jonaustin

``` We're not going to add support for the Pixel 11 series unless it ships MTE support. It has hardware level support for MTE but it shipped without firmware support for it. It's unknown why that's the case and it could be due to flaws in the hardware implementation which have to be worked around. It isn't yet clear if it's going to support MTE but we've decided we won't support it unless it ships. Android 17 QPR2 Beta launched for the Pixel 11 after we made our initial posts about this and added preliminary firmware support for MTE while keeping it fully disabled for the OS even as a developer option. It isn't clear if something is severely wrong with it or not yet so we can't make any commitment to supporting the Pixel 11 series. ```

Comment · 1 month ago

Maybe...at least in the web ui chats I now do several rounds of "carefully review your conclusions" because gpt is incredibly lazy and misses ridiculously obvious things (obvious with a little bit of web research) pretty much every single time it's asked to research anything.

Comment · 1 month ago

I've been working on a pi extension to do this, after frustration with getting the best current local models to stay on track, and just to deal with their relative slowness; gist is that it uses pi hooks to keep the model deterministically on track and beads-rust issue tracking tool to keep everything organized.

A SOTA model writes the initial prompt, and creates the beads issues; then a continuous iteration of plan(local) -> review (sota) -> implement(local) -> review (sota). Until the sota reviewer model is happy with the implementation.

And can mostly just let-it-run; e.g. overnight since local models with mac unified memory are slow.

Still early days, but have had reasonably good success with a Defender (1981) clone and now I'm having it work on a Prince of Persia clone (both Go/ebiten).

Note: I think this is only possible now because Qwen3.8-Flash-Next and 27b are incredibly good models.

- https://github.com/Dicklesworthstone/beads_rust

Comment · 2 months ago

Not all of us live for The Hustle...

Comment · 2 months ago

> This clearly implies that they believe ChatGPT models are AGI and are now willing to say it out loud. > Which I think is a fair interpretation of the term

You can't be serious... Case in point - I just asked GPT Sol High to give me a weekly update of local ai changes.

Here's it's first update, which is complete and utter garbage; i.e. it's a lot of words that says absolutely nothing. That's just a random word generator; AGI? Not even remotely in the ballpark.

https://chatgpt.com/share/6a761ec1-4b38-83ea-a8f7-8f02084e63...

Comment · 2 months ago

fwiw, it absolutely drives me nuts that cli programs are putting configs in that hard to reach folder (long path name and _spaces_ in the folder name ugh)

Comment · 4 months ago

I just started using it on an m4 max 128 and it's the first time since buying the machine a year ago that it feels like local llm "just works" for reasonably decent coding.

Use pi though; claude code has way too much bootstrap context; slows everything way down.