gfosco

I didn't feel like installing a whole redis-server service on my Windows machine, so instead I built a lightweight yet very capable clone. Maybe you'll find it handy for local development.

I set this up today on my 5090 at Q6_K quantization and Q4_0 KV, got 50 tokens/s consistently at 123k context, using ~28/32gb vram through LM Studio.

Yeah, it's token intensive but worth it. I built a very dumb example harness which used IDA via MCP and analyzed/renamed/commented all ~67k functions in a binary, using Claude Haiku for about $150. A local model could've accomplished it for much less/free. The knowledge base it outputs and the marked up IDA db are super valuable.

No you fell for someone elses joke.

Somehow not surprising to see this take on the front page. How embarrassing.

I remember being an early teenager on IRC and joining a small BBS development team. I said I knew Pascal, but that was a fib... I did learn it quickly, though, and used it for many years. I can still find the code online, and see my contribution which I get a kick out of... It was a tool that you entered colored ascii frames for an animated "press any key to continue" prompt, and it output a valid pascal function file for that animation.

Same here... It's still amazing, and I have no legitimate reason to upgrade. Could be looked at as a failure of some sort, that they haven't given us a good reason.

Not mentioned in the article, but one major provider for this is AskSage.ai, which has built up a lot of features around creating government documents, proposals, RFPs, etc.

Depends who you ask. To me, NDT is unbearable too...

Not by themselves... but with a good program built around it, managing the actual state, and very careful prompting, yes. I've been thinking about this for a while, always have the desire to make a game.

Looking to get rid of a few?....

The OpenAI reference came at the end, and it appears it's mostly a fallback... an option, that users must explicitly allow every time. Hardly a dependency. Most of the time, it will be on-device or apple-hosted in "private compute cloud", not connected to OpenAI at all.

At the end of the article, that's my takeaway too. I can't imagine going to this level of effort on the Vision Pro and not keeping it to see how the software develops.

It is extremely silly to think using Gab AI says anything about you, except you want good AI that isn't ruined. Extremely silly.

You really don't have to talk to Hitler, there are 200+ other characters to choose from... and no, I don't think there is a middle ground when it comes to censorship. Either all legal speech is okay, or it isn't. If some legal speech is so problematic, maybe it shouldn't be legal? Until then, yeah, we're going to make the necessary point.

If you want uncensored.. https://gab.ai is also comparing well with gpt4.

I have the Bambu Lab X1C with the AMS and I'm chomping at the bit to get a second AMS and the Hub... If they weren't backordered I would have already. I've had machines from Type A, Ultimaker, a T-Rex 2+ whatever that was, a Raise3D Pro2 plus.. nothing has impressed like the bambu.

You think a font being a few tenths of a millimeter off spec is a "dangerous defect" ? Let me guess... not a fan of Elon.

Perhaps it's time to fight over words. The definition of recall no longer works, and I believe it is constantly used specifically negatively against Tesla. It has an effect on most people, who aren't savvy, who don't read beyond the headline. (i've not participated in these alleged battles before)

It's always worth it... Sometimes, even the downvotes are enjoyable.

WSG, Windows Subsystem for GPT

Connectivity and freedom don't pay salaries or cloud bills, unfortunately. Any rational actor has to focus on maintaining profitability.

Self hosted postgres, not RDS.

Have done minimal downtime major version upgrades using standard replication, switching between two "master" servers. In the five minute range.

I don't recognize the OSI as any authority to define words.

we’ve instead entered a period of “Enshittification”, where the platforms we loved and trusted have been taken over by corporate interests that no longer care what the users want or need.

It's funny to me because we've been here for many years. i.e. Facebook hasn't given two shits what people want or need in a decade. People have been asking for more control of their feed, and complaining about user-hostile decisions made all this time. The other big tech cos are no different, and "trust" has always been low..

This is already being attempted with AI, under the guise of "safety".

A previous HN thread on SoftRAM: https://news.ycombinator.com/item?id=21768268

Many people knew it was a bigger deal, with a casual glance at the chemicals involved. It has been a typical snowjob because the media likes the people in charge.

Heck yeah. There's a pretty decent chance it's still available somewhere. I can still find code I wrote almost 30 years ago for a tiny BBS software.