c0rruptbytes

the iphone mini unfortunately did not kill nor do most small iphones

are they releasing the weights too?

open models don’t imply local models… it just allows more people to run them, most likely with nvidia GPUs

so many inference project, omlx already supports all of this and has a 1000 people trying to optimize it constantly

m5 max really fixed pp with the better matmul support, im sure the m5 ultra will be even crazier

the sparks have much slower memory bandwidth is the trade off

the rr suite seems much better for that

OpenAI measures their internal token usage in “rolexes” - it’s literally a flex to be a token burner

i can imagine insane amount of capital is wasted on these two companies compared to the efficiency elsewhere

it's for me

The 512GB could run GLM 5.3 which is Opus level

Sol is closer to Fable than Opus - I like SlopCodeBench the most as a test - https://github.com/humanlayer/advanced-context-engineering-f...

You add requirements and make previous tests invisible to see how pigeon brained the model is - Sol and Fable seem to rank the same as Opus tends to fall behind

the writing style is so easy to fix, output styles is documented in claude code and you can change it

still don’t think anthropic models are worth the money

auto research the new cool kid on the block - look at https://mlx.fast

Pi by itself is more than capable, OMP is okay but you really don't need much for a great harness (these models are RL trained to hell to be a coding agent, sometimes less is more)

I run a lot of SlopCodeBench - https://github.com/michaelasper/benchmarks

Fable/Sol/GLM 5.3/Kimi are its league (in that order) Deepseek/Opus is solid Qwen 27B is the floor - there's no reason to use Sonnet/Terra/Haiku

For everyday activity - I don't think you need to be using Sol (xhigh) for everything - unless you're made of money - I've found using Luna from OpenAI to be more than enough - it'll outreach to Opus/Sol when it needs to

Haven't had access to Gemini 3.7 but we're getting it at work soon, will give it a go!

Codex CLI is pretty bare bones in a bad way (at least Pi is extensible). Claude code is vibeslopped to the extreme

easier to gate MCP tools? you can allow/deny tools very easily

need competent people to hire competent people first - it seems like their only directive is to raise $MSFT not make good tech

(talking about leadership, not my lovely msft engineers reading this)

plenty of companies offer ZDR and are just as random as OpenAI and Anthropic in their age

looks nice...time to turn it into a pi extension

Pi is excellent for this, its system prompt is tiny

they probably should kill the grok branding...

wow antirez does not sleep

Reminds me of sandboxy - https://github.com/apple/containerization/tree/main/examples...

Also if your thing doesn't work with `pi` out of the box, then low effort

i would call myself a tinkerer but still don’t have a necessity for 3)

it’s more akin to 3d printing to me, i get the design all setup and let the machine do 3) and then get to play with it in 4)

i think i agree, we are going to have /more/ math and now need /more/ mathematicians (we are seeing https://vibemathed.com/)

these LLMs are great are generating arguments but they don't ask questions, we will need mathematicians to shepherd them into more discoveries

i really want to see open weight models crack some breakthroughs

i think over time, the bar will rise, you’d have to imagine how low quality the submissions prior to AI would’ve been since most people aren’t that good at decorating nor have the time

i think as the novelty wears off on the average person, it’ll probably be seen akin to stick figure drawings in the future

perfect benchmark to burn more tokens - convenient

Hestan >>>>>

same, it’s great for first pass LLM docs

my experience right now with fable via bedrock is they collect data, 5.6 has the option for ZDR at least

few months behind? 5.6 came out last week and kimi is benching around the same right now