. You train a million good plumbers. And maybe we need more educators, not fewer, that are empowered by AI in order to focus more on that human connection that makes growth possible because that's what good teaching is. Thank you for the question. Really, really appreciate it. Question three comes from Kurt. It seems to me that the entire AI story is figuring out what to do with all our free time before AI kills us all. Yes, that is one way to put it. And honestly, that tension between potential utopia and annihilation pretty much sums up a lot of my recent conversations about AI.

I often think back to when I spoke with Donald Hoffman and Anil Seth and Howard Gardner and they all echoed something that stuck with me. Intelligence doesn't guarantee wisdom or morality. We are racing to build machines that can outreason us and we should not assume that smarter machines will care about us. Coming soon is an episode with Nate Sores, author of If Anyone Builds It, Everyone Dies. He runs the machine intelligence research institute and he argues that the odds of super intelligence wiping out humanity are far higher than anyone profiting from AI is willing to admit publicly. And this isn't an AI doomer.

This is someone who is in the trenches, who deeply understands how these machines are grown, not programmed, which in itself is an incredible insight. We don't necessarily know what we're growing. Elon Musk once referred to AI as summoning the demon because of the possibility that these machines that are smarter than us could one day wipe out all of humanity. I guess he got over that fear. In any event, I also have an upcoming conversation with Christopher Somerfield, a Oxford neuroscientist at DeepMind, who offers a counterpoint. It's not clear that the current architecture of AI, massive data, scaled transformers, more compute can even reach super intelligence.

The performance jumps we saw in 2023 and 2024 seem to have plateaued a bit. That doesn't mean the danger's gone. It just means the exponential curve may have hit a temporary bend. And depending on when you hear this listener, uh, this could obviously change overnight. Somefield contends that while the existential threat makes headlines, the more immediate threat is displacement. McKenzie projects that by 2030, automation could absorb up to 30% of all work hours. And to put that number in perspective, the unemployment rate in the US during the Great Depression in 1932 was 25%. Now, for people who learn to augment their abilities with AI, it will no doubt be a massive productivity boost.

But for those unwilling or unable to learn the tools, it's a cliff. Entire job categories could vanish. Economically, this could look amazing on paper. higher GDP, cheaper services, but socially that could be catastrophic. A generation of people who feel discarded because their skills no longer matter is a recipe for political extremism and despair. We have to build retraining pipelines and safety nets now, not later when we're going to need them. Back to the question, is AI going to destroy us or save us? Maybe neither. Maybe it's just holding up a mirror, magnifying both our brilliance and our blindness.

We'll see for ourselves, for better or for worse, in the intelligence that humanity creates. So perhaps the future doesn't hinge on whether AI wakes up. Maybe it hinges on whether we do in time. Kurt, thank you for that question. It's a bigger question than I can possibly answer, but I did my best to answer it with data. And for this first goaround with answering your questions on AI based on people that I've had the pleasure of speaking to on the podcast. Thank you for your questions and for tuning in today. If you would like to submit your own question for a future episode about artificial intelligence, just drop it in the comments on YouTube or Spotify or leave it in a Apple podcast review.

We'll be monitoring all of those. You can also send it directly through nickshow.podbean.com. And as always, please visit our sponsors in the description and rate us five stars on whatever platform you are hearing this. Have an excellent day and thank you so much for tuning in. Okay, everybody. Until next time, ask questions, don't accept the status quo, and be curious. [Music] The Nick Stally Show.