Sabitlenmiş Tweet
On my way to vibe code up a b-rep kernel for vibe coded CAD software.
Surely nothing will go wrong.
English
Ryan Drazga
1.1K posts
I think there are levels to the anti-AI in games crowd.
Some people are just 100% no AI at all,
Some people are no AI in large studios
Most people I imagine are no AI "assets", as in models, sound, music, ect.
Then there are no AI in gameplay design crowd.
A gamer especially a singleplayer story-driven game players want deep well thought out stories with attention to detail. One that is opinionated crafted by a storyteller with a vision. (See bioshock, mass effect, ect.)
AI can speed up the development for the systems and assets that support this; However, it really can't and shouldn't replace that vision and the crafting of the story.
This is typically the opposite of what large game companies care about though, as many of these companies now want to make "safe" games faster which are algorithmically confirmed to be profitable (current trend is extraction shooters)
So those companies replacing staff with AI to speed up the development of slop due to their lack of willingness to experiment is what players are truly worried about I would think.
English
To be fair it’s 60B for their data,
Tesla is a physical product which requires resources to build,
YouTube has a lot of the decades of existing content,
Apple again is physical and has brand name,
Cursor has very little moat and rapidly increasing competition with their main value being the data they have collected thus far not the IDE itself being worth 60B
English
People hate on Cursor, or even go as far to laugh at it.
"$60B for an IDE LOL"
"It's just a VSCode fork"
Yeah, and Tesla is just a car.
Youtube is just a video player.
and the iPhone is just a phone
It's honestly hysterical how wrong they are.
GPT 5.5 High Fast and Cursor 2.5 Fast feel unbelievably good in this harness. I have been building non-stop for the last week.
Between Cursor and my Codex sub, I want for nothing.
English
@0xglitchbyte The ironic part is AI (especially the GPT models) are surprisingly good at Zig as well.
English
unironically why zig > rust
ai is supposed to write the boring stuff, and writing code was never boring
Loris Cro ⚡@croloris
I feel compelled to give space also to disagreeing opinions. If this (totally based on real data) graphic makes you lose faith in the future of Zig, I strongly encourage you to do what you think is best for you and stop using Zig, no hard feelings :^)
English
I would argue learning those skill sets can be rapidly accelerated with AI, even if using it to the absolute minimum for a domain.
The ability to rapidly find knowledge you didn’t have before with it that takes a lot of research is a huge usecase especially for less experienced guys in a domain.
English
@JoshDaws Agreed, although importantly what this means is that non-AI skillsets will not go away, and most critically, humans will not be removed from the loop. This alone invalidates much of the marketing hype used to sell investors on specific companies.
English
It’s incorrect to assume work made with AI tools will automatically lack craft and coherent design. There’s still a human directing the tools. They can accept slop or iterate until their creative vision is brought to life.
Ryan Fleury@rfleury
When I enjoy a piece of art, or a film, or a book, or a song, or a game, it’s extremely important to me that it is a carefully crafted, coherent design, made by a human with whom I can emotionally relate, and with whom I share the common elements of the human condition. If any of these things are purely AI generated—yes, even if superficially you cannot immediately tell—it immediately sours the experience, because I feel manipulated, not communicated with. There is no relatable higher purpose behind a scene in a painting, there is no truth in any lyrics in a song, there is no carefully designed experience intended for me by playing a game—it becomes equivalent to a slot machine, and I become the drunk zombie boomer in Vegas repeatedly pulling the lever. I don’t think this is an uncommon feeling at all, and because the economy is the cumulative effects downstream of human desire, I think there is far less demand for such things than—for instance—psychopaths in Silicon Valley would like investors to believe.
English
As a young guy who has vague ideas of what I want, I use “vibe coding” as a starting point just to even see what ideas work and feel like,
The issue is to go beyond that you need experience on how to turn that into the non slop you talk about.
Which is hard to get as the market is rapidly pushing towards a desire for devs who just use AI for everything, to get rapid work done.
So there is a lot of pressure to show off your AI coded slop projects just to get work, while there is decreasing incentive from companies for you to understand that work deeply to learn how to get to the non-slop version.
English
“Vibecoding”, i.e. ~hands-off usage of LLMs to rapidly generate code without regard for the actual code’s contents, for novel applications, can literally never be non-slop, because—as I’ve described before—there is not enough bits of information content in prompts to express the user’s exact desires in sufficient detail, and the desired solution is not expressed in training data (due to the problem’s novelty).
Only a sentient human developer can relate to another human user to determine what is desirable, and design the software such that it accomplishes this desirable outcome, and carefully verify that it is doing that, rather than something else (potentially undesirable).
This is true even for the combinatoric space implied by the training data, for instance if the novel problem is merely novel in that it combines pieces of existing solutions. There needs to be a guiding force to know what to combine and how.
The more detailed the prompt becomes, the more human oversight (the more human-guided round trips with the LLM), the closer it becomes to actual code (i.e. detailed execution instructions for a computer).
Ryan Fleury@rfleury
@yacineMTB Contradiction of terms
English
@LLMJunky Meanwhile: Skyrim has been updated to run on your smart blender
English
OpenAI is not a serious company
Roomba?
Peloton?
Ninja Crock Pot?
Nintendo Wii?
Meta Glasses?
Nest?
You tryna tell me NONE of these are on the roadmap?
smh
Andrew Ambrosino@ajambrosino
circling back on what we aligned on here for Q2/Q3 prioritization
English
@Fish__Stixx @mikusingularity i would love to see paradox tackle a game like this, just stellaris but a space race sim
English
I tried out the Solar Expanse demo since it finally came out this month.
It is a space colonization management game for the whole Solar System that takes Delta-v and interplanetary transfer windows into account.
English
English
we are in the "way too many pylons" era of image gen
Air Katakana@airkatakana
this is the new openai image model two forges, map isnt fighting spirit, way too many pylons, wall makes no sense, cannon in main, 1 building that doesnt exist, dragoon but no cybernetics core, no gas at all
English
Ryan Drazga retweetledi
I'm writing Go again (for what, you'll see later...). `go doc` and `gopls` are like agent superpowers and its shocking how productive agents are out of the box at writing [good] Go code versus other languages I've used (including the JS ecosystem). Also, Go + Zig is a good mix.
Go for the higher level and concurrent stuff and then no-libc Zig code plus the Zig compiler for zero dependency cross-compiled cgo with high-performance characteristics (minimize cgo boundary crosses). Chefs kiss.
Its funny because a lot of the shitty ergonomics of Go CLIs like `go doc` and `gopls` (prev. stuff like `go oracle` or `guru`) are totally obviated by agents and not just that but in a twist of irony they're excellent for agents.
Don't worry, its not Ghostty. Ghostty and libghostty will remain pure Zig; it's a fantastic fit and a perfect pairing. This is for something else.
"Wait, I thought you said Go has no place anymore?" I was wrong, mostly because agents are so productive at Go. I won't bring in other languages in this discussion because I don't want to feed the crabs, so to speak. lol.
English
Ryan Drazga retweetledi
Meet Kimi K2.6: Advancing Open-Source Coding
🔹Open-source SOTA on HLE w/ tools (54.0), SWE-Bench Pro (58.6), SWE-bench Multilingual (76.7), BrowseComp (83.2), Toolathlon (50.0), Charxiv w/ python(86.7), Math Vision w/ python (93.2)
What's new:
🔹Long-horizon coding - 4,000+ tool calls, over 12 hours of continuous execution, with generalization across languages (Rust, Go, Python) and tasks (frontend, devops, perf optimization).
🔹Motion-rich frontend - Videos in hero sections, WebGL shaders, GSAP + Framer Motion, Three.js 3D.
🔹Agent Swarms, elevated - 300 parallel sub-agents × 4,000 steps per run (up from K2.5's 100 / 1,500). One prompt, 100+ files.
🔹Proactive Agents - K2.6 model powers OpenClaw, Hermes Agent, etc for 24/7 autonomous ops.
🔹Claw Groups (research preview) - bring your own agents, command your friends', bots & humans in the loop.
-
K2.6 is now live on kimi.com in chat mode and agent mode.
For production-grade coding, pair K2.6 with Kimi Code: kimi.com/code
-
🔗 API: platform.moonshot.ai
🔗 Tech blog: kimi.com/blog/kimi-k2-6
🔗 Weights & code: huggingface.co/moonshotai/Kim…
English
why this get so many vues? is zig popular now? how do I feel special and superior now???
maybe odin?
Shreyas Mididoddi@Shreyassanthu77
anybody want a zig s3 client that's tiny, 0 deps and fully uses the native 0.16 async IO and streams?
English
@0ximjosh @sun_apanasovich Real shirt cheaper then the virtual item it’s based off of…
English
@SebAaltonen @jamonholmgren It’s interesting to watch the development of Kitten Space Agency where they are putting a lot of work into a custom engine to deal with these issues.
English
In a custom engine, I could personally choose 32-bit fixed point. If we have 1048km * 1048km world size (height = 1048km too). We have uniform density of 4096 units in one meter (~0.2mm). Same storage and CPU cost as 32-bit floating point. But you need to make positions (3xuint32) and vectors (3xfloat32) separate types. It has engine-wide implications.
English
Devs are … fun.
Every reply is either:
1. You should do origin shifting (the thing I literally say I’m doing in the second paragraph)
2. Wow what a dummy just use 64 bit floats / integers (sigh…)
3. Wow that’s actually hilarious, leave it in (best replies)
Jamon@jamonholmgren
Floating point precision gets pretty bad at 500 kilometers from the origin point 😅 Working on an origin shifting system, but it's slow going, given my terrain system doesn't support it and I'm having to do a bunch of C++ wrangling, compiling, reloading ... not the fastest feedback cycle
English
Ryan Drazga retweetledi
We just released Gemma 4 — our most intelligent open models to date.
Built from the same world-class research as Gemini 3, Gemma 4 brings breakthrough intelligence directly to your own hardware for advanced reasoning and agentic workflows.
Released under a commercially permissive Apache 2.0 license so anyone can build powerful AI tools. 🧵↓
English
@duanebester Did you increase the value of that software, or decrease costs by more then 1200 (IE if you would have needed 2 engineers vs one for that timeframe)
If you can measure that, which may not always be measurable in numerical values, this question can be answered.
English
Say I spend $1200/mo on AI tokens, but I’m twice as effective at software. Does my value double or halve (with a bias of +/- $1200)?
English
@jeffreyweichsel This sounds like a solid foundational role to hire for
English
WANTED: PSYCHOHISTORIANS
Seeking exceptional candidates with advanced training in stochastic processes, sociology of masses, and predictive modeling of large-scale human behavior.
Candidates must demonstrate:
• Mastery of n-variable differential equations governing societal dynamics
• Ability to extrapolate historical trends across millennia with statistical rigor
• Willingness to operate under conditions of utmost discretion
Successful applicants will join a small, elite team developing a revolutionary framework for forecasting the future of the Galactic Empire. Work involves collaboration with Dr. Hari Seldon and select colleagues. Compensation: intellectual fulfillment + secure accommodations on Trantor (or relocation to designated peripheral projects).
This is not a position for those seeking public acclaim or imperial favor. Only those who understand that the fate of trillions may rest on precise probabilistic analysis need apply.
Reply via encrypted channel: Project Prime Radiant – Reference “Seldon Initiative.”
Vague inquiries will be ignored. Talent will be noticed.
English