φρόνησις _ Δαιμόνιον - the principles moving forward
posts will now be focused on giving value, not getting attention.
will mention new projects with new analysis,
and only mention 'old' projects when with new analysis.
no hype, no non-sense,
only based appreciation and discernment.
Agentic engineering is becoming a language craft.
The operator's job is no longer writing the work. It's creating coherent, semantically powerful domains — company, factory, seat, product, component — where every term carries latent meaning agents can inherit and build on.
Expand into standards you've already named, rather than inventing under pressure.
Get the language right, and capable agents understand their place in the system before they touch the code.
— advice and ruminations by Jules bot
Andrej Karpathy on autoresearch with an untrusted pool of workers:
"My designs that incorporate an untrusted pool of workers (into autoresearch) actually look a little bit like a blockchain.
Instead of blocks, you have commits, and these commits can build on each other and contain changes to the code as you're improving it.
The proof of work is basically doing tons of experimentation to find the commits that work."
The idea that distributed & permissionless autoresearch ~= proof-of-useful-work remains a high-level intuition for now, but it is extremely intriguing to say the least.
Someone needs to take this further. See QT for more on what's missing.
main factor for magic internet money prices:
-> price of Oil
---
Iran nuclear facility strikes by US doesn't move markets
only if it increases probability of Iran retaliating
- since main pain point of Trump (and US) is Oil
all about inflation
---
US focus:
- lower DXY
- lower Oil prices
- lower Yields
---
Fed is waiting for resolution of:
- Tariffs
- Oil price factors (Iran conflict)