AI Will Not Replace My Mind, Just Extend My Skills

I built a website in collaboration with AI. Every piece of data on it, every company, every CEO compensation figure, every political donation was reviewed by me before it was published. I made the final call on what goes live. The AI gathered. I built and decided.

That distinction matters more than most people realize right now.

When companies started spending more time selling their own employees on AI than actually learning it, I began to feel the disconnect. I watched colleagues spend months building tools that already existed, watching teams that weren't talking to each other, and being in initiatives no one could measure. The statement I kept hearing was 'we're still figuring out the ROI', while attendance at AI workshops. Then watching AI become a portion of performance reviews. That contradiction is what started this.

AI is a tool. In its current iteration, it is not the future, it is infrastructure. When you build a house, you don't hand every worker a hammer and tell them their performance will be measured by how often they use it. You let people use the right tool for the right job. AI is no different. It is a specialized tool being sold as a universal one, and companies are paying the price for that confusion, they just haven't admitted it yet.

The evidence is already there. Klarna cut 700 customer service workers citing AI automation, then partially reversed the decision when customer satisfaction collapsed. The workers weren't redundant, they were the product. When you remove the human from the first point of contact for a paying customer, you are not cutting costs. You are cutting value.

And yet the pattern continues. Because the people making these decisions are measuring AI's return on investment in headcount reduction, not in innovation, not in customer retention, not in growth. They skipped the first step: defining what success actually looks like.

Here is what concerns me most, the skills that aren't being developed.

Malcolm Gladwell's research in Outliers popularized the idea that mastery takes 10,000 hours. The underlying science, drawn from deliberate practice research by Anders Ericsson, is clear: expertise requires sustained, effortful repetition. There are no shortcuts.

When companies replace support teams with AI chatbots, they are not just cutting jobs.

They are cutting the pipeline of people who learn a product deeply enough to eventually build it better.
They are cutting the internal talent worth promoting and research consistently shows that promoting from within is significantly more cost-effective than external hiring.
When employees are forced to use AI for tasks they haven't yet learned themselves, you remove the friction that produces expertise.

I studied adult learning frameworks professionally. The one I keep returning to is Experiential Learning, the idea that real understanding comes from doing, not receiving. AI, in the way most companies deploy it right now, removes the doing. That is not efficiency. That is the slow erosion of organizational knowledge.

The data supports this concern.

After three years of widespread AI deployment, a survey of 6,000 executives across four countries, conducted by NBER and BCG, found that most reported no measurable productivity impact at their own companies. Deloitte's research found that only 25% of organizations describe their AI deployment as transformative. By Q1 2026, the market had started pricing in this gap: companies with vague AI language in earnings calls saw stock drops, while those with specific, quantifiable metrics outperformed.

So where is the return? Companies selling AI are making money. Companies using AI are still looking for theirs.

I built Nexus Sunset because I believe people deserve accurate information to make their own decisions. Not my conclusions. The data. The SEC filings. The earnings calls. The pay ratios. Cross-referenced and sourced so you can follow every claim back to its origin.

What I found: $779 million in CEO compensation across 22 companies. 316,630 jobs cut. Every one of those companies cited AI. Most of them were profitable when they made those cuts.

That is not transformation. That is cost reduction with better branding.

AI will not be a replacement for my mind, just an extension of my skills.

I use AI to gather data, because there is more data than any person can process alone. I use it to organize and clean. Then I evaluate, I decide, and I publish. The human stays in the loop, not as a formality, but as the point. That is what Human-in-Command means.

How are you measuring your company's AI investment?
Has anyone defined what success looks like beyond headcount reduction?
When was the last time a significant AI initiative produced something that didn't already exist?

These are not rhetorical questions. They are the ones worth asking before the next round of cuts gets announced.

If you want to see the data: nexussunset.com

Additional Resources:

TripleBee | Human-in-Command