Excellent overview of what’s happening and what’s coming. Well, until the next breakthrough. I suspect the skills entrepreneurs have developed will suit this new world.
⚡️This is the moment AI stops being a productivity tool and becomes a replacement architecture for elite cognitive labor.
Citadel is not a random corporation automating low-level admin. It is one of the most competitive intelligence machines in finance. The work Griffin is describing sits near the top of the white-collar pyramid: research, modeling, financial reasoning, market analysis, scenario work, probably pieces of strategy design and investment process. If that work is moving from “PhDs over months” to “agents over days,” then the protected class is no longer protected by intelligence alone.
That is the earthquake.
For years, the comforting story was that AI would eat repetitive white-collar work while elite judgment stayed safely human. That story is breaking. The machine is now moving into work that looked elite because it required credentials, stamina, math, domain knowledge, and long-form synthesis. A lot of that work turns out to be decomposable into agentic loops: gather data, structure problem, run model, test variants, summarize findings, compare assumptions, stress scenarios, refine output, escalate uncertainty.
That does not eliminate the human at the top. It makes the top human massively more leveraged. The portfolio manager, senior analyst, or strategist who can frame the right question and judge the output becomes more powerful. But the pyramid underneath them gets thinner. The machine does the grind. The human becomes conductor, evaluator, risk owner, and taste layer.
That breaks the apprenticeship model.
The junior analyst’s old job was not just to produce work. It was to become someone through the work. The grind built pattern recognition. The model-building built intuition. The memo-writing built synthesis. The repetition built judgment. If agents now do the repetition, firms get efficiency today while quietly destroying the training pipeline that produced senior judgment tomorrow.
That is the social bomb inside this.
Finance can automate the ladder faster than it can rebuild the ladder. Law, consulting, software, accounting, corporate finance, marketing, research, medicine, and education all face the same problem. The entry-level layer was always partly inefficient, but it was also how humans absorbed tacit knowledge. AI attacks the inefficiency and accidentally attacks the formation process.
The winners become extremely powerful.
One elite operator with agents can produce what used to require a team. One small fund can run research breadth that used to require institutional scale. One independent analyst with the right workflow can compete far above their formal weight class. That is the opening.
But inside big institutions, the same force compresses headcount. Fewer juniors. Fewer middle managers. Fewer generic analysts. More pressure on everyone to prove actual judgment. The credential stops being enough. The market asks a colder question: can this person command the machine toward truth better than someone else?
That is the new meritocracy.
The most valuable skill becomes agentic command: knowing what to ask, how to decompose a problem, which outputs are fake, where the hidden assumption lives, when the model is overconfident, what data matters, what contradiction breaks the thesis, and when the machine has produced coherence without truth.
Griffin feeling depressed is the tell. He saw the labor impact inside the walls before the public narrative caught up. This is not about a chatbot writing emails. This is about high-end cognitive production being mechanized.
AI is revealing that a shocking amount of elite knowledge work was structured pattern labor protected by credential scarcity.
Once agents can perform that pattern labor, the real scarce asset becomes judgment under uncertainty.
Everyone else gets repriced.
@EvanLSolomon@kbsdigital Imagine if government-funded research generated ongoing licensing revenue for the public — not as tax revenue, but as contributions into a sovereign fund owned by the people.
Adam Chambers on Buffalo fans singing O Canada: "I think that is a strong representation of the relationship between our two countries ... it would be nice if politicians on either side of the border got to the table to instead of using the relationship for their own domestic political benefits."
My notifications are absolutely flooded since announcing that you can @ me on X.
If you haven't tried it yet, what are you doing?
If you're:
- Raising capital
- Hiring for your startup
- Going to a city and want to meet professional connections
Or even just see someone who is looking for these things.
Tag me in the conversation or DM me!
At the end of the week, the follower who tags me into the most interesting thread gets a Boardy affiliate badge.
This just makes sense on so many levels. Lower environmental impact, quiet clean energy and surplus revenue generated for the community. There are so many more rivers up north. @timhodgsonmt
🇨🇦 BREAKING: Canada just a broke a world record for the largest gathering of people dressed as dinosaurs.
We're not kidding.
682 people turned up in costume to commemorate @UCalgary's 60th anniversary.
We just posted all the pics.
Oh, Canada 🦖
https://t.co/53tCarAmU0
Canadians don’t hate Americans. We were certainly shocked at the threats of annexation. Just imagine if Americans were threatened. But as their biggest customer, being told they don’t need us just makes us ask…who else in the world needs/appreciates our money, products and services? No drama , just pragmatic action. And we seem to have a Prime Minister that agrees.
Honestly, this is the most accurate diagram I've seen.
Waterfall: You plan for 18 months and deliver exactly what nobody needs anymore.
Agile: You deliver something usable at every step, but the CEO keeps asking, "Where's the car?"
AI: You get the car on day one. It has six wheels, the doors are on backwards, and it has a rocket launcher. You spend more time making it yours than actually "building"; it's shaping. owning. verifying. That's what the best AI developers do now. They don't build. They shape and own.
Hope for reducing GHG in the atmosphere was rekindled when the Powerwall sold out and an affordable EV called the model 3 was about to launch. If it sold out quickly like the Powerwall, there was a shot. Batteries meant “time shifting” energy. And EV’s, if they caught on, would start replacing gas distribution channels. Also, knowing batteries and EV’s were tech, (and tech innovates and reduces costs) meant there was hope. But then you need clean power generation. And here we are. To your point, it’s not belief, it’s about money. If solutions are cheaper/save money and are more convenient, people and businesses switch. Just look at the battery lawnmowers. No more lugging gas cans to the station. They are quieter, getting cheaper and batteries have gotten better. Now if we could just all get along in the world. Who knew the world would need to be reminded again about how important “love the Neighbour” is. Love your posts btw.