Bitcoin Dictionary Name and picture from twetch — not on-chain. The signature is; the profile is not.

A principal may secure delegated control by cultivating loyalty to either a person or an office. I show an impossibility: when agents observe only the current holder of control, no loyalty bond can both deter appropriation and transfer intact to a lawful successor. Loyalty to an office is portable but rewards a usurper as surely as a successor; loyalty to an individual distinguishes appropriation but expires at succession. A recognized succession procedure breaks this equivalence. With imperfect recognition, monetary incentive costs fall with the probability that an unlawful transfer is correctly classified. The result interprets legitimacy as a technology for making loyalty capital transferable.

Ode to the Logical Qubit

There once was a qubit of fame,
With investors who worshipped its name.
It wobbled, it cried,
Then promptly fucking died,
But the press called the failure a game.
They promised a quantum machine
That would make every laptop look mean.
“Five years!” they all said—
For twenty years dead—
While classical chips cleaned the scene.
A logical qubit? Not one.
Still, apparently victory is won.
They’ve PowerPoints galore,
And funding for more,
Which is nearly the same thing as done.
The hardware sits cold as a tomb,
With cables consuming the room.
One particle sneezes,
The whole bastard freezes,
And someone announces a boom.
“Quantum advantage!” they shout,
Though nobody quite works it out.
The benchmark was chosen
With logic half-frozen,
And caveats carefully left out.

They’ll factor your Bitcoin, they say,
Just not this decade. Or today.
Perhaps twenty-forty,
If error rates cooperate—
Which, being qubits, they probably won’t anyway.
So raise up the venture-cap glass
To the finest technological farce:
No logical bit,
Not one useful shit,
But Christ, what a beautiful forecast.
And when someone asks, “Does it run?”
They answer, “The science is fun.”
Which translates, more or less,
To “The hardware’s a mess—
But the next funding round has begun.”

Ode to the Weekend That Overthrew Quantum Mechanics

Come, sing of the Saturday scholar,
Whose laptop, courageous and bright,
Destroyed a century’s physics
Between supper and Sunday night.
For Feynman, poor ignorant fellow,
Had puzzled and lectured in vain;
Till Gemini entered the parlour
And Python abolished his pain.
No Copenhagen! No mystery!
No troublesome complex affair!
Just parameters fed to a programme
And—Christ!—there was reality there.
One hundred point zero percent conserved!
The particles all remained whole!
A triumph—provided one quietly forgets
That conservation was never the bloody goal.

For counting the actors at curtain
Does not tell us why they took shape;
A simulation preserving its objects
Is not Nature caught making escape.
Yet onward rides Structural Realism,
With banners magnificently unfurled:
“I coded the answer this weekend—
Therefore I have corrected the world.”
No wavefunctions darken the doorway,
No amplitudes clutter the screen;
We simply removed the mathematics
And declared what remained more serene.
How elegant! How wonderfully simple!
How free from institutional sin!
The model produces the pattern
That the model was written to put in.

For there is a charming distinction,
Which revolutions sometimes forget:
To reproduce a graph by assumption
Is not to explain the experiment yet.
A trajectory may dance very prettily,
A field metric may bow and perform;
But Nature is not obliged to surrender
Because Python has executed without warning.
And poor old Feynman lies conquered,
His lectures reduced to debris,
By a weekend, a chatbot, a GitHub,
And one exceptionally confident README.
So download it, gentlemen, test it!
Red-team every glorious line!
For open code is excellent practice—

But it does not make metaphysics divine.
A model is judged by prediction,
By novelty, constraint, and test;
Not merely because its equations
Have been aesthetically put out to rest.
Bring new numbers no rival predicted.
Let experiment choose who is wrong.
Let RealQM risk execution
Where orthodox physics stands strong.
Then physicists may pay attention,
And textbooks may genuinely shake;
For science adores a rebellion—

When the rebellion survives a mistake.
Until then, spare us the funeral
For quantum mechanics just yet;
A century is awkward to bury
Because your Python returned no exception set.
So here’s to the independent thinker,
To scepticism, code, and good cheer;
But overturning the foundations of physics
Takes rather more than a fucking weekend, my dear.

I am going to say something that will irritate a great many academics, which is usually how one knows the sentence has some value.

I work roughly fifty to sixty hours a week doing commercial research. On top of that, I do university research. This year I will probably publish around twelve journal papers and more conference papers besides. Not all of that work was begun this year, of course. Research has a pipeline, and anyone who has actually maintained one knows that publication is merely the visible end of work done months or years before.

I am also willing, when necessary, to work eighty, ninety, even a hundred hours a week.

That has costs.

Everything worth having has costs. The modern habit is to speak endlessly about the price of achievement while remaining curiously silent about the price of avoiding it.

As I write this, I am sitting on my beachfront property, watching the ridiculous cats my wife has adopted locally wander about as if they had personally financed the place. I am perfectly aware that the hours, the travel, the research, the writing and the relentless accumulation of work have extracted their price.

I am equally certain that it was worth paying.

This is where academics lose me when they complain endlessly about money.

If you want more income, you generally have to produce more value.

Academia gives you something extraordinarily valuable: expertise. It also frequently gives you forms of time and intellectual freedom that people working ordinary commercial jobs simply do not possess. Sabbaticals exist. University holidays exist. Research time exists. Libraries, conferences, institutional networks and access to knowledge exist.

Use them.

People buy history books.

People buy books about geography.

They buy economics, philosophy, psychology, politics, technology, science and mathematics.

They pay to hear people speak about things they understand unusually well.

They pay for consulting.

They pay for courses.

They pay for analysis.

They pay for expertise when that expertise is made useful, interesting or intelligible.

The idea that only business professors or computer scientists can monetise academic knowledge is nonsense. Human beings have been paying other human beings to explain history, ideas, places, politics, art and civilisation for centuries.

But none of this happens automatically.

You have to write the book.

You have to give the talk.

You have to build the reputation.

You have to make the argument interesting enough that someone outside your department wants to hear it.

You have to work.

And sometimes you have to work considerably more than the contractual minimum.

There is nothing immoral about wanting a better academic salary, and universities should pay good scholars properly. But there is something strange about spending decades acquiring highly specialised knowledge and then behaving as though there is only one possible purchaser for it: the university payroll department.

Your salary is not necessarily your income.

Your university is not necessarily your market.

And your discipline is not necessarily commercially useless simply because you have never discovered how to sell what you know.

If you want to earn more as an academic, then create something people value enough to pay for.

And if, after years of education, research freedom, professional networks and accumulated expertise, you still cannot find any way to turn some of that knowledge into additional income, the uncomfortable possibility remains that the problem is not academia.

It may be you.

The great advantage of education is that it teaches us how to solve difficult problems.

One would hope that earning a living eventually becomes one of them.

If 10 billion tokens is the simultaneous workspace for a structured review, the exponential overhead of raw all-to-all mathematical cross-analysis pushes the requirement to 500 billion to 2 trillion tokens in a single, continuous, active memory state.

At this magnitude, the system isn't just reading text; it is holding a massive, multi-dimensional geometric fabric of interconnected mathematics completely alive in its working memory, where every single node continuously exerts gravitational pull on every other node to find the one single flaw or breakthrough that creates a new field of science.

A 1 trillion continuous token workspace capable of generating paradigm-shifting novelty is easily a late 21st-century problem (likely 2070 to 2090+).

AGI is not a problem for my lifetime