Neden bu kadar verimsiz ve sevimsiz bir süreç bu?
Kasanın imalatı keyif kaçırdı.
İnsanlığa yazık.
华强北 Mac mini 生产过程
看的我也想进去打螺丝了~
Readers added context they thought people might want to know
视频展示的是华强北迷你PC组装过程,并非苹果Mac mini。该视频也被正确描述为迷你电脑生产,华强北以生产通用/仿制电子产品闻名。
x.com/EngrUnlocked/s…
ifun.de/so-wird-der-ma…
shenzhen-trip.com/guides/huaqian…
35
*Deepseek V1.4 Flash 1017*
Cheap inference with long context support.
Give it a try.
zenmux.ai/invite/YFJ91N
43
Opus 5.5 is safe.
Fable 5.5 can take as much time as it needs.
Muse feels sooooo benchmaxxed. :(
1
71
i think there is problem with Muse Wonder Full model.
i m not feeling that much intelligence boost as in shown benchmarks.
it feels it is hard-benchmaxxxxed model to the core.
soo sad.
soooo sad...
1
1
76
Distractosphere retweeted
Try Underdog: free, capable, private AI running locally on your device
New build and Saluki 27B are out. Under 8GB, beats Qwen 3.8 27B at tool calling
Like, repost, and comment one task you wouldn’t give a cloud agent to get an invite🦴
Today we're releasing Underdog Saluki 27B
Based on Qwen 3.8 27B, Saluki is almost 7x smaller than its full-precision counterpart while it keeps 96% of its benchmark performance and beats Qwen at tool calling
It's the best model <8GB that runs in Underdog
Saluki is available today under Apache 2.0. Try today in Underdog underdog.ai/saluki
this is so interesting that nobody talking about this model set: desertant.com/ @desertantlabs
xitter.gq/desertantlabs/status/2…
Today we're launching Desert Ant Labs: a European frontier AI lab building on-device intelligence.
18 models across audio, vision, and text. SDKs for Swift, Kotlin, and JavaScript.
No tokens. No logins. Nothing leaves the device.
desertant.com
46
I ditched Ghostty for Tern, and I’m extremely happy with the switch so far.
So what’s actually different between them?
At this point, Tern feels like it could become the foundation of a new terminal paradigm.
Ghostty, on the other hand, feels like the pinnacle of the traditional terminal world.
And I mean that as a compliment to Ghostty. At least on macOS, it’s probably one of the best implementations of the classic terminal model.
Tern feels different. It feels like it’s trying to rethink the terminal itself rather than simply build a better terminal emulator.
And there’s another important difference: Tern also runs on Windows.
Huge thanks to @can1357 for building this terminal.
I’m seriously curious to see what OMP 2 is going to be like.
1
77
OpenAI is about to flip the pricing model.
GPT-5.6 and GPT-6 become Sol, Terra, and Luna unlimited.
Astra works like Anthropic Fable: one pool, no more 5-hour caps.
$200 → ~$400 Astra
$500 → ~$1,000 Astra + max-speed, possibly unlimited Sol / Terra / Luna
"human collaboration is a superintelligence technology"
- Ilya Sutskever | September 30, 2022
xitter.gq/SmokeAwayyy/status/209…
76
started using astra with %88 usage limit with light effort on big project assessment. lets see...
132
go check until it says yes
codex --model gpt-6-astra exec "have i got gpt-6-astra yet?" --skip-git-repo-check
warning: Model metadata for `gpt-6-astra` not found. Defaulting to fallback metadata; this can degrade performance and cause issues.
ERROR: {"type":"error","status":400,"error":{"type":"invalid_request_error","message":"The 'gpt-6-astra' model is not supported when using Codex with a ChatGPT account."}}
ERROR: {"type":"error","status":400,"error":{"type":"invalid_request_error","message":"The 'gpt-6-astra' model is not supported when using Codex with a ChatGPT account."}}


















