AI is destroying your skills

AI is destroying your skills and making you more helpless and reliant on these machines. I'm going to talk about something I'm sure a lot of you are already familiar with as you're using AI more and more in your work. You're starting to lose that critical thinking, problem-solving muscle, right? Like you got this weird slowness in your brain. And if I shut down all the LLMs in the world just as an experiment and ask you to solve a complex programming problem, Would you be able to do that? Or would you just stare at your screen and you won't even know what the first step is, right? There's a term for this thing, and we're going to call it cognitive surrender, right? And you can probably guess what it means too. But basically, people are more and more inclined to just surrender their ability to think and just let AI make all the decisions for them, right? It's lazy, but it's also convenient. And humans are always going to choose convenience and laziness over all other solutions. And you might have noticed this phenomenon with other people who use AI heavily, or even with yourself if you're using AI a lot. Like before AI, you used to think through the complex problems, right? And come up with a solution that works. And it may not be the best solution, but it will still get the job done. And it will be something you create yourself. Now what I'm seeing is people just don't even bother to think about the problem. They just go straight to this prompt machine and type into this magical machine to solve the problem for them. And when you do that you skipping the most important part of the problem solving process which is critical thinking So AI is destroying your ability to think right that what we discussed so far but nobody nobody really cares about this because humans are always going to choose laziness and efficiency we're always going to pick the most efficient and lazy solution to a problem right so no matter how many people raise their voice against this most people are just they're just going to let the ai think through all their problems for them now some people have suggested using guardrails and other systems to make sure you're, let's say, reviewing the AI output, right? Now, since this channel focuses on programming, let's talk about that. In our case, these guardrails would be stuff like reviewing the code, right, that AI generates. Now, let's just be completely honest and cut the bullshit, guys. Nobody is reviewing so much code on a daily basis, right? We're talking thousands and thousands of lines of code being generated every fucking day, right? No human being is going to review all that code. And beyond all that, any workflow that requires a human being to review massive amounts of output for errors is doomed to fail sooner or later. And I say this as someone who uses AI for coding regularly. Like it's very hard to get humans to review massive amounts of output consistently, right? Whether it's code, whether it's writing, marketing material, whatever it is, right? Like our brains just aren't built for this kind of work. If this output is correct, even just 90% of the time, like if this is just correct, 90% of the time, you just going to cognitively check out of the whole review process and you just going to give it a quick glance and just approve the changes you not going to be inclined to carefully review every single line of code and look for any subtle bugs or vulnerabilities and so this just ends up becoming a ticking time bomb that's gonna explode in your face sooner or later right like sooner or later the ai is going to hallucinate and you're going to miss it not to mention you're also losing skills the more you do this you're losing the skill to to write code yourself to solve programming problems, right? And what's even more alarming over here is the more capable the AI becomes, the less capable you're going to become because you're just going to keep asking it for more and more complex tasks and it's going to do a good job at them. And sooner or later, you're going to forget how those tasks even work, right? You're going to lose those skills. Just a quick example. Let's say, let me get some empty space over here. Let's say you're asking it where you should deploy your app, right? And your choices are between Vercel or AWS, right? And you ask the LLM for an answer and the LLM just says you should deploy on AWS, right? How do you know that this is a good decision? Maybe your unique app is, let's say, cheaper and easier to deploy on Vercel, right? because we're selling a serverless platform and your app just happens to be much more suited to that kind of architecture right how would you know the difference right especially if ai is going to atrophy your skills in real time right remember the better this thing gets the worse you get These LLMs because of their hallucinatory nature they can never be fully autonomous because you can never trust these models to not hallucinate wrong output. And so you always need to have the skills and the experience to review the LLMs output. You can't ignore that. And if you make these agents autonomous, meaning they can take final decisions on their own without a human in the loop, You're just surrendering the human reasoning that goes into determining whether the agent's decisions are even good or not. And this is what we were discussing in the beginning of the video, right? This cognitive surrender right there, right? And even though AI is going to help us automate a bunch of complex stuff, there's one huge downside to automation that nobody's talking about. When you automate routine cognitive tasks, right, and you hand them over to an AI agent or something, you stop getting the routine practice, right, the daily practice of those particular tasks. And over time, you lose that skill and you become more reliant on these machines, right? So all of that's why, and I've talked about this in a previous video as well, but that's why the people who are experts in their field, like senior developers and engineers, they're the ones who are producing the most high quality output with AI. And the more that we let these AI agents run wild without actually carefully reviewing its output, the faster we'll lose these skills. So please be careful when you're using AI and don't let it atrophy the skills that you've spent years learning. That's all, thank you for watching.