gian

weaponized incompetence

[gian @notgodcomplex96]

1) what

impressive.

[Richard Davey @photonstorm]

I asked Fable 5.2 to "go wild" making a game, and all I got was this Pong clone.

And then I played it ...

👇

[gian @notgodcomplex96]

yes. google finally did something good

[martin_casado @martin_casado]

Ok. Carbon is *really* good at code.

[gian @notgodcomplex96]

@nachonavs Its his first season with pre season and full season debut, so hes by definition, a rookie.

[Max Rovensky @MaxRovensky]

he's certainly a better one than you

[gian @notgodcomplex96]

@itsjustcornbro Also agree. My stuff is pure scripts lol

[gian @notgodcomplex96]

github actions are a scam. waste of time ans money

ci is fine and necessary. just not there

[Josh Valdez @jmvldz]

great, so my GitHub Actions spend has been useless

traditional CI with agents just does not make sense

[gian @notgodcomplex96]

@itsjustcornbro pyramid and pray.

[gian @notgodcomplex96]

brilliant, passionate and humble guy is how i would describe this guy

I have close to zero doubts that he will succeed sooner or later.

I recommend to check his stuff.

My 2 cents.

[Czar102 @Czar102]

A few months go I did an experiment with the @touch_market team.

I traded for ~5 min on Touchmarket, then asked everyone the question:

What notional did I just trade?

I made about 15 predictions, each 20-120s in duration. Mostly $50 stake.

Then I told them.

The notional I traded was in the ballpark of $2m.

No other trading platform known to mankind can provide this kind of capital efficiency and upside for a trade, especially for short timeframes.

All this happens while being immune to liquiations/stop losses. Even if the price goes far against the trader, the bet remains on until expiry.

[gian @notgodcomplex96]

1. Open claude code
2. Close it

[Lennox Saint @lennox_saint]

ok why the fuck aren’t more people talking about this??

1. open Claude Code
2. type /model and pick Fable 5.2
3. set your sub-agents to Opus 5.5 (otherwise they copy Fable)
4. give it a big job and add this:

[gian @notgodcomplex96]

this is getting scary!!

Fable 5.2 now...

> proceeds to showcase a video of a website mockup, some random useless orchestration thing / skill package and a 10sec three js demo

[gian @notgodcomplex96]

@AlphaFlying2487 youre a senna hater. U dint understand shit about f1. opinion rejected

[gian @notgodcomplex96]

@BaneOfWallSt Both gasly and colapinto would out qualify antonelli in the same car.

btw the nice kid fia implant should focus on staying inside the track. luckily for him rules dont apply.

[gian @notgodcomplex96]

ppl starting to realize

hes the most talented rookie, including antonelli

but the car its all that matters to make a difference.

[manex☁️ @manex_4]

ok colapinto is mad mad talented 😮‍💨

[gian @notgodcomplex96]

zig wouldve overtaken rust if their community wasnt anti ai, sadly they will be left behind.

[quakewang @quakewang]

虽然我写了快 9 年的 Rust 了, 但是在 AI 时代我反而强烈推荐性能和内存占用在同一水平线的 Zig

以前大家吹 Rust 吹得最狠的就是它那套严苛的内存和借用检查，觉得能帮人类程序员兜底，但到了现在这个已经不需要人类写代码的时代，这反而成了拖后腿的累赘

AI 生成代码追求一个快速迭代和秒级试错，特别是在 0.17 版本之后, 增量编译几乎都是写完1秒钟后就能编译好跑测试看结果

对比 Rust, AI 为了战胜编译器往往得反复读编译错误甚至强行绕弯路，各自 struct trait 生命周期满天飞, 增量编译一次要十多秒, 改动一次硬生生被拖慢几十倍

再加上 Zig 内存分配全靠显式传参,边界极其清晰,还能直接无缝调现成的 C 库, 在 AI 已经能帮我们把关基础逻辑, 拼的是迭代效率和底层掌控力情况下, Zig 这种轻量透明不需要 AI 和编译器较劲的语言设计反而有巨大的优势

[gian @notgodcomplex96]

i moved the process out of the context. agent asks for the next step (some gates each one has diff skills etc) and only gets what that step needs, hard rules are checks at PR time. nothing to get stuck in, and it "runs free in between"

the process scales with risk. small stuff goes straight through, the full loop only kicks in on paths that can actually hurt (auth, money, state). and most checks are plain scripts, not more model turns

[gian @notgodcomplex96]

@moreproteinbars Gm Tina, hiw have u been? Long time no see you!!

[gian @notgodcomplex96]

larpin on hinge as Nvidia employee

wish me luck

[Pumkin @Pumpkiwami]

“chatgpt made it, not ai”

[vas @vasuman]

It’s not just NVIDIA. Sources confirm Google's Gemini team was also leaking confidential information to China.

China is reportedly asking them to stop.