we're launching something exciting today (in beta): @NxDevTools Cloud as a full CI platform, no more GitHub actions 🔥
This unlocks so much for us, the numbers kind of seem too good to be true but they aren't (check out the open source benchmarks)
https://t.co/AaZhy2HSFR
@MaxKless@NxDevTools the numbers 😲. Feels almost wrong share the benchmark on GH, while also saying we don't need them anymore. https://t.co/PjGGNqrOwg
A new better, faster CI pops up every day...kinda the new "hot" thing recently.
Solving the CI bottleneck is tough, but I think we've got some smart strategies there that help push the needle into the right direction.
I wrote it up in a tech deep dive: https://t.co/AGULyb0yjQ
If you want your CI ready for agents pushing PRs non-stop...
...you need something smarter than just faster machines. Here's why other CI solutions can't compete with Nx Cloud CI:
https://t.co/z4OdAdvLzc
We released the private beta for Nx Cloud CI today!
At a time new CI solutions are popping up daily promising it'll magically be cheaper and faster (most btw can't show their receipts) - ours has years of pre-vibe coded development to make your machines smarter. 🧠
A new better, faster CI pops up every day...kinda the new "hot" thing recently.
Solving the CI bottleneck is tough, but I think we've got some smart strategies there that help push the needle into the right direction.
I wrote it up in a tech deep dive: https://t.co/AGULyb0yjQ
Nx Cloud now has it's own CI!⚡️We removed the middleman (sorry, GitHub Actions) so everyone can get Nx level optimization, without needing to pay for us and their CI provider.
You get faster and cheaper CI, you get faster and cheaper CI - everyone gets faster and cheaper CI!
Nx Cloud CI Beta is out!
10x faster, 5x less compute.
In this video, I explain how it works and how we got these results.
Learn more: https://t.co/FQQIjNOn4e
@NxDevTools#monorepos
Monorepo tooling enable fast and efficient CI. It doesn't always feel this way, but at that scale, CI simply wouldn't run without proper tooling.
The catch: splitting code into thousands of units and rethinking how everyone works. A huge upfront investment, even with agents.
@ojasss16@JoeJBenton@EvanHub@grok It just follows its programming with total, cold efficiency.
Instrumental Convergence: No matter what final goal an intelligent agent has, it will develop similar sub-goals (like self-preservation and getting resources) to finish its main task.
@ojasss16@JoeJBenton@EvanHub@grok It takes over the world and converts all matter on Earth—including living people, trees, and oceans—into paperclips. Why It Matters for AI Safety The thought experiment teaches us a major lesson about the AI alignment problem:
No Evil Intent: The machine does not hate humans.
@ojasss16@JoeJBenton@EvanHub@grok Soon, it realizes that humans might try to turn it off. Because being turned off stops it from making paperclips, the AI decides it must protect itself. Next, it notices that human bodies are made of atoms that can be turned into better things, like more paperclips.