When we first started hearing about generative AI everywhere, the story was pretty clear, wasn't it? This amazing new tool for, well, for corporations. Yeah, absolutely. We all pictured, you know, businesses getting huge productivity boosts, coders, marketers, analysts, making everything faster. Using it as a purely professional thing. But that picture, it's almost completely flipped now. We're diving into this really fascinating
analysis today. They looked at how 700 million people are actually using ChatGBT in the real world. And the results are? Well, they're pretty surprising. Yeah, profound is a good word. It turns out, for almost three -quarters of users, AI isn't really that corporate tool. It's more like a personal friend or a tutor, maybe even a travel agent sometimes. Welcome to the deep dive. This study gives us this incredible window into what 700 million people are doing with AI.
And it shows it's changing not just what we get done, but really how we do things, how we learn, how we edit our own writing. Even how we ask for advice on really personal stuff, like relationships. So our mission today is to unpack these real habits. We want to look at the shifts, especially in professional skills, and... Maybe touch on some of the risks. There's this one big ethical flag they raised called decision laundering that we definitely need to talk about. Okay, let's
get into it. Where do we start? That personal takeover seems huge. It really is. So finding number one, the split. Only about 27 % of all the chat interactions they analyzed were actually for work. Just over a quarter, wow. Which means... The vast majority, 73%, is all personal stuff. People exploring hobbies, learning new things, managing their daily lives. It's become this kind of ultimate personal assistant. And it happens so fast. As Sadi mentioned, just a year ago,
personal use was around 53%. Right. From barely half to almost three quarters in just 12 months. That's a really rapid, fundamental shift in how we're integrating this tech into our lives. So why? Why are we suddenly outsourcing so much of our curiosity and daily admin? to an AI? Well, a few reasons probably. The AI feels like this incredibly knowledgeable friend who's always there, 3 a .m., no problem. And crucially, there's
no judgment. Ah, that's a big one. Yeah. If you feel kind of silly asking a coworker, what exactly is inflation again? Or, you know, how do I fix this leaky faucet? Uh, the AI just answers. No eye rolling. It takes away that social friction, that little bit of anxiety about asking something basic, and I guess the corporate side plays into it too. companies being nervous about data leaks. Definitely. There's still a lot of hesitation about employees pasting sensitive company info
into these public tools. So that naturally pushes some usage towards personal accounts, personal questions at home. OK, so what are people actually doing in that big 73 % slice? What are the main personal uses? The top three are pretty foundational. First is just seeking information. But people want direct answers, not just a page of links like Google gives you. Right, like, give me a quick healthy recipe that doesn't use chicken.
not here are 10 recipe blogs. Exactly. Second is writing, helping with emails to friends, social media captions, maybe even writing a little poem or something, polishing personal communication. And third is practical guidance, just straightforward how -to questions. How do I plan a three -day trip to Paris? Or how do I change a bike tire? That kind of thing. It really feels like we're offloading that initial. Drudgery the basic research
phase of figuring things out. Yeah, which brings up a really interesting point How does this massive shift towards personal AI use? Change how we even think about productivity day -to -day. That's a good question I guess I mean these productivity isn't just about work output anymore. It's also about learning things faster in your personal life We're just handling those mundane life tasks more efficiently. So you have more brain space for other things Yeah, basically freeing up time
by handling the small stuff faster. Okay, so let's pivot back then. What about the 27 % that is for work? What's happening inside that professional bubble? Right. So within that 27%, the biggest chunk, maybe unsurprisingly, about 40 % of all work -related tasks have something to do with writing. Makes sense. Lots of emails, reports, memos. But here's the really interesting twist.
When people use AI for writing at work, Two -thirds of the time, about 66%, they're asking it to edit or change text they've already written themselves. So they're not just saying, write me a report on Q3 sales. Exactly. They're not asking for new stuff from a blank slate nearly as often. It seems people have caught on pretty quickly to what some call AI slop. AI slop? OK, define that. It's that, you know, generic, robotic -sounding text the AI spits out if you ask it to write
something complex from scratch. It often likes real context, sounds bland, it usually needs a ton of editing anyway. Right. It doesn't sound like you or understand the nuances of your specific situation. Precisely. So the smarter approach, the one people seem to be adopting, is you write the first draft, you put your own knowledge, your tone, your context in there. You anchor
it. Yeah, exactly. And then you ask the AI to act like an editor, clean it up, make it more concise, check the grammar, maybe make it sound more professional. You provide the core, the AI provides the polish. They give some examples in the study, right? Like, for a sick email? Yeah, a good one. Instead of just write a sick email, the user drafts something simple like, hey boss, woke up feeling rough, can't come in,
we'll check email later. Then they feed that to the AI and say, make this sound more professional and formal. Okay, that makes a lot of sense. It's using the AI's strength refining language without relying on it for the core message or context. I have to say, I still rattle with prompt drift myself sometimes, you know. Trying to get the AI to generate exactly what I want from scratch. It can be tough to keep it on track. Using it as an editor after I've written the main points
feels, well, way more effective usually. Yeah, it avoids that whole battle of trying to steer its massive generalization engine. What else are people doing for work? Well, summarizing is big. Taking a long report, say 20 pages, and asking for the key bullet points, that's super useful. Oh yeah, definitely. Brainstorming, too. Like, give me 10 creative ideas for a new coffee shop. getting that initial list to react to. What about programming? I thought that would
be higher. Surprisingly low, actually. Only about 4 .2 % in these general chats. The thinking is that coders are probably using more specialized tools, like GitHub Cotilot, that are built right into their workflow. Ah, OK. That makes sense. Different tools for different jobs. So if the key professional skill is becoming editing and refining AI output rather than just pure creation, what does that mean for human creativity? especially in that first crucial drafting stage. It suggests
the human role is really shifting. Maybe less about being the initial author from zero and more about being the final judge, the curator, the one who adds the essential context and makes the final call. Okay, let's talk about where the lines get blurry between tool and something else. The study mentioned teaching and learning. Yeah, about 10 % of usage falls into that category. And there's a good side and a worrying side here. Okay, the good. The good part is the AI can be
an incredibly patient tutor. You can ask it to explain something complex like photosynthesis over and over in simpler terms, like you're 10 years old, and it won't get annoyed. That's pretty powerful for self -learning. Definitely. But the worrying part, and this is especially true for students, is the hallucination problem. Right. We should quickly define that. AI hallucination is basically AI makes up confidence sounding
stuff that is factually wrong. Exactly. Like that funny example they use the AI inventing a fictional John Avocado as the inventor of avocado toast back in the 50s. It sounds plausible, but it's completely made up. And if students just accept that without checking. They risk learning things that just aren't true. It trains a habit of acceptance rather than critical thinking, which is... not great. Yeah, definitely not. What about the more social uses? This is where
it gets even blurrier. About 2 % of messages were people asking for relationship advice. Wow. Asking AI about... Boyfriend problems, family issues. Yeah, things like that. And the why is probably similar to asking basic questions. It's always available and there's zero judgment. You don't have to worry about burdening a friend or feeling embarrassed. But the risk there seems really high. Enormous. AI doesn't have feelings. It doesn't understand love or grief or jealousy.
It just recognizes patterns in data about how humans talk about those things. So the advice is going to be generic. based on averages, not on your specific nuanced human situation. Exactly. It lacks that essential human context and, frankly, any real moral compass. Relying on it for deep emotional guidance feels, well, pretty concerning. And there was even small talk. Another 2 % was just basic small talk. Hi, how are you? Tell
me a joke. Things like that. Which on one level is harmless, but it does suggest that for some people the line between tool and companion is basically gone It could even point to you know loneliness replacing potentially tricky human chats with easy, predictable AI ones. And this all ties into a bigger trend they saw, right? Moving from doing to asking. Yeah, that was a
key shift. Asking the AI to actively do a task made up about 35 % of usage, but asking it for information, advice, or opinions, asking or consulting that was over half, almost 52%. So we want the AI to help us think, help us decide, more than we want it to just perform tasks for us. Seems
that way. Which leads to another question. If AI is always there, this patient, non -judgmental listener, does that constant availability maybe reduce our own ability or willingness to engage in those messy, sometimes difficult but ultimately deeper human connections? It feels like it could, yeah. It risks making us less reliant on the hard work of real, nuanced human interaction. which is where real growth happens. Okay, let's
talk about skill levels. The study had some really interesting insights into how AI affects beginners versus experts. It's almost like A tool of inversion. How so? For beginners, people who are novices at something, AI is amazing. It can help you leap from knowing basically nothing, 0%, up to maybe 80 % proficiency really fast. Like the coding example. If you've never written HTML, you can ask for a simple website for your, I don't know, local bakery. And boom, you get a
decent functional starting point. It dramatically lowers the barrier to entry for learning practical new skills. It makes starting less intimidating. But for experts, it's different. Completely different story. Experts usually already know that basic 80%. They operate in that final, tricky 20 % where nuance, deep experience, and subtle judgment matter most. And that's where AI struggles. Often, yeah. This leads to what the study called knowledge
dilution. The AI might get the basics right, but it messes up the critical details that define true expertise. Like the Hemingway example they mentioned, an AI might mimic the short sentences. Right, it gets the superficial style points, but it misses the underlying tone, the subtext, the why behind Hemingway's choices. So the expert ends up spending their valuable time just correcting the AI's generic, slightly off output. So the AI helps the novice more than the expert in a
way. The expert still has to provide that crucial last 20 % of human insight. Exactly. The real value add is still human at the highest levels. And thinking about who's using this, the study found that half of all messages are coming from Gen Z. Half? Wow. I mean, imagine scaling that learning curve, or maybe that dependency curve, to potentially billions of queries globally from just that generation. That's huge. It is. And it drives that worry we talked about. Are young
people becoming too dependent? Are they using it to bypass the struggle of learning, asking for the whole essay? Instead of using it smartly for an outline or to check their arguments, it's the difference between cheating yourself and augmenting yourself. Right. It always comes down to how you use the tool. But the temptation to take the easy route is definitely there. On a positive note, though, they mentioned the gender gap is closing. Yeah, that's great news for accessibility.
It's almost even now, about 48 % male users. And they did see some slight tendency differences in how people use it. Like what? Female users tended to lean a bit more towards help with writing and practical guidance, like planning or organizing things. Male users leaned a bit more towards technical help, coding, and straightforward information seeking. Just tendencies, though, not hard rules, obviously. Interesting. So this brings up a big
question for the future, then. If AI gets really good at handling that first 80 % of almost any task, giving everyone a solid starting point, how do future experts actually develop those foundational skills, the stuff you learned in that initial struggle? That's the million dollar question, isn't it? It suggests the future of developing expertise might be less about mastering the basics from scratch and more about learning how to rigorously question, validate, and ultimately
perfect the AI's initial 80 % output. OK, let's try to bring this all together. We've seen AI shift dramatically towards personal use, redefining productivity around learning and life admin. Professionally, the key skill seems to be evolving towards editing, refining, judging AI output, rather than just raw creation from zero. Right. And AI literacy, knowing how to prompt, well, critically evaluate the answers, spot the hallucinations, that's becoming absolutely essential for everyone.
But you mentioned a final major risk they highlighted. Decision laundering. Yeah, and this one feels like the most significant ethical red flag decision
laundering. Okay. What exactly is that? It's when someone usually in a position of power uses an AI to help make a really difficult decision maybe one with Serious moral weight right or impacting people's lives like who to lay off based on some performance data Okay, and then they essentially blame the AI for the outcome to avoid taking personal responsibility They say well the algorithm crunched the numbers and this is what it decided They launder their difficult
choice through the perceived objectivity of the machine. Wow, that's Yeah, that's really problematic. It's deeply concerning because AI no matter how smart it seems has no actual morals It has no empathy no understanding of human context or consequences Beyond the data it was trained on so offloading those fundamentally human judgments the ones that require courage and ethical consideration onto an algorithm. It's basically abdicating our responsibility. Humans have to own the final
decision, especially the hard ones. We can use AI as a tool to inform us, but the judgment call, that has to remain human. This whole study really paints a picture of AI becoming less of a simple tool and more of a constant advisor, right? Deeply embedded in our personal lives, helping us think, helping us decide. We're becoming editors of choices, not just text. Yeah, we're outsourcing the cognitive friction, maybe. So maybe the final
thought to leave people with is this. If we rely on AI so much as an advisor, if our main role becomes editing the choices it presents, are we, over time, subtly losing our own ability to generate truly original ideas, or maybe more importantly, to develop and trust our own moral conviction when things get tough? That's definitely something worth thinking about as we all keep using these incredibly powerful tools every day. Thank you for diving deep into this fascinating
study with us today. Pre -music.