Ah, yes, the Brite Bootswatch skin.

No need to mess around further.

Only if you treat the generated code (and the costs; time and money, even if infinitesimal) as non-disposable.

One can (and probably should) make the argument that if your initial/incremental belief was incorrect, one should start again with a fresh prompt or roll back the tree and prompt again, disposing of the now-incorrectly generated code.

I catch myself doing this all the time. My model (local) isn't nearly as fast as something cloud-based, so there is a real incurred cost of time that is hard to shake. Very often I'll need the model to reconsider and rewrite the plan after it's already done quite a bit of work. Rolling back the code is tough, but I should resort to it more often.

Six months ago this ceased to be the case for some (if not most), and the quality of LLM generated code has increased over time.

I'm sorry to be the bearer of bad news: human coding has not improved a lick since then.

At a high level, we're still at the stage of AI development where we're taking cues from nature.

Take the most recent qwen and deepseek models with offloadable n-grams, which function (both in name and vaguely in capability) like human memory "engrams".

Huh, TIL "engram" is not just an alternate spelling of "n-gram".

Can confirm. I've a single R9700 and have maxed out at 45 tokens/second on llama.cpp with Q4 Qwen 3.6 27B (with MTP)

The way this comment was going I was pleasantly surprised to see that the iPhone Duo adhered to the ISO standard.

From Apple's diagonal measurements [1], it looks like the screen is actually a few mm off from the closest ISO 216 sizes.

d = 7.58in = 192.532mm [1]

d^2 = w^2+h^2

Unfolded:

h = d/sqrt(3) = ~111.1584mm

w = h*sqrt(2) = ~157.2017mm

Closest ISO 216 size: A6 (105x148)

Folded:

D = 5.36in = 136.114mm [1]

H = D/sqrt(3) = ~78.5855mm

W = H*sqrt(2) = ~111.1366mm

Closest ISO 216 size: A7 (74x105)

[1] https://www.apple.com/iphone-duo/specs/

I imagine that would be to account for the bezel. The outer display is slightly less than half the size of the inner display, so to get the same aspect ratio they would have to adjust it slightly.

As a thin client to access cloud models this is an astounding waste of money.

I think you miss the point and value add. The always on computer isn’t to run the client only, it is to run all the things a user would normally. The laptop becomes a terminal for a computer that can do anything your laptop can. If you don’t have the need or desire for that, then yes, use a rpi or whatever.

> local models are close to the point where, with a sufficiently well-architected harness, you can get results that are on par with the experience you'd have with cloud inference

In my opinion, 98% of the work most devs would send to an AI can be capably achieved with a local model and a frontier-level model is overkill.

The goalpost moving feeds right into Anthropic and OpenAI's interests.

16GB VRAM could load a small quantised qwen 27B model but it would be a ways away from a frontier cloud model.

Though keep in mind not being beholden to shenanigans from said cloud companies (and interference from government entities!) is definitely worth something intangible.

> That’s less than 164USD worth of GLM5.3 tokens on the inference market.

I can cherry pick stats too.

The other day I heard mention of someone paying $200/mo for Claude Code.

At those rates my local LM setup pays for itself in a single year.