0xA

10.3K posts

0xA banner
0xA

0xA

@NeuralAA

Formerly @GiannisWorld got suspended at 10k Figuring things out as I go

Turn On 🛎️ เข้าร่วม Şubat 2018
651 กำลังติดตาม1.6K ผู้ติดตาม
ทวีตที่ปักหมุด
0xA
0xA@NeuralAA·
Excited to share what ive been working on.. I Built my first product and finally shipping it after a ton of work on it.. I am introducing Prism Canvas.. upload a PDF, get a visual map of every concept inside it. Each concept comes with its own breakdown: summaries, key stats, timelines, comparison tables, and live interactive simulations you can actually play with (sliders, toggles, and more.. the whole thing). You can also chat with your document and pull up the source at any point. My first real product launch would appreciate anyone giving it a look, and all feedback is appreciated Get your first month for free using code “LAUNCH” at checkout More below 🧵 👇
English
4
2
14
2.3K
0xA
0xA@NeuralAA·
@mweinbach You got 10 m3 ultras Leave the 27 billion models for peasants like us
GIF
English
0
0
0
101
Max Weinbach
Max Weinbach@mweinbach·
Damn Qwen3.8 27b is not a terrible model Still would use a larger one like MiniMax M2.7 or GLM 5 but it's good for incredibly basic stuff. The RL really favors 1 MAYBE 2 tool calls rather than deep research or work
English
6
0
47
5.1K
0xA
0xA@NeuralAA·
@thdxr Meanwhile apple don’t give a shit💀
English
0
0
0
64
dax
dax@thdxr·
you're probably underestimating how crazy things are
dax tweet media
English
231
649
8K
723.5K
0xA
0xA@NeuralAA·
@thdxr “5. cause a sev1 incident” Would be an issue if I vibe coded but I vibe engineer
English
0
0
0
485
dax
dax@thdxr·
guys we fixed the aws console 1. open cloud shell 2. npx opencode-ai 3. it already is authed with aws + will pickup bedrock models 4. ask it to do everything aws 5. cause a sev1 incident
dax tweet media
English
39
35
1.2K
75.8K
0xA
0xA@NeuralAA·
@mweinbach I remember when I had a somewhat growing account and had this issue before elon’s terrible app banned me for absolutely ZERO reason
English
0
0
1
155
Max Weinbach
Max Weinbach@mweinbach·
I'm blocking AI bots now screw this
English
10
3
144
8.6K
Hoops Crave
Hoops Crave@HoopsCrave·
Zohran Mamdani announces Eid will officially become a public holiday in New York City.
Hoops Crave tweet media
English
292
3.2K
50.5K
908.4K
Tibo
Tibo@thsottiaux·
Do people like this? We don't do this for codex because it exists to help you and it's important that you remain the owner and accountable for your work without AI taking credit. At the same time it does mean that you can't trace how popular codex is among repos.
Yuchen Jin@Yuchenj_UW

I noticed something interesting: Claude Code auto-adds itself as a co-author on every git commit. Codex doesn’t. That’s why you see Claude everywhere on GitHub, but not Codex. I wonder why OpenAI is not doing that. Feels like an obvious branding strategy OpenAI is skipping.

English
586
74
2.6K
340.7K
Reshad Rahman
Reshad Rahman@ReshadFCB·
🚨 BREAKING: Barcelona will be open to hearing proposals for Kounde this summer if a good offer arrives. @MatteMoretto #Transfers 🇫🇷💰
English
536
606
11.5K
1.2M
0xA
0xA@NeuralAA·
@frontendTony @leerob @axolytea @rnescalz Aint no backhanded compliments lmao these are great people and ive always said this and their work is impressive And I hope they know it being built in k2.5 doesn’t mean their work is for nothing like some people say
English
0
0
0
25
Lee Robinson
Lee Robinson@leerob·
Yep, Composer 2 started from an open-source base! We will do full pretraining in the future. Only ~1/4 of the compute spent on the final model came from the base, the rest is from our training. This is why evals are very different. And yes, we are following the license through our inference partner terms.
Fynn@fynnso

was messing with the OpenAI base URL in Cursor and caught this accounts/anysphere/models/kimi-k2p5-rl-0317-s515-fast so composer 2 is just Kimi K2.5 with RL at least rename the model ID

English
358
202
2.8K
1.4M
0xA
0xA@NeuralAA·
@scaling01 Not just personal problems if I need to get something done and ask gemini It has access to time and points it out with every response and tells me to go sleep.. the fuck💀
English
0
0
0
42
Lisan al Gaib
Lisan al Gaib@scaling01·
talked to Opus 4.6 for a couple of hours about personal problems and it has this weird response mode where it's very commanding "put the phone down", "close the laptop", "Save this conversation. Set the reminder. Go to sleep.", do this, do that not sure how I feel about it
Lisan al Gaib tweet media
English
536
74
4.5K
408.4K
0xA
0xA@NeuralAA·
@robinebers But why Just be good faith and say yk thats the whole thing Anyways it don’t matter that much and they did good going back on it
English
3
0
0
57
0xA
0xA@NeuralAA·
It feels like anthropic when they begin to make any product they ask “How badly can we build the V1 of this and add so much overhead with zero upside” lmao This isn’t a coding with AI issue at all btw I don’t get some of their decisions
Felix Rieseberg@felixrieseberg

A small ship I love: We made Claude.ai and our desktop apps meaningful faster this week. We moved our architecture from SSR to a static @vite_js & @tan_stack router setup that we can serve straight from workers at the edge. Time to first byte is down 65% at p75, prompts show up 50% sooner, navigation is snappier. We're not done (not even close!) but we care and we'll keep chipping away. Aiming to make Claude a little better every day.

English
0
0
1
67
0xA
0xA@NeuralAA·
@tibor_tee i spent 15 minutes looking where the hell I can get the non fast composer lmao cause it wasnt in the usual place
English
0
0
1
17
0xA
0xA@NeuralAA·
@OpenAIDevs Naaahhhhh its just not good im not gon lie to you Its almost like its hellbent on making terrible ux
English
0
0
17
1.1K
0xA
0xA@NeuralAA·
@tibor_tee Fucken hell thats beautiful
English
0
0
1
12
Tibor (Tee)
Tibor (Tee)@tibor_tee·
Sunday is gonna be so much fun riding out here again
Tibor (Tee) tweet media
English
5
0
16
337
0xA
0xA@NeuralAA·
@magicseth “Warning: If you have fewer than thousands of documents and moderate traffic, you probably don’t need any of this yet. Seriously — ship your app. These patterns are for after you’ve succeeded.”
GIF
English
0
0
1
761
Seth Raphael
Seth Raphael@magicseth·
🦞 I’m now an OpenClaw maintainer ⚡ I work at Convex So naturally I scaled ClawHub to support 1M WAU. The future is crustacean-powered and runs on the best backend in the world. Here’s how: stack.convex.dev/optimizing-ope…
English
23
25
363
163.5K
0xA
0xA@NeuralAA·
@jamesacowling @magicseth I love this Will read it another 5 times But love stuff like this keep more of this coming its really helpful 🙏 Although I have 100 users but I still want to know about it💀
English
0
0
0
109
James Cowling
James Cowling@jamesacowling·
ClawHub was vibe-coded to 1M weekly actives (plus a lot of bot attacks tbh) but started getting expensive. @magicseth optimized it to massively reduce their Convex bill and we're feeding this stuff back into our prompts to help your agents do this themselves. details below
Seth Raphael@magicseth

🦞 I’m now an OpenClaw maintainer ⚡ I work at Convex So naturally I scaled ClawHub to support 1M WAU. The future is crustacean-powered and runs on the best backend in the world. Here’s how: stack.convex.dev/optimizing-ope…

English
3
0
46
8.3K
0xA
0xA@NeuralAA·
@magicseth I absolutely love it Great read I love stuff like this Will probably read it another 5 times
English
0
0
0
262