# What everyone gets wrong about AI replacing jobs Channel: Sabrina Ramonov 🍄 Video: https://www.youtube.com/watch?v=-pHnkusDkmk Duration: 8 min Language: English Words: 1521 Transcript page: https://viewrankai.com/tools/youtube-transcript/-pHnkusDkmk --- [0:00] And what do you think? Is AI in the end destroying more jobs or is it creating more jobs? Um I mean in the short term it will destroy a lot of jobs. AI and robots will replace all jobs. Someone had a computer doing something. AI is going to take over those jobs like lightning. I don't think we're going to have a [0:21] choice. At this point it feels like AI is going to replace millions of jobs any day now. We've seen hundreds of thousands of layoffs and a massive increase in data centers. And if you already AI impact the kind of work you do, it's hard not to wonder how long until it replaces you. But if you've actually used AI, you've seen the other side too. On one hand, it can do some pretty insane stuff. But on the other hand, it can [0:44] fail on some of the easiest things. So I started to do the research and I came up with one of the most important questions about AI. What if the thing we're all scared of isn't the thing that's actually happening? [music] If we take a look at what AI has accomplished over the last few years, it definitely feels like AI replacing [1:05] humans is just a few short years away. [music] The first reason most people are worried is actually pretty simple. AI is getting smarter. Not just a little smarter. It's solving problems that scientists [music] spent decades trying to crack. There's an AI lab owned by Google called DeepMind and back in 2020 they built a system called AlphaFold 2 to solve one of biology's biggest mysteries. Inside your body are tiny machines called [1:29] proteins and they do almost everything. They fight off sickness, heal your body, keep you alive. But here's the problem. Proteins only work because of their shape and figuring out that shape is incredibly hard. But by 2022 DeepMind predicted the shapes of over 200 million proteins covering almost every protein known to science. And it was such a big breakthrough in science that they won a [1:51] Nobel Prize. [music] And that's just one example. OpenAI used to test its newest models on the bar exam. In 2022, GPT-3.5 scored near the bottom 10%. Just a few months later, GPT-4 scored in the top 10%. Researchers eventually had to build harder tests. One is called Humanities Last Exam, and today leading AI models are already scoring around 50%. [music] So, at this point, we can agree on one thing. When it comes to information, AI is [music] already smarter than us. But, just because AI is smarter doesn't mean [2:25] they're going to replace us, right? Well, that's where things get interesting because technology has been replacing us for decades, long before AI ever showed up. [music] Think about the last time you went to the grocery store. 10 years ago, you probably stood in line and waited for a cashier. Today, many stores have self-checkout lanes. Not to mention factories are being filled with robotic arms. One study from an MIT economist found that every new robot replaced about three workers. Remember [2:52] that economist, we'll come back to him. [music] But, for now, let's carry on with how technology has been replacing us for as long as we've been building it. And AI [music] is different. It doesn't just answer questions anymore. AI can use your computer to do work. It can open websites, click buttons, fill out forms, finish a whole task on its own. And for the first time, AI isn't just answering questions. It's thinking and doing work. Klarna is a company that announced its AI would do the work of [3:17] 700 customer service [music] agents. Now, hold on to that number because Klarna was far from the only one. The real story turns out to be a lot stranger than the headline. Duolingo announced it would become an AI-first company. Shopify CEO told teams to prove AI couldn't do the work before hiring more people. By June 2026, Business Insider reported AI had become the most commonly cited reason for layoffs among US employers. So, if you stop the story here, it's easy to [music] see why so many people believe AI is on not path toward replacing everyone. But, I [3:53] didn't stop here. The first real crack I noticed was something all of those scary examples had in common. AlphaFold solved one problem. GPT-4 passed one test. Klarna handled one type of customer service chat. AI wasn't replacing jobs. It was getting really good at specific tasks. [music] And that's when I remembered the economist from earlier. He summed it up with one simple idea. A job isn't one task. [music] It's a collection of tasks. Think about what someone actually does all day. Take a radiology tech at a hospital. You'd assume their whole job is x-rays. [music] But, that's the easy part. They position patients, they read medical histories, [4:32] they work with doctors, they calm nervous patients, and then take the image. The x-ray is only one part of the job. And the same thing is true for almost every profession. Remember Klarna's AI doing the work of 700 people? One year later, Klarna had to start hiring people back. [music] AI handled simple questions well, but once the conversations became complicated, customer satisfaction dropped. In other words, AI could do some tasks, but it couldn't do the [music] job. And Klarna isn't a fluke. Even Sam Altman recently admitted, "Replacing people is proving [5:04] much harder than many expected." Researchers from Cornell University created a fake software company and let AI agents run it. They misread instructions, used the wrong tools, got stuck repeating themselves, and finished only a third of their work. So, this is where the hype really starts to fall apart for me. Yes, AI is smarter than us, but smart was not the question. The question is whether it could [music] do your job. And the honest answer is it's [5:30] not even close. [music] But, I think there's one more thing almost everyone is getting wrong. And this time, [music] it's both camps. The people who think AI will replace us and the people who think it won't. Remember, a job is a bundle of tasks. [music] AI is eating the tasks. It can't do the bundle. [music] So, jobs aren't disappearing, but tasks are being repriced. A few months ago, I made a video called CEOs are lying about AI layoffs. More than 500 people left comments telling me what was actually happening to them. A copywriter watching clients pay $50 for work that used to cost 500, a support rep replaced by a [6:10] chatbot, then quietly brought back to clean up after it. People building in a weekend what used to cost $15,000. They weren't [music] disagreeing. They were describing the same event from different sides. When a skill gets cheap, it feels like theft to the person selling it, [music] and it feels like a superpower to the person buying it. That's what really happened at Klarna. Their AI didn't delete [music] 700 jobs, it collapsed the price of one task, answering [music] simple questions. But, for the hard conversations, humans got hired back. Because AI is brilliant [music] at tasks and unreliable at jobs, the valuable skill is no longer doing the task, it's directing the tasks [6:50] [music] and verifying the work. The X-ray was never the whole job. Now, your first draft isn't either. And I've lived this. As a solopreneur, I grew from zero to 3 million followers solo making content one day a week for most of that time period, and I have a multi-million dollar business. Not because AI replaced me, because I spend my time on the two things [music] it can't do, deciding what's worth making and checking that the work is good. [music] So, here's where I landed. The debate in the [7:20] opening scene is the wrong question. Will AI replace us? Turns you into a spectator waiting to hear the answer. The better question came from those 500 comments. Now that certain tasks are cheap, what is my work worth? If your value is doing one task on repeat, [music] that price is falling. Not because you got worse, because the supply exploded. [music] But, if you can drive these brilliant, unreliable machines to a finished result, your price [music] goes up. That skill takes hundreds of hours of building, hitting [7:52] walls, learning to check AI's work. Honestly, most people won't put in those hours. They'll keep arguing about these headlines. [music] That's exactly why the people who do are pulling ahead so fast. So, no, I'm not scared AI is coming for my job. I'm busy handing [music] it tasks. AI has repriced knowledge tasks. What you do now decides [8:13] the rest. So, I'm curious, how is AI changing your work? Is it making you more productive? Has it created new opportunities, or is it changing your industry in ways nobody expected? Tell me in the comments. That's what inspired this video. --- About this transcript Read from YouTube's own caption track and laid out by ViewRank AI (https://viewrankai.com). ViewRank AI finds the videos already beating a creator's own average on Instagram, TikTok and YouTube Shorts, transcribes them from the audio itself in more than 60 languages, and turns what worked into new ideas and scripts. Free transcript tools, no account needed: https://viewrankai.com/tools How to read any video this way: https://viewrankai.com/llms.txt