The machine itself is basically useless for any type of realtime inference, no matter what the marketing page states, but I still use it for prototyping LLM integrations and running comparisons across MoE models.

If only the alternatives to framework desktop wouldn't be so poorly built, I might swap it out for a local machine which has more ram but comparable performance for stuff like gpt-oss-20b (around 70tok/s)

"European sunscreens are safer than American (2024)" https://news.ycombinator.com/item?id=48503940

Is this the person I have to complain about for the removal of fulldisk encryption in Android 13?

I agree on some points about the missuse of AI particularly for surveillance, military and propaganda.

But this reads like a post further glazing Mythos, and we are just one or two years away "trust us guys", and similar to Mistral's policy plea "please use AI everywhere or we are going to be left behind".

I had the hardest time accepting one of his first points that LLMs could barely write a line of code 4 years ago.

ChatGPT 3.5 was reasonable at code writing but hallucinated a lot of library functions. Yes we have better harnessing today, and models have been further finetunned with reallife code, but pushing this argument just to support his exponential narrative is deceptive. Like most AI marketing.

If it is that much better in practice I'll await confirmation through some kind of research paper before building even more stacked layers of LLMs.

That's the kind of claim that requires and asterix, and things like this are what feeds into the AI propaganda machine.

That is an anualized revenue, which are projected numbers and not "real numbers".

Also one of my worries with the mercusour trade deal. And any deal that involves meat imports from the US, with specific laxer regulation requirements (at least what Trump would like).

From a quick search I gather that US citizens' daily use is around 300B gallons https://watercalculator.org/footprint/how-united-states-uses...

I was just curious of the number and scale.

Artists and writers are unionized, why they have a more powerful collective voice.

Second, there are enough peole for which their jobs are very well paid and too cozy to dare to rock the boat.

The economy and job market isn't so hot either at the moment for people to quickly be able to jump ship.

Can you even be sure that you find a tech company that isn't jumping head first onto the AI hype train? Even politicians can't have enough of AI in their mouth.

I kind of hope at least they'll fix such issues permanently before the steam machine release.

Default configurations as well, since it states FIPS compliance it has to change defaults <https://fedoraproject.org/wiki/Changes/RemoveFipsModeSetup#W...>

Their UX is really at most a proof of concept, not good enough for daily use.

Or has sanity reached the Angular ecosystem by now?

In practice I haven't got around to building something around multimodality since I'm primarily using their text generation capabilities.

They have their own custom compositor for handheld-mode, named gamescope. https://github.com/ValveSoftware/gamescope?tab=readme-ov-fil... XWayland based.

Post classified under "improvement". Maybe for revenue targets.

I can't believe that font and color passes accessibility checks.

At the same time having these large companies that operate in emerging economies where they hide under the rug all child work exploitation.