Apollo data shows the Magnificent Seven widening their revenue-per-employee lead over the S&P 493 between 2019 and 2026, highlighting the growing productivity gap between Big Tech and the broader market.
AI isn’t replacing people, its just making skilled workers somewhat faster and better at their jobs. I use AI pretty heavily in my work these days. I work in cybersecurity and if you go into the vendor room at any conference and throw a lawn dart in the air, it will almost certainly impale some vendor talking about their “AI Enabled” tool. Thing is though, some of those tools are actually really useful. When you are trying to dig the three malicious events out of tens of thousands of routine events which look very similar, an LLM trained on that sort of data can be really good at surfacing the bad stuff. Get yourself a few million log events per day, with attackers trying new stuff all the time, and you start to need something better than static algorithms and heuristics to find evil. Actually getting in front of attackers and stopping their attacks, before data walks out the door, gets even harder.
The problem arises when you have inexperienced folks using those tools and management is not willing to put the money into training and having more skilled workers to back the less trained workers. Part of that is just not knowing how things work and part of it is the AI lying very convincingly. Someone who knows systems reasonably well might look at one of those hallucinations and immediately call bullshit. Someone without that knowledge might assume the AI is right and kick off a pretty disastrous series of events. But this means that the company might just be spending more on workers plus AI tokens than they did on just workers.
And that’s really the crux of the problem for businesses. AI isn’t a magical worker replacement program. Its a tool to enhance workers. And it won’t magically make marginal workers less marginal. It’s like handing a wet behind the ears kid a chainsaw. He’s not going to suddenly become Paul Bunyan and be the worlds most efficient lumberjack. He may not even be OK at the end of the day. On the other hand, get that chainsaw into the hands of an experienced lumberjack, and he might just be cause for better environmental regulations. The problem is, eventually every business will be using AI tools and any competitive advantage they offered will disappear. AI will just be a tool everyone uses and not using it will be like that lumberjack showing up to work without an axe and dressed in high heels.
AI isn’t replacing people, its just making skilled workers somewhat faster and better at their jobs.
The primary motivation of most CEOs for adopting AI is to replace as many salaries as possible. If it has other useful uses, that’s great, but the primary corporate dream is to increase profits through job replacement.
Which isn’t happening. AI can be run autonomously for simple tasks, but not the complex ones which are actually what costs the most. The quality of LLM output depends on the quality of input. GIGO holds.
It will become hopium that you’ll end up sell your kids to be raped by the Epstein class because you’re so desperate for a bowl of hot soup to stave off starving to death.
AI isn’t replacing people, its just making skilled workers somewhat faster and better at their jobs. I use AI pretty heavily in my work these days. I work in cybersecurity and if you go into the vendor room at any conference and throw a lawn dart in the air, it will almost certainly impale some vendor talking about their “AI Enabled” tool. Thing is though, some of those tools are actually really useful. When you are trying to dig the three malicious events out of tens of thousands of routine events which look very similar, an LLM trained on that sort of data can be really good at surfacing the bad stuff. Get yourself a few million log events per day, with attackers trying new stuff all the time, and you start to need something better than static algorithms and heuristics to find evil. Actually getting in front of attackers and stopping their attacks, before data walks out the door, gets even harder.
The problem arises when you have inexperienced folks using those tools and management is not willing to put the money into training and having more skilled workers to back the less trained workers. Part of that is just not knowing how things work and part of it is the AI lying very convincingly. Someone who knows systems reasonably well might look at one of those hallucinations and immediately call bullshit. Someone without that knowledge might assume the AI is right and kick off a pretty disastrous series of events. But this means that the company might just be spending more on workers plus AI tokens than they did on just workers.
And that’s really the crux of the problem for businesses. AI isn’t a magical worker replacement program. Its a tool to enhance workers. And it won’t magically make marginal workers less marginal. It’s like handing a wet behind the ears kid a chainsaw. He’s not going to suddenly become Paul Bunyan and be the worlds most efficient lumberjack. He may not even be OK at the end of the day. On the other hand, get that chainsaw into the hands of an experienced lumberjack, and he might just be cause for better environmental regulations. The problem is, eventually every business will be using AI tools and any competitive advantage they offered will disappear. AI will just be a tool everyone uses and not using it will be like that lumberjack showing up to work without an axe and dressed in high heels.
The primary motivation of most CEOs for adopting AI is to replace as many salaries as possible. If it has other useful uses, that’s great, but the primary corporate dream is to increase profits through job replacement.
Which isn’t happening. AI can be run autonomously for simple tasks, but not the complex ones which are actually what costs the most. The quality of LLM output depends on the quality of input. GIGO holds.
That’s neither what its being sold as nor is it a reflection how tomorrow’s capabilities are being advertised.
And that is why the collapse of current hype is inevitable.
It will become hopium that you’ll end up sell your kids to be raped by the Epstein class because you’re so desperate for a bowl of hot soup to stave off starving to death.
I have no idea what you are trying to say.
That people are only going to be more desperate.
Of course.