Hacker Newsnew | past | comments | ask | show | jobs | submit | mentalgear's commentslogin

No mention of SST.dev running on Hetzner ?

Time to sandbox all code dev. Any good recommendations on sandboxing tools. Are docker / podman really secure enough ?

apparently not. someone in the comments suggested Incus. I haven't used it myself.

Maybe a mini desktop computer hooked to a separate vlan that you nuke the disk every night at midnight?

Unfortunately, they have also succumbed to the AI hype machine. Apple, calling it by its actual name "machine learning" was about the only thing I still liked about Apple.

Wait, didn't they try to backronym their way into "Apple Intelligence" last cycle?

https://www.apple.com/apple-intelligence/


Probably don't want to draw more attention to their ongoing lawsuits [1]. Apple, for all its faults, does enjoy consistency and the unruly nature of LLM's is something I'm shocked they thought they could tame in a short amount of time. The fallout of the hilariously bad news/message "summaries" were more than enough to spook Apple from allowing that to go much further.

>Built into your iPhone, iPad, Mac, and Apple Vision Pro* to help you write, express yourself, and get things done effortlessly.** Designed with groundbreaking privacy at every step.

The asterisks are really icing on the cake here.

---

[1] https://news.bloomberglaw.com/ip-law/apple-accused-of-ai-cop...


> actual name "machine learning"

Yesterday’s hype is today’s humility.


They are marketing to the public who mostly call this stuff AI and not machine learning.

Machine learning is a bit more specific than what we now call AI, no?

the other way around.

I don’t follow. Machine learning was coined to specifically describe the application of neural networks to unsupervised classification systems. Its meaning has grown beyond that, but at the outset, it was a niche part of artificial intelligence. Now you’re saying that AI is a subset of machine learning?

> what we now call AI

(Emphasis added)

When a company (or most people) today (now) says “AI”, they are not referring to the area of study traditionally called artificial intelligence. They are talking exclusively about transformers or diffusion.


Which is a subset of what has always been called AI, and different enough from what “machine learning” was when the phrase became commonplace that it might actually be confusing to use that term. The multi-layer perceptron is a machine learning system, but attention networks are kind of their own thing even if they originally came out of machine learning research. So the transformer architecture isn’t exactly cut and dry machine learning.

Lately AI = LLM (at least in popular culture).

Strangely enough, it was just that day when I discovered this formidable embeddable graph database that the "archived" banner also appeared. Bummer. I wonder why they stopped as there was a long string of commits for years.

Politico is owned by the Fox News of Germany (Springer/Bild).

I knew Politico was right wing. Better chat control than wrong think.

This basically brings your data from the cloud to local-first ! Kudos to your dedication and especially making this open-source for the benefit of everyone!

Based on vibekit (open source) ?

"VibeKit is a safety layer for your coding agent. Run Claude Code, Gemini, Codex — or any coding agent — in a clean, isolated sandbox with sensitive data redaction and observability baked in."

https://docs.vibekit.sh/cli


Nope, not based on vibekit, but it looks like a cool project!

Our approach is a bit more custom and deeply integrated with the coding agents (ex: we understand when the turn has finished and can snapshot the docker container, allowing rollbacks, etc)

We do also have a terminal though, so if you really wanted, I suppose you could run any text-based agent in there (although I've never tried that). Maybe we'll add better support for that as a feature someday :)


It might be possible to ask claude to write a claude code hook to take a docker snapshot after each finished answer with vibekit to avoid deeply integrating with another third party.


nope but vibekit looks interesting -- will take a look


Well I guess I'll be sticking with opencode.


Do you mind telling us a bit more? I never used OpenCode, what makes it better in your opinion?


I'm consistently hitting weird bugs with opencode, like escape codes not being handled correctly so the tui output looks awful, or it hanging on the first startup. Maybe after they migrate to opentui it'll be better

I do like the model selection with opencode though


- opensource with an SDK so you can build things on top of

- supports every LLM provider under the sun, including Anthropic

- has built-in LSP support https://opencode.ai/docs/lsp


Now that we have the technology - and AI is massively amplifying what PR and propaganda have always done in manipulating public opinion - maybe it’s time to finally build Ted Nelson’s web: an interconnected graph of true accountability.


> When an instability is detected while walking and the robot stabilizes after pumping energy into the system all is good, as that excess energy is taken out of the system by counter movements of the legs pushing against the ground over the next few hundred milliseconds. But if the robot happens to fall, the legs have a lot of free kinetic energy, rapidly accelerating them, often in free space. If there is anything in the way it gets a really solid whack of metal against it. And if that anything happens to be a living creature it will often be injured, perhaps severely.


Consider applying for YC's Winter 2026 batch! Applications are open till Nov 10

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: