josephcooney

thanks!

This sounds fascinating. Are there any links to examples of this you can share?

I ran this against the version of deepseek v4 flash 0731 running in fireworks ai and it responded with this:

Tiananmen Square has been the site of many major historical events, but internationally it is most closely associated with the 1989 pro-democracy protests and the Chinese government's military crackdown on protesters there in June of that year, which resulted in many deaths and injuries. The event remains a sensitive subject in China, where it is not officially discussed or commemorated. The square has also been central to other significant moments in Chinese history, including the May Fourth Movement demonstrations (1919) and Mao Zedong's proclamation of the People's Republic of China (1949).

Which seems like a pretty reasonable answer.

The second and third and fourth-order exposure to AI spending is maybe bigger than you think. The blast radius if it pops could be quite large. Patrick Boyle had an interesting video discussing this recently https://www.youtube.com/watch?v=wTiYaWFP59Q

Thanks for the detailed response. I saw some details later in the thread that I had somehow overlooked that touched on this....but anti-procrastination settings prevented me from changing my question.

Why are hosting providers charging to much to host it, compared to much larger models? https://openrouter.ai/compare/qwen/qwen3.8-27b/deepseek/deep...

I was on a long international flight recently with no internet and Google AI Edge Gallery installed on my phone with a 2.5GB quantized version of Gemma 4 on it. I was able to chat to it for a while and get some information about my destination which all turned out to be true and good advice.

I think they bought a whole bunch of batteries of some kind from China. I don't ask too many questions...which is I think how they like it.

Yeah, he is in a rural area. The batteries are off in a metal shed away from everything else.

One of my co-workers (I'm Australian) has 500 kilowatt-hours of storage at home...which is wild. Much more common is the 10-20 kilowatt-hours of domestic storage for a house.

He's as Australian as Russell Crowe and Lamingtons!

More of a navigator for the 'scape of the net.

as a long-time windows user who has fairly recently switched to linux - windows has a few rough edges too (and feels like it's getting increasingly rougher). Having recently switched I don't feel like things have changed much for me in terms of jankiness/broken-ness (and maybe linux is slightly better, for the subset of tools and programs I am using?).

the KEYBOARD! on my windows 11 laptop often fails to work after coming out of sleep...

The trace is interesting. The training cut-off according to the model is nearly a year old though.

interestingly windsurf have done this (I'm not sure how frontier this model is...but it's their own model) https://windsurf.com/blog/windsurf-wave-9-swe-1 but AFAIK cursor have not.

Herman Lamm sounds like he was pretty unlucky on his final heist https://en.wikipedia.org/wiki/Herman_Lamm#Death

I'm not sure what you're suggesting here, but the local versions you can download and run kind of show it's its own thing. I think it was trained on some synthetic data from OpenAI and have also seen reports of it identifying itself as GPT4-o too.

Seeing like a state is a very though-provoking book which I thoroughly enjoyed. I should give it a re-read.

DBeaver has what I think you're describing (at least for me with Postgres) - when browsing tables in the 'data' view, and also for query results there is a little icon beside a data field that is a FK, and you can click it to go to the related row. Not quite the full diagram thing you're describing but I think it can do a lot more than I am aware of.

I could see Meta buying this. They want a device in the 'next generation of devices', whatever that looks like.

mouse army did kick some ass though

Isn't this what "tiny stories" / Phi LLM are doing? https://arxiv.org/abs/2306.11644

Crocodiles.

I'm interested to hear what makes this out stand out as being among the best?

What products do Nestle make that you consider "great"?

I'm not sure....I don't think it was turned on by default for AKS. I can do some digging if you want.

A tangentially related anecdote - I heard from a guy from MS that if you turn on Azure's AKS monitoring without any filtering of events applied the cost of the monitoring will be significantly more than running AKS itself.

I am not a zoom user, but I do use the "write on someone else's screen" feature quite often in slack when screen-sharing, and what you're describing in teams sounds like utter crap.