craftkiller

> Often times the math savvy crowd will even help explain stuff to people

Perhaps, but HN apparently is pretty hostile to people explaining why something is not clear and how it could be more clear based on all the downvotes. I'll keep that in mind: "If you don't understand something, shut up!"

Yes, but you need to consider your audience when posting links to this site. For example, if I posted to HN a link to a really good cookie recipe, it wouldn't be a good fit for this site. The author did nothing wrong, but that doesn't make it a good submission to HN.

This article is a bit more of a grey area. It is on a subject that is deeply relevant to this site and I'm sure some HN readers will get value from it. My point is that an even better submission to HN would have been an article on the same topic but written for programmers rather than mathematicians.

This article does two things which make it unhelpful for me:

1. It uses math notation. This is a software development site, wouldn't this be much more clear to this audience written as psuedo code instead of a complex series of greek and latin letters?

2. It doesn't step through a concrete example. What is hollow R raised to the (M * d_v) ? Is it a number, a set, a list, a matrix?

To use an example I understand: RAG used to seem complicated to me when only math people were talking about it in math words. For example, let's take a line from wikipedia on embeddings:

> the representation is a real-valued vector that encodes the meaning of the word in such a way that the words that are closer in the vector space are expected to be similar in meaning

What made it clear was learning that generating an embedding is just a function(string) -> list[float], and RAG is just taking a list[float] for your prompt and finding other list[float] that are close to it in your dataset.

So what would make this article more clear, is step through a simplified example of putting a value into a memory store and querying that memory store, showing the actual data structures.

While this is going to be fine for many users, wouldn't this continue our reliance on STUN for UDP hole punching? I'd certainly prefer a world where peer-to-peer networking is less reliant on work-arounds created for the NAT era.

That RFC does look like a nice improvement! Until then, since both devices have a ULA address, wouldn't simply deleting the printer from your phone/computer and re-adding it resolve the issue? Doing it when your public internet is down would probably even be best to ensure you're getting the ULA instead of GUA, thereby avoiding this issue entirely in the future.

(Admittedly I haven't had a printer for over a decade and I have never printed from a phone, so maybe I am missing something.)

My first job out of college passed the authentication over SSL/TLS, but the page that *contained* the login form? That was sent in the clear.... Truly dark days.

> if your Internet connection goes down and the router withdraws the prefix

My router advertises both the public prefix from my ISP and a ULA prefix for internal LAN connections (essentially the IPv6 equivalent to RFC-1918). This gives me stable local addresses that do not change and work without public internet access.

https://en.wikipedia.org/wiki/Unique_local_address

> without privacy extensions

Privacy extensions are enabled by default on every major operating system.

Why are you still using Twitter? It has been years since the purchase, you've had plenty of time to switch.

My 12 year old IPS monitor still has no burn-in. I grew up in the age of necessary screensavers, I ain't going back.

That sounds pretty difficult. We would need a way to signal to the municipal collectors what each thing is, and they would need to be able to store each type of thing separately. For example, are these batteries lithium, lead, or nimh? Those go into separate bins. What type of paint is it? those goes into separate bins. How are we supposed to indicate that a container contains used motor oil?

The simple trash, recycle, compost stuff works because it is only 3 categories so you can have 3 different colored bags/bins. Adding in the special waste stuff is adding 10+ categories. We'd either need a full rainbow of bins or some special stickers to mark each item (which would be yet-another barrier).

While I'd love at-home pickup, what we really need is for people to just be better.

In NYC it is free to dispose of such stuff properly (paid for with taxes). Of the sites I've been to, both have been easily accessible without a car. They've been incredibly easy to use: you walk up and either hand your stuff to the employee or put it in the correct bin yourself based on the signs. I've never had to wait in a line, you just walk in, drop your stuff, and go. And yet every week I still see people disposing of their electronics improperly by putting their smashed televisions on the curb for regular trash pick up instead of taking it to a special waste site. Hell is other people.

https://www.nyc.gov/site/dsny/what-we-do/programs/special-wa...

> current

That's the problem: if everything stayed at the current level of development, we would be fine. There is no reason to think that will be the case. Intelligence per watt has been rapidly improving [0]

> 1000x

Coincidentally, that is exactly the efficiency gain promised by ASICs like Taalas [1]

[0] https://hazyresearch.stanford.edu/blog/2025-11-11-ipw

[1] https://taalas.com/

What kinds of problems do you think someone will pay you to solve after AI advances for another 10 years?

This is my dilemma as well. I can't think of a role that won't be impacted so I don't know what to train for. When I was younger, my fallback was always going to be "truck driver" but that's going to be one of the first dominos to fall.

> we will live animals in a zoo

Why would we still be alive? Once we have advanced enough AI inside not only computers but robots, the rich will no longer need us. They will have the robots farm the land, clean their homes, fly their planes, etc. We will then have no income and therefore no way to feed ourselves. Most people do not own enough farm land to keep themselves alive.

In the past, such imbalances of power were resolved because the rich still relied upon the poor for everything (security, food, etc). I think we're approaching the point where even armed insurrection is no longer an option with the advances in automated killing machines. We're not there yet, but it's coming.

What makes you think those jobs will still be available?

Uber : Getting replaced by Waymo

Uber Eats : They've been experimenting with ways to replace humans for this. I think the most likely scenario isn't those silly robots on the sidewalks but rather just a Waymo. Have the restaurant send someone to carry food from the restaurant to the Waymo idling at the curb, and have the customer come out to the street to pick up their food from the Waymo. We lose the door-to-door convenience but I think a lot of people would take that trade to eliminate the cost of tips and paying the delivery person.

Security Guard : AI monitoring security cameras and automating calls to police sounds trivial. There are also already "security robots" that move around so they could be dispatched for scenarios which do not warrant a call to the police.

Handyman / physical jobs : While I think these will last longer than the others, I think they won't last much longer. We've already invented all the necessary hardware. The limiting factor was the software. What happens when you take an LLM from the year 2030 and give it a pile of tokens to write the software to drive the robots to do plumbing/building/whatever.

Currently we have C-level people saying "I want a website that sells widgets". Then product managers make a plan for all the pages/functionality on that site. Designers style the site. Managers divide up the work into smaller tasks. Tech leads steer a group of software engineers, and software engineers do the actual implementation.

2 years ago AI was a novelty and essentially just "fancy autocomplete". Today a skilled software engineer can use it to mostly one-shot the whole website with a detailed technical prompt and then iterate on it from there.

What happens 2 years from now? If progress continues, I expect we'll start walking up that chain of command. First you eliminate the software engineers and just have the tech leads do the prompting. Then you eliminate the tech leads and have the managers do the prompting. Then you eliminate the managers and designers and eventually you end up with star-trek-esque programming where some C-level person is just telling their computer "I want a website that sells widgets. Good! Now make the navigation bar blue. Could you make the font bigger?"

So what are your plans for when the wealthy no longer need you?

That's how I felt before I discovered tiling window managers. Now my flow is very similar to yours, but since I was wrong before, I can't rule out the possibility that I am still wrong and there might be something better out there. For example, Plan 9 had some interesting UX ideas like making the window manager infinitely nestable and mountable over the network. It also had the neat behavior in acme where to execute a command, you middle-click on some regular text. In the end, I'm not a fan of Plan 9's UI because of the reliance on the mouse, but I think it shows there's still room for novel innovation out there.

Here's an example of removing a branch that was posted to HN a little over a month ago: https://www.greyblake.com/blog/branchless-rust/

Assuming you are in the happy camp, what do you intend to do for food after all the software engineers have been replaced? The only people I can imagine as happy are those who are either close to retirement or wealthy enough to live off savings/investments.

The iso itself is read-only. When I need to update software I build a whole new iso and replace the existing one. For any folders/files that need persistence across reboots, I store that data separately from the iso, either in a virtual nvme drive or a mounted 9pfs folder from the host. The iso mounts the persistence drive/9pfs via /etc/fstab. So in the end, the software and config files are read-only (until I replace them with a new iso) but my databases and data files are read-write.

Getting into the unnecessary details: in classic live ISO fashion, the root (/) filesystem is tmpfs. Then I have at most 1 persistence drive and/or at most 1 9pfs mount. From there, I use bind mounts to only persist the few folders/files that I want preserved. For example, I'll have my persistence drive mounted at /persist and I'll bind mount "/var/lib/etcd" to "/persist/var/lib/etcd". Then when I reboot, everything outside of "/persist" is wiped out and I'm back to the initial iso + anything in /persist. So it is very similar to running a docker image with a volume mount, or running tails with persistence.

I build it all via NixOS and I use Impermanence to manage the bind mounts to the persist drive: https://wiki.nixos.org/wiki/Impermanence

One instance where rsync's delta merge comes up for me: For all of my virtual machines, I build a bootable live ISO so that way I can cleanly swap out the full ISO with a new one when there are software updates. This means that every time I want to change a config file inside the virtual machine, I end up building a new ISO and rsyncing it to my server. rsync speeds this up significantly since most (but not all) of the ISO is unchanged.

My last apartment had the overflow drain hole but it wasn't attached to anything. The water just went directly into the wooden cabinet underneath. Madness. Literally worse than not having the hole because it reduced the capacity of the sink. This is what happens when landlords who DGAF hire handymen who DGAF to do all the work on their units.

Here's the content of the tweet so you don't have to give twitter any more traffic:

  Hugging Face has been acquired by NVIDIA

  It is quite exciting to be a part of this journey! NVIDIA has been an active supporter of the llama.cpp project. For more than a year now, their engineers have actively contributed to the codebase, collaborated with the community and provisioned hardware for development and testing purposes. The local AI ecosystem has largely benefited from our joint efforts.

  Going forward, llama.cpp/ggml will stick to its founding principles. One of the most important  qualities of the project is to be hardware-agnostic. Therefore the development and support of all backends will continue to be done as usual - driven and shaped by the community. Open to everyone who is willing to participate. The existence of such an independent software platform is crucial for the rapid adoption of AI locally and for bringing it closer to the user.

  Now, with such a significant partner as NVIDIA supporting us, in addition to the wider Hugging Face team, I believe that our long-term goals will be easier to achieve. We will remain focused on building one of the most exciting projects based on the most exciting technology of our lifetime and making it truly available to everyone in the most accessible and efficient way possible.

My solution was to self-host PowerDNS and then sign up for the free DNS mirroring from hurricane electric. That way, I can administer my DNS records any way I want (these days I usually just manually edit PowerDNS's sqlite database) and if/when my PowerDNS server goes down, hurricane electric is still serving my records so the domain keeps resolving fine.

That seems like an expensive way to block ads considering traditional ad blockers have been doing that with a tiny fraction of the compute for decades.

How so? I've had no issues running llama.cpp with vulkan compute on my AMD graphics card (9070 XT). llama.cpp exposes an OpenAI API endpoint, which seems to be the lingua franca for AI applications. Does llama.cpp not work with datacenter AMD graphics cards?

Ah I didn't know Apple put their repair parts under a different domain. I stand corrected!

> That sure doesn't apply to this situation though does it?

It sort-of does. Naturally this situation can be resolved much cheaper with a flasher, but in this situation you absolutely have the option to replace the "broken" motherboard, while still keeping your screen, laptop chassis, battery, ram, SSD, wifi card, expansion cards, sound board, webcam. Another commenter said the board is available on Ebay for $270 which is going to be less than just the ram for a new laptop.