The AI Employment Explosion, Part 1
The AI Boom Won't Remove Labor
AI makes the future of work look scary. Headlines constantly warn that “jobs will disappear.” Many predict that Universal Basic Income will become necessary for survival. In this vision of the future, only billionaires make money, while everyone else lives at the goodwill of the government. The government taxes AI and redistributes it back to the helpless masses, who have nothing to do with their lives but watch Netflix and play video games.
But what if these predictions never come true? What if AI creates more jobs than it destroys, and work isn’t going anywhere? In fact, the CEOs banking on mass layoffs may not only be mistaken, but setting themselves up for failure.
This series is dedicated to showing why the AI doomer scenario is wrong. I am not betting on AI failing or succeeding. If AI fails, then the future of work barely changes and the doomers are wrong by default. The real question is what happens if AI succeeds. This series assumes it does and argues that its success will create more jobs than we can imagine.
AI Agents to the Rescue?
Chatbots like ChatGPT and Claude are just the beginning of the AI revolution. The prompts they handle can speed up research, simplify writing, and even generate images from a short description. But that’s not what transforms the world. Real change comes from AI agents that act on their own, run 24/7, and handle a wide range of tasks. The arrival of OpenClaw 1 looks much closer to what the AI doomers are imagining: an agent that can operate independently and doesn’t need you.
The real test isn’t chatbots. It’s whether AI agents can run on their own. If they can, then the doomer scenario becomes plausible. And if anyone should be able to trust an AI agent to take over real work, it’s someone like Summer Yue, Meta’s Director of AI Safety and Alignment. Right? 2
But Summer found out the hard way that OpenClaw can’t be left alone. She tested it on her personal email first. No problems. So she moved it to her business inbox. That’s when all hell broke loose. It immediately started trashing thousands of emails. She told it to stop. It didn’t. She had to run to her computer and shut it down herself. Fun times.
To see how this actually played out, take a look at her conversation with OpenClaw.
Who wouldn’t want AI to nuke all their old emails? None of it’s important, right?
At least OpenClaw said it was sorry. It promised it would never do it again. Pinky promise.
AI Will Get Better. Right?
Of course it will. AI agents will improve and likely reach mass adoption. But what level of failure is anyone willing to tolerate? Would you install something like OpenClaw into your business email? What if it were right 99.9% of the time? What about 99.99%? Would that be enough? Or would you still feel the need to keep an eye on it?
Now imagine installing OpenClaw into an X-ray machine or a heart monitor. Or putting it in charge of BlackRock’s biggest funds. Or making it the spokesman for your family business. No need to watch it. No need to check its work. What could possibly go wrong? It would never do anything out of line. Right?
Now imagine AI agents running millions of small businesses without any human supervision. There are an estimated 300 million businesses worldwide, so imagine if only 10% adopted AI. That’s 30 million agents.
Even at a 99.999% success rate, 30 million agents making 100 decisions a day would still produce 30,000 failures. And that’s before we even consider hospitals, banks, governments, or personal users.
How many people do we need just to babysit 30 million agents?
And remember, today’s AI agents fail up to 63% of the time on complex tasks, even assuming only a 1% error rate per step. 3 That’s nowhere near 99.99%.
If you have any doubts about AI being able to run on its own without monitoring, then you already understand why the human element isn’t going anywhere. If we can’t trust agents to operate freely, someone has to oversee them. And that person must be an expert in the domain the AI is working in and know when it’s not behaving correctly. That’s not a disappearing job. That’s a high-paying one.
Conclusion
The purpose of this article is not to convince, but to build intuition for why humans are still needed. For now, just consider what could go wrong if we allow AI to operate without oversight, and begin to see why the human element isn’t going anywhere.
Join Crypto Confidence
If you’re thinking about where all of this is going, you’re not alone. One of the biggest shifts happening right now is that understanding these systems is becoming a skill in itself. That’s a big part of what I focus on in Crypto Confidence, where I teach people how to actually use AI and crypto instead of just reacting to them.
If you want to learn how to operate in this new world, you can join Crypto Confidence here.
Final Word
AI won’t replace you. But someone using AI might.
"For every complex problem there is an answer that is clear, simple, and wrong." — H.L. Mencken
Disclaimer:
The information in this publication is for educational purposes only and does not constitute financial, investment, or legal advice. Always do your own research before making any financial decisions. Cryptocurrency investments carry risk, and past performance is not indicative of future results. I actively invest and trade in the crypto markets, and my personal portfolio and holdings change frequently. Nothing I share should be interpreted as a guarantee of performance or a recommendation to buy or sell any asset.
https://openclaw.ai/
https://www.businessinsider.com/meta-ai-alignment-director-openclaw-email-deletion-2026-2
https://venturebeat.com/infrastructure/ai-agents-fail-63-of-the-time-on-complex-tasks-patronus-ai-says-its-new