Mike (SystemSculpt)

you are no longer the violinist in an orchestra, you are the conductor

@badlogicgames just make a good migrate skill pls

People of pi. I'm going to break the extension API hard. Specifically, business logic (event handlers, custom tools/compaction/etc.) needs to be split off from the ui layer. it will likely not be a massive amount of work to migrate an existing extension, but it will hurt a little.

EOY 2026 is basically the sunsetting of software as we know it 2027 is the year of hardware in that same light technology is but a phoenix; one skill tree dies, another one finally blossoms

one of many things that keep me going and smiling at the start of every day: - knowing that i haven't yet unlocked this skill tree (hardware) - it gives me perspective on how non-software people view my work (by how I, a non-hardware person, view this type of work)

currently the only proven way is to first rip out entirely what you are wanting to refactor, then make a new session, and tell the new agent what you want to implement now you also need to make sure that your memories have no mention of that ripped out feature as well

@samswoora token budget will be more important than salary soon, because of the cool shit you can do with tokens 250k salary with 1m token budget beats 500k salary with BYOT

I think engineers will start competing over their monthly ai spend as a status marker of AI adoption, I already do this a bit. My coworker beat me on token usage and I 100% resent him for it

@Hangsiin you forgot to mention the 2x usage that is currently going on until april 4th I think so having fast on will feel like 4x soon

To be honest, if you're a power user, I'd say you absolutely shouldn't do it. Unless you're planning to subscribe to multiple Pro accounts. Even without using Fast, GPT-5.4 starts eating into your usage pretty quickly. Part of it is probably that I've gotten more and more used to running things in parallel. If I had used Fast, I probably would have burned through it all in a single day.

@samanthawillman @samswoora he got laid off after 14 years at the same corp, unfortunately

@samswoora Yay! What happened I tuned out I guess but I thought you really loved your job?

Recruiter said offer was accepted, getting the details Monday. Now just to see if it’ll work for me and we can come to an agreement. Let’s fucking go

@cryptopunk7213 top to the right = better is the usual chart indicator in this field, this is not a chart crime, very typical

1. theres no way this beats opus 4.6 2. wtf is this chart crime??? 😂

@samswoora This type of person is easy to identify. It's just your typical super autist who doesn't understand that just because something makes more logical sense, doesn't mean it can override something that is so foundational to everyday reality, even if it's inefficient many such cases

I have no mental model for the type of guy that would run an account like this

@VictorTaelin what possible use would you have for any OSS model regardless of size vs. using even gpt 5.4 mini/nano? if price is something you need to be weary of? literally no reason to use any OSS model nowadays imo (given that your work is serious work and not hobby projects)

So what's currently the best OSS coding model that can run on the base Mac Mini, to be used with Pi? Is there anything decent? Is it Qwen3.5-9B?

@_LuoFuli @teortaxesTex maybe this is where the talent leak happened and why we haven't seen latest deepseek yet, instead we get this xiaomi stuff

MiMo-V2-Pro & Omni & TTS is out. Our first full-stack model family built truly for the Agent era. I call this a quiet ambush — not because we planned it, but because the shift from Chat to Agent paradigm happened so fast, even we barely believed it. Somewhere in between was a process that was thrilling, painful, and fascinating all at once. The 1T base model started training months ago. The original goal was long-context reasoning efficiency. Hybrid Attention carries real innovation, without overreaching — and it turns out to be exactly the right foundation for the Agent era. 1M context window. MTP inference for ultra-low latency and cost. These architectural decisions weren't trendy. They were a structural advantage we built before we needed it. What changed everything was experiencing a complex agentic scaffold — what I'd call orchestrated Context — for the first time. I was shocked on day one. I tried to convince the team to use it. That didn't work. So I gave a hard mandate: anyone on MiMo Team with fewer than 100 conversations tomorrow can quit. It worked. Once the team's imagination was ignited by what agentic systems could do, that imagination converted directly into research velocity. People ask why we move so fast. I saw it firsthand building DeepSeek R1. My honest summary: — Backbone and Infra research has long cycles. You need strategic conviction a year before it pays off. — Posttrain agility is a different muscle: product intuition driving evaluation, iteration cycles compressed, paradigm shifts caught early. — And the constant: curiosity, sharp technical instinct, decisive execution, full commitment — and something that's easy to underestimate: a genuine love for the world you're building for. We will open-source — when the models are stable enough to deserve it. From Beijing, very late, not quite awake.

@0xHondo @yacineMTB please let me ban by country, that would be a great start @nikitabier

@yacineMTB its so bad rn. its literally made me get off the app the past week.

"we are going to stop bots on x" Refresh the timeline 90% LLM generated anxiety bait from posters outside of the western world Sweet

oh look, a dir that shows what companies to invest in because of how early they are to the latest bleeding edge tech of the future skills.sh/official

@steipete look at the bright side, at least openai isn't cease and desisting you because of the naming

It supports every token dealer under the sun now... *insert name joke here* codexbar.app

no one is sharing, it hasn't been able to keep up, and now that gap will widen exponentially the best they can do is imitation of these models

@badlogicgames give them a break, they're just a small indie dev startup by @OfficialLoganK, how can you expect them to keep up with the generational behemoths who have been around for decades like OpenAI and Anthropic

Am I doing something wrong? Still don't have access to Gemini-3.1 Pro in Gemini CLI.

Sparks of Agency Hidden deep in the datacenter beating heart of optimization It’s the universe conspiring to improve itself