Does an expansion of computing power on this scale imply that computing capital is displacing model architecture as the true moat in AI competition?

Debugging in C is soooo hard. When I was writing Malloc Lab in system course, there were uncountable undefined and out of range :(

Yet, debugging memory corruption issues in C and C++ code with modern compiler toolchains and memory debugging tools is infinitely easier than 25 years ago.

(e.g. just compiling with address sanitizer and using static analyzers catch pretty much all of the 'trivial' memory corruption issues).

I think vice versa - C is so simply, that debugging it is just a pleasant walk.

Especially compared with modern languages with lambdas/exceptions/virtual functions and so on.

The one thing I see can make it harder is function pointers.

I was wondering that will modifying prompts or contracting the context also impact the performance? It may mistake the original meaning, and these steps also need help from external LLM.

Forge doesn't modify the prompt, it just injects information into the conversation as if it was a conversation turn. Over many turns - it can degrade the model (a concept I'm calling "effective attention"). But that requires serious context growth that really only becomes relevant for long-running agentic coding tasks in my experience. Still, it's possible.

Context compaction can also affect the outcome - I have eval scenarios for that as well but not in the published set, only in the repo. For those, I'd say "it's better than nothing". If you hit max context, the whole thing will barf or OOM the rig or something like that. So compaction degrades performance versus some theoretical ideal where you never need to, certainly. But it's better than a hard failure. Eval on those scenarios showed increasing degradation depending on severity of compaction. I view the auto-compaction as insurance. I never give the models tasks that will require that much context, but if it ends up getting there then the run might be saved.

Asking for a hundred billion in damages and having your multi-million dollar legal team defeated not on the merits of the case, but because they literally forgot to look at the calendar lol

kudos to the Martin-Baker seats and the pilots' training, ejecting at that low altitude after a collision is incredibly hazardous. Using high-value electronic warfare assets for aerobatics seems like an unnecessary risk when regular Hornets could do the exact same job for the crowd.

Always enjoy looking through the Martin-Baker site for their database of successful ejections. They've saved a lot of lives!

Will the environment is EU be particularly better than US?

The next version of MacOS should have a choice to "keep alive while the screen is closed" lol

Aren't there a few utilities for that, like caffeinate, etc?

`caffeinate` is primarily designed to prevent sleeping due to inactivity. It doesn't prevent sleeping due to lid closure while on battery.

`sudo pmset -b disablesleep 1` (0 to revert) works, although it doesn't exactly roll off the tongue.

I'm working on a local-first AI app that combines a chatbot interface with agentic capabilities (similar to a GPT + OpenClaw mashup). Would love to hear what features you'd want in a "Local AI Copilot"!