xendo

Around that time I built a CLI to access and manage monitoring cameras that my company is selling. After giving a demo to my leadership I strongly adviced against releasing it to public. Giving agents access to some stuff is bad for customers.

I appreciate how they waited till SpaceX IPO.

Exactly, for long running vibe coded stuff that I don't care about quality getting big and smart model is the only option. But for high quality changes where I need to have control and understand everything, where I do everything in small chunks - I can use basic model like Sonnet.

Even with examples it's still not convincing. I'm working on real products so I don't have time to waste comparing models that won't be relevant next month.

Maybe it depends on the task, but the biggest productivity gains are from boiler plate generation, and there it's as easy as "generate me the boiler plate". Even if you can learn some very specific workflows today they would be model dependent and mostly obsolete within a month or two.

I work with AI everyday, despite what many people suggest there is so little to learn. After a couple of hours you are good to go. You don't even need gstack.

Sure, but for many folks the distraction is irresistible. It was difficult already to put care and craft into a product, having a slot machine for your attention makes it damn impossible.

In the old days, producing all those things would be tremendous learning opportunity. Today it's a pure waste, not producing income is not a problem, not producing anything is.

AI make easy work even easier, at the same time it shortens the attention span making it more difficult to do any difficult work. That's why there is so little real progress despite huge productivity gains.

Any idea what makes for such a diff between your and theirs numbers? Batching? Or could they do a crazy prefix caching across all nodes to reduce the actual processing.

You can argue that any UI is like a game engine in that sense. Some make sensible choices and don't need to pretend they have to render at 60fps.

Highly recommend gron. https://github.com/tomnomnom/gron

I thought you needed 100k lines of code for a C compiler. https://www.anthropic.com/engineering/building-c-compiler

Highly recommend 'Proving Ground' book, it's fiction but talks exactly about this.

Seems they have already removed

RustFS is written in Rust, a memory-safe language, so it is 100% secure

From their docs.

https://github.com/rustfs/docs.rustfs.com/commit/cd1ece3c5f5...

Thanks, my first post ever. After almost 20 years in tech I finally found something I feel strong enough about to share my voice.

I'm really worried about all the people this transformation will leave behind. Judging by my big tech co, most leaders don't care.

We can't really property test most code. So it comes down, as with everything, to good judgement and experience.

You really think that Rohit was driving Alexa or AGI in the right direction?

Prediction for 2026 - investors will be shitting bricks.

Are there any viable nginx alternatives that support HTTP3 and are mature for prod workflows?

Any idea if the same dataset can be used to improve human reasoning? Let's say I manually analyze 817 math examples, would that be optimal strategy for me to improve my math reasoning? Can the same distilation process be applied to leetcode?

Sharing your IAM credentials is like sharing your password. Just don't do it, regardless of the intentions. Even if this one doesn't steal anything it creates a precedence that will let people think it's ok and make them easier targets in the future. Besides, bedrock already has a console, so what's the point of using your UI?

They also announced speech to speech and any to any models for early next year. I think you are underestimating the effort required to release 5 competitive models at the same time.

Some independent latency and quality evaluations already available at https://artificialanalysis.ai/ Looks to be cheap and fast.

SteamOS on Steam Deck is based on Arch Linux.

They are using JDK17, which segments the cache into profiled and non-profiled code. Both at around 120MB by default, it’s not that difficult to hit.

Are there any officially supported Google products?

I’ve been using ‚gron’ with good results

They just know first hand that online advertising is not worth the money!

No the OP, but we had similar issue — our service was leaking when allocating native memory using JNI. We onboarded Jemalloc as it has better debugging capabilities, but the leak dissapeared and performance improved. We never got around to root causing original leak.