Ah yes, we must be getting closer to their IPO date.

Strongly recommend looking into Proxmox hypervisor as your base OS. Then install as many Linux VMs/LXCs as you like, with backups/snapshots all taken care for you. I’m on an Intel 8th gen CPU and have 3-4 VMs running, including Jellyfin, Home Assistant OS, an assortment of random docker images and a dedicated headless Debian VM for various AI projects. The CPU is more than enough, your 12th gen target will breeze through it. Get more RAM than you think you need - it will become the bottleneck for what you can do, not your CPU. I’m running 64GB and don’t regret it!

For video encoding/decoding like Jellyfin/Frigate, make sure you stick with Intel CPU for hardware codec support via QuickSync, otherwise you’ll need to add a GPU.

Uhhh… does anyone want to tell him?

My favourite part of this is that Zuck was so out of ideas that he tried to reinvent Google Glass at great expense and is running into exactly the same social brick wall that project did.

Perhaps this is more for you: https://makerworld.com/en/models/3134457-steam-machine-style-tissue-box-cover-1-1-replica

If you’ve actually read the sections of the Epstein files around this incident, including the artwork and diaries by some of the survivors, it’s very damning. The mainstream media simply aren’t touching a lot of it.

“Ex-Barclays boss denies having sex with child dressed as Snow White after Epstein emails”

Fixed that for you Guardian.

Only for those in the US it would seem.

We need more open hardware standards. Late-stage capitalism ruins everything.

And we’re about to watch Anthropic go through the same thing with their IPO. Late-stage capitalism sucks.

Read my post above to a different commenter, there’s a new type of ASIC being talked about that’s very different to an NPU or GPU, no RAM necessary.

LLM ASICs are different to NPUs. The idea is to encode a specific model’s weights (all billions of them) as a gigantic resistor network on the chip itself. Analog, where ohms law becomes the processor, and no RAM is necessary - the chip IS the model. It would pave the way for cost effective, fast, low-power local inference. Given the analog circuitry, there would be issues like interference, temperature sensitivity and so on, but we’re seeing models get good results even when heavily quantised, so the idea has legs. The big disadvantage is you’re stuck with that one model forever, so we’re likely to see these first appear in non-consumer applications like automotive, defence etc. A pretty exciting concept that solves some of the environmental issues around AI.

More info for the curious: https://taalas.com/the-path-to-ubiquitous-ai/

I’m really interested to see what comes of the LLM ASIC idea - it worked wonders for hardware video codecs. I could see the smartphone industry picking up again on annual releases worthy of upgrades if they included dedicated AI chips not beholden to a cloud subscription. I think we’re reaching the point that the frontier models are becoming “good enough” to use for a while before uploading to the next.

It’s good they’re pushing, but it’s not enough! Other countries do 10-year leases. Australian renters really are treated like dirt.

I swear this is the real reason AI companies are concerned about watermarking - not for the public good, but so they can keep their model training inputs clean.

Pauline Pantsdown just as relevant 20 years later.

I honestly think this is one of the biggest challenges that needs to be solved. The constant jargon, poor naming choices and assumed prior knowledge of pages of text is extremely fatiguing. It only gets worse as the volume of LLM-generated content in your project increases.

We created an AI usage code of conduct at our workplace. One of the golden rules is “Don’t make other people’s jobs harder.”

As an example, we tried getting AI to generate weekly bug reports from analysing the most common crashes and asserts for the week, assigned to the right person with relevant context. It was a great idea, but too often the bug description and even the title required way too much jargon and comprehension to get through, and often the findings weren’t quite right anyway, so it was only useful as a marker of “look at this crash please”.

These walls of text people share with each other are another great example. If you’re going to do that, you really have to ask it to talk as simply and as briefly as possible with no context assumed. And then read it/correct it yourself!

Reminds me of Sam Altman saying he uses AI to make a podcast about his kids sports game so they can listen to it on the way to school. As if talking to your children wasn’t an option.

All these tech bros creating solutions in search of a problem.

I remember reading an article a while back about high powered lasers that could essentially take down drones via direct line of sight, but haven’t heard much more since then. I guess they’re difficult to target from the ground if they’re flying at very low altitude.