Шел, шел и упал 🇷🇺🐍

💟rust💟

[Шел, шел и упал 🇷🇺🐍 @minmax777]

@rasulseidagul @OlexGameDev >webslop
Yes, and that's okay.

(We also use a WebView for our launcher)

[Шел, шел и упал 🇷🇺🐍 @minmax777]

@WektorAlex @Avedis72 А ты сугробы в идеальной форме держал?

[Brian Armstrong @brian_armstrong]

We were able to fine tune an open weight model (Qwen3.5-9B) on proprietary internal data at @coinbase, and it out performed frontier models, at much lower cost and speed, for a specialized task (fraud prevention in this case).

I think we'll see more of this pattern inside companies. The frontier models are amazing generalists. But increasingly you'll see specialist models out perform on more narrow tasks, either working on teams, or by themselves.

We're making it easy to fine tune models internally for our employees. We even have a skill for it in our internal harness that makes it easy to spin these up and start training them:

"/fine-tune-and-serve-llm I want to fine-tune Qwen for payments risk review. help me get started"

I imagine we'll have some backend services that could be replaced entirely by this new architecture.

[Шел, шел и упал 🇷🇺🐍 @minmax777]

@JMNewsNetwork_ fascinating behavior

[maniac (malfoid enjoyoor) @maniacdotsol]

i was a functioning human being before i discovered Malfoid.

[Шел, шел и упал 🇷🇺🐍 @minmax777]

@vital_pavlenko Профессия всегда менялась, быть "прогером" и означало учиться новому.

Все что от тебя требуется - просто код дальше и все, как и всегда.

[Шел, шел и упал 🇷🇺🐍 @minmax777]

@Capicua25x @jun_song Sir, you've already done it. You can relax now.

[Шел, шел и упал 🇷🇺🐍 @minmax777]

@TheProgGuy95 @investwithpaul Вот это поворот, да, теперь понятнее.

[иду в прайм @intoprime]

Быть праваком - ахуенно!

[Шел, шел и упал 🇷🇺🐍 @minmax777]

@johncodes You should try submitting a merge request to GIMP or LibreOffice. Then you'll understand. Some open-source projects just aren't welcoming to new contributors. I completely understand his decision.

[Шел, шел и упал 🇷🇺🐍 @minmax777]

@karo_aslan55037 @teortaxesTex Ты же понимаю, что ресурс A100 не восполним? Ведь да?
Или для тебя все цоды только и делают что обслуживают я-такси? Какой дегенерат.

[Шел, шел и упал 🇷🇺🐍 @minmax777]

@akulyan12 @2edPanda @hagalazi @It_is_not_dying Эм, еду за тебя тоже кто-то пережевать должен?

[Шел, шел и упал 🇷🇺🐍 @minmax777]

@chnyxiownf60408 @poteto @ConnorTalksAI This is obvious to everyone except those two.

[Шел, шел и упал 🇷🇺🐍 @minmax777]

@RayFernando1337 This is the way

[Eric S. Raymond @esrtweet]

It's true, there is a bit of a monastic atmosphere about a lot of math departments. There was at University of Pennsylvania when I was a theoretical
-math major back in the 1970s.

I've been pretty brutal towards the faction of mathematicians that wants to reject OpenAI's findings because they "didn't ask for this work to be done". I won't apologize for slamming their sense of entitlement and their claim to own mathematics as though it were some kind of holy ground not to be desecrated by clankers.

On the other hand, I have to say that I get it. When you're doing a thing that is so taxing that it demands complete concentration of the will - monastic devotion - it's easy to believe that people who don't partake of the discipline don't have the right to play where you do.

I'm a software engineer at a level where that takes monastic devotion, too. Maybe not to the intensity that pure mathematics does, but - I too have felt some sadness and some sense of loss as the Singularity bulldozes assumptions about programming that have surrounded me for 50 years.

And then the pragmatic part of me comes around and says "Yeah? Get over yourself. There's so much more we can *do* with the new tools..."

So here I am, leaning into AI-assisted coding pretty hard. Embracing the storm. It will not defeat me. I will ride it, master it; I will become the storm.

If you are in any knowledge-centered field, it's either embrace the storm or a collapse into stagnation and irrelevance as people with more courage and more flexibility power-zoom away from you, accelerating towards an unknown future.

Software engineers mostly seem to be handling this pretty well. A minority of open-source projects are prohibiting AI assisted contribution. I am confident they will quickly become laughingstocks.

I think and hope that most mathematicians, like most software engineers, are brave enough. Willing to bet there's an age split, too, with the younger ones much less fazed.

And me? Old, but too stubborn to quit. I was a mathematician once myself, I published research. I guess my message for upset mathematicians is: don't take counsel of your fears. You can do this. Be too stubborn to give up.

[Шел, шел и упал 🇷🇺🐍 @minmax777]

@alexrkzmu @tetelman Русские? Где?

[Шел, шел и упал 🇷🇺🐍 @minmax777]

@A_Smaslom @Skipper_Marzell fable 5

[Шел, шел и упал 🇷🇺🐍 @minmax777]

@dhh In practice, this doesn’t always work well out of the box. In most cases, CLI commands will be run with "| tail -n 20", which means half of the review output may be lost.

Explicitly instruct the agent to save its findings to a file.

[Шел, шел и упал 🇷🇺🐍 @minmax777]

@n3kroo_ @LundukeJournal Как отлично что извращенная левая мразь сама отваливается из it.