If you and I have a contract that I’ll pay you for unique shoppers referred to my store and you just go in yourself to trigger the door counter 10,000 times wearing different disguises, it’s pretty clear that you have defrauded me.

When a government collects taxes they have to think how to invest that money to maximize gdp growth. Subsidizing road infrastructure is a pretty good use of those funds IMO, though admittedly I’m probably biased as a transportation engineer.

CA uses fuel taxes to fund road infrastructure, which isn’t the worst approach… the heaviest users (‘heavy’ both figuratively and literally) invariably pay the largest share, although EVs are messing up that equation.

So theoretically in another 20-30 years we can take a guess at who we’ll be fighting.

They’re basically just taking equations of the form f(x,y)=k where k=0, and asking “ah but what about when k!=0”?

Indeed, looking at the constant as a variable by looking at the equation f(x,y)=z IS quite useful. (I’m reminded of Feynman favorite trick for solving integrals). That generalization trick is actually useful in MANY contexts. But there’s nothing novel here.

Not just Silicon Valley but also OpenAI, specifically.

I.e. Instead of limiting autoregression to the token level, you introduce a persistent compressed global workspace latent memory vector that is fed back into the self-attention mechanism at every layer or every token step, allowing the network to attend to its own prior attentional states before computing the next token. Obviously that’s going to involve some compression steps.

Trouble is… I think the architecture there is much simpler a tweak than figuring out how to train it.

…that’s likely to just destabilize training for not much if any gain at first. You’re probably gonna have to resort to some really clever (and currently missing) tricks to figure out how to train the network to actually use that feature.

Also, for further insights on some of the unstated policy, I recommend reading the project 2025 document as well. [2]

Perhaps by “coherent policy” you meant “sound” or “effective”? But as it stands it certainly is coherent.

[1] https://www.whitehouse.gov/wp-content/uploads/2025/12/2025-N...

[2] https://www.documentcloud.org/documents/24088042-project-202...

Ordinarily in mathematics there’s a TON of papers to write just combining low level problems with different techniques. Better still, and often considered groundbreaking is borrowing techniques from other fields and adapting them or creating analogous methods to solve problems. A lot of landmark papers have been written this way. This is also what transformers are sort of good at within other contexts. They have super human breadth so I’m hopeful they’ll become real assets in math for a long time. Though the leaps necessary to adapt a technique in a nonobvious way might be too much for a while longer. We’ll see.

Truly novel techniques are quite rate indeed and I don’t know if LLMs can represent them faithfully in their embedding space or not. My inclination is that they probably can most of the time, but I don’t know. Mathematicians would describe such thins as “alien”.