notsylver

I feel like there is a reasonable compromise between "yeah they have full internet access" and "separate airgapped rooms that require multiple levels of authorization to access" that would make this a lot better without that much more work. I feel like they're doing it intentionally to show how dangerous these models are and that the government must step in and protect them

I desperately want to see some of these projects used in production by the companies or people making them, at best it might prove that it isn't just a sloppy publicity stunt

There is a PR for that https://github.com/librepods-org/librepods/pull/655

I tried it out, it seems to work well. Hopefully it's not blocked by patents

It seems like the right answer is usually the longest of the choices, I managed to get a few just by picking the longest. It would also be nice if there was a "I don't know" instead of guessing and skewing the results by getting it right, though maybe thats accounted for

I believe they added it back after people noticed, archive.org has versions where its gone

I am fine with the price increase, for me its how sneaky they're being about everything. If they sent a few emails about the recent changes I wouldn't care, but it feels like they do not want customers to know which is the last thing I want from a password manager.

I really want to use linux but for me its the software. Games I play use anticheat that doesn't work. Software I use for 3d printing doesn't work without hours of workarounds. VR is poorly supported if at all. Usually every year I'll try linux, use it for a month or so, then be forced back to windows after running into a hard block, with the dozen ads forced down my throat welcoming me back. Hopefully one day I can commit to it.

I use a medium sized Australian ISP and got the same, maybe they just have whole regions blocked...?

There's also a locked door in one of the buildings, but I couldn't figure out what either mean if anything. It would be cool if the radio was used for something though

I pay for github copilot solely for completions and use codex for actual agent work. I really wish I could pay like $5/month for just completions or there was a good local alternative for them.

This feels like a buzzfeed quizz for developers. If you think about each step long enough you can't really get a wrong answer

I always thought eventually someone would come along and make a hardware accelerator for LLMs, but I thought it would be like google TPUs where you can load up whatever model you want. Baking the model into hardware sounds like the monkey paw curled, but it might be interesting selling an old.. MPU..? because it wasn't smart enough for your latest project

You might be confused with their new-ish US datacenter? Hetzner is still European-owned.

I feel like cursors solution is still the best answer. Let the model suggest edits in whatever format it prefers using as few "extra" tokens as possible and have a small model figure it out. I don't use cursor anymore but when I did it was impressive how consistently it worked, I think there was a single time it failed. 70b might be overkill though...

I think people are misunderstanding. This isn't CT logs, its a wildcard certificate so it wouldn't leak the "nas" part. It's sentry catching client-side traces and calling home with them, and then picking out the hostname from the request that sent them (ie, "nas.nothing-special.whatever.example.com") and trying to poll it for whatever reason, which is going to a separate server that is catching the wildcard domain and being rejected.

I've been waiting for something like this for ages. Cursor making me pay $20/month when all I use from it is autocomplete was always a little annoying, especially as they changed the UI to push agents more and it got in the way. I was even considering doing it myself but wasn't sure about gambling on models small enough to run locally being smart enough to do anything useful.

I threw together a vscode extension to run it and while the extension is rough, the model seems decent. I'm trying to keep my expectations contained, in the past local models have been absolutely terrible for inline completion, this seems much better already. I hope this kicks off more competition.

This really needs some examples or a demo that doesn't require signing in. As-is this is just a pricing page with a sign in button and no information on what it actually does?

It seems to let you access head tracking data, so now I'm really curious if it would be accurate enough to use with games (eg, microsoft flight sim/arma 3/euro truck simulator 2 head tracking). There is probably a lot of other interesting use cases for it too, but I'm stuck with windows for now so :(

I've really wanted to fine tune an inline code completion model to see if I could get at all close to cursor (I can't, but it would be fun), but as far as I know there are no open diffusion models to use as a base, and especially not any that would be good as a base. Hopefully something comes out soon that is viable for it

I tried a dozen or so images. For some it definitely failed (altering details, leaving damage behind, needing a second attempt to get a better result) but on others it did great. With a human in the loop approving the AI version or marking it for manual correction I think it would save a lot of time.

This is the first image I tried:

https://i.imgur.com/MXgthty.jpeg (before)

https://i.imgur.com/Y5lGcnx.png (after)

Sure, I could manually correct that quite easily and would do a better job, but that image is not important to us, it would just be nicer to have it than not.

I'll probably wait for the next version of this model before committing to doing it, but its exciting that we're almost there.

I didn't do any videos, just pictures, but considering how little I found for pictures I doubt you'll find much

I scanned everything into 48-bit RAW and treat those as the originals, including the IR scan for ICE and a lower quality scan of the metadata. The problem is sharing them - important images I manually repair and export as JPEG which is time consuming (15-30 minutes per image, there are about 14000 total) so if its "generic family gathering picture #8228" I would rather let AI repair it, assuming it doesn't butcher faces and other important details. Until then I made a script that exports the raws with basic cropping and colour correction but it can't fix the colours which is the biggest issue.

I digitised our family photos but a lot of them were damaged (shifted colours, spills, fingerprints on film, spots) that are difficult to correct for so many images. I've been waiting for image gen to catch up enough to be able to repair them all in bulk without changing details, especially faces. This looks very good at restoring images without altering details or adding them where they are missing, so it might finally be time.

If CV cheats are good enough that people are using them (and then getting banned), and other people are willing to pay >$1000 for "undetected" cheats (that still get them banned)... wouldn't making custom hardware that is just a capture card and USB keyboard+mouse running one of those CV models that sends the inputs back over a "real" keyboard work?

I have tried fully switching to bun repeatedly since it came out and every time I got 90% of the way there only to hit a problem that couldn't be worked around. Last I tried I was still stuck on some libraries requiring napi functions that weren't implemented in bun yet, as well an issue I forget but it was vaguely something like `opendir` silently ignoring the `recursive` option causing a huge headache.

I'm waiting patiently for bun to catch up because I would love to switch but I don't think its ready for production use in larger projects yet. Even when things work, a lot of the bun-specific functionality sounds nice at first but feels like an afterthought in practice, and the documentation is far from the quality of node.js

You already need very high end hardware to run useful local LLMs, I don't know if a 200gb vector database will be the dealbreaker in that scenario. But I wonder how small you could get it with compression and quantization on top

I'ts fun, I think it needs queues for different game modes because with 150 players you almost always get horded by neighbours. Being able to queue for a team game would make it a bit easier to learn I think

I've had this too, especially it getting stuck at the very end and just.. never finishing. Once the usage-based billing comes into effect I think I'll try cursor again. What local models are you using? The local models I tried for autocomplete were unusable, though based on aiders benchmark I never really tried with larger models for chat. If I could I would love to go local-only instead.

I showed the model a picture and any text written on that picture and asked it to guess a latitude/longitude using the tool use API for structured outputs. That was in addition to having it transcribe the hand written text and extracting location names, which was my original goal until I saw how good it was at guessing exact coordinates. It would guess within ~200km on average, even on pictures with no information written on them.