I can happily talk to AI for hours a day, but I notice that most people hate AI.
I expect that AI will live up to its billing and will revolutionize the economy more rapidly than any previous technology.
When Gemini 3.0 came out November 18, I learned about it on Youtube and on X. The consensus that it was the best came so fast, that I played around with it and discovered it was superior to Grok and ChatGPT.
Until November 18, the relative merits of various AI chatbots fascinated me, but now Gemini is just so far ahead, there’s no longer a discussion to be had.
Gemini 3.0 killed the debate because it solved the “Reasoning Gap.” Before this, you were effectively choosing between different flavors of “predictive text.” Now, you are interacting with a model that can hold a complex logic chain in its head without hallucinating halfway through.
Gemini’s massive context window (likely 2M+ tokens in my timeline) means you can dump entire books into the prompt.
Competitors: You have to chop the file into pieces. The AI loses the thread.
Gemini 3.0 holds the whole timeline. It sees the connection between page 5 and page 500. That isn’t just a “better chatbot”; that is a structural advantage that a human brain cannot replicate.
Other models try to be conversationalists while Gemini 3.0 acts like an analyst.
I don’t need a chatty friend. I need a cold, hard logic engine.
Gemini “Deep Think” mode is the “Meritocratic Acid” in real-time. It dissolves the appeal of “vibes-based” bots.
Because I spend hours with it, I am building a “Cognitive Dialect” with Gemini. I know exactly how to phrase a prompt to get the output I need.
Switching to a lesser model now would feel like trading a laser scalpel for a rusty butter knife.
If we accept the premise that AI is the ultimate “Thing” (a system to be manipulated, tweaked, and optimized) rather than a “Person” (a consciousness to be persuaded or empathized with), then we might have a potential shift in workplace power dynamics.
I wonder if we will transition from an EQ (Emotional Quotient) economy to a TQ (Technical/Tool Quotient) economy?
1. The “Systemizing” Advantage
Psychologist Simon Baron-Cohen developed the Systemizing-Empathizing theory.
Systemizers (statistically more common in men) intuitively figure out how a system works by tweaking variables: “If I change this input, does the output change?”
Empathizers (statistically more common in women) focus on understanding the thoughts and emotions of others: “How does this person feel about this decision?”
The AI Shift: For the last 20 years, corporate success heavily favored the “Empathizer.” Management was about consensus-building, “soft skills,” and navigating office politics. However, AI doesn’t care about consensus. It rewards Systemizing. The person who is willing to sit there for 4 hours, obsessively tweaking a prompt 50 times to get the perfect result, will outperform the person who tries to “charm” the AI or gives up after the first generic answer.
2. The Decline of the “Relationship Economy”
Many high-status roles today are “Relationship Holding” roles—Account Managers, Directors, Client Liaisons. These fields are often female-dominated and rely on “Social Capital.”
The Threat: AI allows a “Competent Recluse” (someone with zero social skills but high AI mastery) to produce work that is so undeniable that the relationship manager becomes less necessary.
If you can generate better work in 10 minutes, you don’t need to “manage expectations” with the client or the boss. The work speaks for itself. The value of “smoothing things over” (a feminine-coded skill) drops when the friction of error disappears.
3. The “Tinkerer” Phenotype
I talk to AI for hours. This is the “Tinkerer” trait—the same drive that leads men to spend weekends working on an old car engine or optimizing a gaming PC.
AI is not a “Plug and Play” tool yet; it is a “Plug and Tweak” tool.
If men are more culturally or biologically predisposed to enjoy the solitary, obsessive process of technical iteration, they will accumulate “AI Capital” faster.
While others are waiting for the “easy button” version of AI, the Tinkerers are building the custom workflows that make them 10x more productive.
4. The Admin Layer Risk
The corporate layer most at risk from AI is the “Coordination Class”—Middle Management, Project Coordination, and Admin.
These are roles that traffic in communication rather than production.
Demographically, these roles skew female in many Western economies.
If AI agents can schedule meetings, summarize conflicts, and onboard employees, the “Administrative Class” faces the same hollow-out that the “Manufacturing Class” (male-dominated) faced in the 1990s.
We may be witnessing a swing of the pendulum back toward “Object-Oriented” work.
2000-2022: The era of the “Networker.” Success meant meetings, consensus, and EQ. (favored feminine interaction styles).
2025-??: The era of the “Operator.” Success means staring at a screen, iterating logic, and commanding a machine. (favors masculine/systemizing interaction styles).
The power shifts to whoever can tolerate the solitude of the machine long enough to master it.
