Why AI Will Never Replace Our Jobs: The Human Edge in the Age of Automation

Ethan
0

AI TAKING OVER

“AI will replace our jobs”. This is the most ridiculous statement I’ve ever heard. Ever since ChatGPT came out, there’s been this general fear that AI will replace everyone’s jobs and that all of our skills will suddenly become completely useless. And it’s not surprising why people think this. 

For one, people have seen these tools in action first hand. They’ve seen realistic photos spawned out of thin air, 3000 word essays being typed out in 30 seconds, and hundreds of line code being drafted up like it’s nothing. 

At the same time, we also have all these articles that are like AI that will replace 47% of all US jobs within the next 20 years or AI is set to replace 300 million jobs or AI could replace up to 8 in 10 jobs, experts say. And it’s not just journalists and media organizations who are making these bold claims either. 

AI could replace up to 8 in 10 jobs


We have valedictorians from the most elite colleges and CEOs from the biggest tech companies that are all propagating this notion. Elon, for example, has been claiming that we’ll all need universal basic income because robots will take all our jobs. Meanwhile, Sundar Pichai has been preaching that AI could be more profound than fire or electricity. 

But, you wanna know a secret? Sundar has been saying the exact same thing for over 5 years now. That’s just his go-to statement when he’s pitching the importance and impact of Google’s AI. Also, remember that study that was suggesting that 47% of all US jobs will be replaced by AI within the next 20 years? Well, that study was published 10 years ago, so theoretically, we should already be feeling the pain. 

Yet unemployment is still at near all-time lows despite recession fears, high inflation, high interest rates and apparently AI taking over. And the reason is obvious, AI is not going leave all of us jobless relying on some sort of UBI from the government because AI will not be replacing the value of humans, let me explain. 


Why People Are Switching From Android to iOS :The Future of Smartphones


HUMAN RESILIENCE

To understand why AI doesn’t stand a chance, all we have to do is look back at history. In fact, we can rewind 12,000 years to the year 10,000 BC. This probably feels like forever ago but in the context of human development, it’s actually extremely recent. 

HUMAN RESILIENCE


Modern humans are thought to have come about sometime within the past 200 to 300,000 years, so 10,000 BC is part of the most recent 5% of human history. Up until this point, the life of the average human was pretty simple. Virtually every single human spent the vast majority of their time on survival. They would hunt for animals, gather berries and water, create weapons using rocks and sticks, and ignite fires to keep them warm and cook food. Basically all of their time was just spent on survival. 

But, somehow, around this time, some idiots around the world figured out that if you put a seed into the ground, pour some water on it, and give it some time, you could produce food on demand. Now, imagine if these nomads were like “Brooo, if we could just scale this agriculture thing, man no one would have to work anymore, especially if we got good at it. 

Imagine if just 1% of the population could make enough food for everyone. Then we could all just sit around and never have to worry about survival again.” 

Why AI Will Never Replace Our Jobs


Oh wait, that’s exactly how it is today. Most people in western countries never have to worry about food but 98.7% of Americans are not unemployed. In fact, it’s the other way around.

farmer salary in US

 

The people who produce all of our food are usually the least paid, the least respected, and the least valued even though they often work the longest, the hardest, and are essential to our survival. Now, theoretically speaking, we could’ve made that nomad dream a reality. 

We could’ve had some sort of system where all of us only have to work like 1 week every year on the fields. The rest of the time, we could just sit in our huts and have food handed to us. Thinking back, this did start to happen. As the efficiency of agriculture skyrocketed the amount of free time people had also skyrocketed. 

But people didn’t just opt to sit around; instead, they started mastering crafts. They became skilled bakers, carpenters, butchers, soldiers, stonemasons, and tailors. We spent our time building entire cities with hundreds of roles, jobs, and mechanics all centered around these agricultural hubs. And agriculture is just one example. 

Such shifts have occurred thousands of times during human history during the advents of concrete, electricity, light bulbs, telephones, locomotives, cars, planes, trucks, highways, vaccines, computers, smartphones, the internet, and who knows what else.

I mean, just think about this. 60% of all workers are currently in occupations that did not even exist in 1940. Let me phrase that in a different way. The majority of all working people are working at jobs that did not even exist a single lifetime ago. So, how in the world is the advent of AI any different from what we’ve been doing over the past 12,000 years? 

Well, many would argue that it’s the speed of AI progress that makes it so dangerous. Not only will AI replace low level factory jobs or cashier jobs but it’ll also replace white collar jobs like doctors, lawyers, and engineers. And to do this, I will once again call BS. Let me explain. 


Why Amazon Is Happy They Are Getting Sued :How It Could Benefit Them


THIS TIME IS NOT DIFFERENT

People often like to claim that it’s the speed of AI advancement that makes it so disruptive but this too is easily disprovable. Yes, AI is advancing faster than anything we’ve seen before, but guess what, so did every other revolutionary technology. That’s kind of why it’s revolutionary. 

It took agriculture 10,000 years to catch on but it only took a thousand years for cities to catch on. It only took hundreds of years for electricity, only decades for tech, and only took years for smartphones. So, it’s only natural that AI is able to catch on within months but really it’s not even that fast. 

THIS TIME IS NOT DIFFERENT


Researchers predict that AI’s computational power will double every 6 to 10 months. But this isn’t even all that crazy when you consider Moore’s law. Back in 1965, Gordon Moore, the eventual founder of Intel, predicted that the number of transistors on integrated circuits will double every single year and he was spot on. 

the number of transistors on integrated circuits will double every single year


Over the next 10 years, the number of transistors on microchips jumped from a few hundred to tens of thousands. But after this fast initial jump, Gordon would revise his prediction as progress slowed to transistors doubling every 2 years which has largely held true over the past 50 years. 

So, objectively speaking, the intelligence of tech was advancing the fastest back in the late 60s and early 70s. But if you extrapolated that to argue that this was the fastest period of tech adoption, you would’ve been completely wrong. Microsoft and Apple weren’t even a thing yet. And there were still decades until Google and Facebook would even be founded. 

Really, it wasn’t till the 2000s and especially after the launch of the iPhone, that tech adoption became ubiquitous. Now, this isn’t to say that it’ll take decades for AI to become ubiquitous. It’s obviously advancing and being adopted faster than computers but it’s still naturally going to take time. 

The bottomline is that with each and every one of these revolutionary introductions, it’s humans who are creating these changes and propagating them. So, naturally, these changes will be limited by the laws of supply and demand. 

Just because you make a great AI doesn’t mean that it’ll be instantly adopted. It’ll only be adopted when there is a demand for it and there will only be a demand for it when the technology and a large portion of society are ready for it. So, this time is no different. 

I’d wager that AI will actually catch on much slower than you might expect and the best example of this is the failed automation of factories. 


How Microsoft Was Blackmailed into Acquiring Nokia: Unveiling the Inside Story



AUTOMATION THAT NEVER HAPPENED

Before everyone was worried about AI taking over white collar jobs, people were scared about automation taking over factory jobs. And this makes sense given that back in 1979, factory employment was at an all-time high. 

AUTOMATION THAT NEVER HAPPENED


19.6 million Americans worked factory jobs, so fancy robot arms and assembly lines obviously made them feel uneasy. And they were right, factory jobs did indeed decline by 35%....over the next 40 years. 

In other words, all of these crazy computer chips that went from a thousand transistors to 50 billion transistors weren’t even able to eliminate half of factory jobs over the next 40 years. They weren’t even eliminating 1% per year. And that’s not even accounting for all of the factory jobs that were shipped overseas as opposed to being replaced by robots. 

The number one reason for this is that robots suck at handling edge cases. Our AI overlord Elon is most familiar with this as he was nearly driven into bankruptcy because of it. You see, as a staunch AI and automation supporter, I don’t think you’d be surprised to hear that Elon tried to create a completely automated Tesla factory. 

He hired the best software and AI engineers on the planet, sunk billions of dollars into building his grand gigafactory and had hundreds of thousands of people waiting to receive their Model 3s. Theoretically, all he had to do was press the go button on the factory, deliver a bunch of cars, and make a crap ton of money. Easy right? 

Well, all of this changed when he actually pressed the go button. There’s really no pleasant way to put this. The reality was that his automated factory was absolute trash. One small misalignment in in the sheets of metal and the robots couldn’t screw in the bolt. One small misalignment in the paintshop, and the robot would paint the window.

Such matters are trivial for humans to handle but for robots, such matters made them completely ineffective. Elon found himself constantly stopping the production line to fix all the errors that these robots were making. This destroyed the productivity of his factories, leading to Tesla burning billions of dollars and getting single digit weeks away from bankruptcy. 

Do you know how Tesla recovered from this manufacturing hell? Well, everyone including the tech guys had to work on the factory line and work 100 plus hours every week just to get Tesla to stop burning cash. Aka, Tesla had to deautomate their factory, and Elon would end up admitting “Yes, excessive automation at Tesla was a mistake. To be precise, my mistake. Humans are underrated.” 

Elon would end up admitting “Yes, excessive automation at Tesla was a mistake.


If this is the case with simple repetitive factory automation, how in the world is it realistic to assume that AI can just automate software engineering just because it can write algorithms or solve hard leetcode problems? It really makes no sense at all. 

The vast majority of a software engineer's time is not spent on doing the fun obvious stuff. In fact, the obvious stuff takes no time at all. The vast majority of time is spent on doing boring but essential stuff. How do we handle all of these different errors? How do we scale this code seamlessly? How do we address all the edge cases? 

This is precisely where engineers spend their time and this is precisely what AI cannot do. I mean, just look at Tesla autopilot. They were able to get lane assistance on highways working extremely fast, but getting it to work everywhere else is proving extremely difficult. 

Now, with that being said, it would also be stupid to think that AI won’t affect software engineering. It absolutely will, but we’re talking about a slow shift away like with manufacturing as opposed to something that just transpires overnight just because ChatGPT has a 100% accuracy on Leetcode. And this brings us into the last argument regarding AI replacing jobs which is that it’ll replace more jobs than it creates. 


Why Patients Are Suing BetterHelp: The Dark Side of BetterHelp


JOB DEFICIT DEBUNKED

The idea that AI will take away more jobs than it creates is really quite stupid because AI nor any new technology is not a zero sum game and it’s quite limiting to look at it from that perspective. Take the internet for example. In the early days, the internet was largely used for research purposes. 

What if people thought that the internet would take away jobs just because it would put a lot of libraries out of business and in return only give a 1000 jobs to coders? Obviously, that statement is laughable today, but is it really that far fetched back in the day? 

40 years ago, no one would’ve even been able to fathom the idea of multi-hundred billion dollar social media companies, ecommerce platforms, cloud providers, database solutions, streaming giants, fintech offerings, food delivery, ride sharing, online booking services - the list really goes on forever. It’ll be the same way with AI. 

From our current position, we can only envision how AI will disrupt the jobs and careers that we already know. We have no chance of even conceptualizing all of the jobs and careers that AI will create. 15 years ago, it wasn’t even possible to make money as an “influencer”. 

Today, being a YouTuber is the number 1 dream job in America. Also, not only did tech create a bunch of jobs but it created a new first class. Think about all of the richest people in the world. They are owners of tech companies, executives of tech companies, employees of tech companies, partners of tech companies, affiliates of tech companies, investors of tech companies, and so on. 

Truth be told, if you weren’t involved in tech within the past 20 years, you were probably left behind. And it’s no wonder why. These companies are able to create value faster and at larger scales than ever before. 

This will only be truer with AI as AI makes work magnitude more efficient. So not only will AI not make everyone jobless hobos but it’ll create an unprecedented number of insanely highly paying jobs along with hundreds of trillionaires because the only way that it plays out any other way is if humans fundamentally change. 


What Happened To Nokia And How Is Nokia Even Still Alive?



If humans fundamentally change and somehow become satisfied with what we already have. The current tech we have, the current economy we have, and the current society that we have. As long as people are always striving for more and more, you can rest assured that we’ll never end up in some sort of AI dystopia where robots do everything and humans do nothing. 

There will always be something for us to do and if there isn’t, you can be sure as hell that we’ll make something to do and that’s why the idea of AI replacing everyone’s jobs is simply ridiculous. AI content, however, is ruining YouTube and Blogging.

Post a Comment

0Comments

Post a Comment (0)