Disagree. @lionkor compared them to a hammer, and @ragequittah is saying they're not like a hammer.
The narrow gap between downloadable and frontier models is tangential to this. If you want to expand on the "hammer" metaphor, the downloadable models are a small construction/demolitions firm, and the frontier models are a big construction/demolitions firm.
And in this analogy, we're still working out what planning permission laws are while both are being hired all over the place.
People can love the Japanese for their unique characteristics and they can love the Japanese for the things they appropriated from other cultures. Just like the British.
Fact is though, in many ways, Japanese culture is a lot more caring, considerate and kind than the British.
>I doubt that purely wordless, symbolic system was truly the ultimate pinnacle of operational clarity.
Agreed, but only because Brits aren't really good at designing things for longevity, whereas the Japanese are great at it.
Cookies are not the meta for tracking. Fingerprinting [1] is highly effective across IPs and across incognito/browser profiles.
The only browsers I've found that defeat it on desktop are Mullvad and Tor Browser as well as most antidetect browsers.
Mullvad (essentially Tor Browser without the Tor network built in for better usability on the web) and Tor browser are blocked from using or signing up for most websites because of IP reputation and the privacy protections which make them look like VMs and bots.
Antidetect browsers are used by marketers, web scrapers, social media account managers, ecom, among other low lifes. These use different browser profiles to avoid fingerprinting by isolating and randomizimg identifiers while constantly patching(use a few week old version and you'll see it gets detected as something to be blocked, not detected as in correlated cross profile, just blocked) to avoid detection. I'd like to see a privacy focused antidetect browser, but perhaps open source makes it easier for the fingerprinters to find detections.
Your second kid only takes half the effort. Half it again for a third. New parents always try to be perfect. It takes a while for them to get pragmatic.
Some of the inference engines can process multiple requests in parallel more efficiently than doing them sequentially. Not sure of the exact mechanism but e.g. llama.cpp's llama-server can do this (you tell it the number of slots to have when starting, then fire HTTP requests at it and it batches them together when it can).
Waiting for the hooman (or tool calls) won't help either, of course.
A female high on meth gets disoriented and dies from heat exposure in the mountains. As per article she willingly separated from the guy whom she sent to the top of the peak "to continue to get pic for social media". He probably should have know better and just go down with her and call it a day, or not getting high on meth in dangerous environment in the first place but thats about it.
Unless you have a better article on that, that really ain't evidence of anything.
> ...a bit of an odd question: how well do LLMs losslessly compress, as in for cold storage?
TBH this is like the near last ranking consideration in cost for being able to download and run this. Even though HDD and SSD prices have gone nuts as a result of the recent demand/shortage, it's not like 1.5TB of space costs a lot.
Even if you fed it into xzip with the most cpu intensive compression options and it didn't compress at all (eg: like trying to xzip an AV1 video, or whatever), it's still the cost of a single fast food hamburger meal in $/TB. The real concern is the RAM to run it.
This is on the level of "just use rsync bro". I prefer the console experience of clicking a button to order, plugging it in and playing. Valve is doing the vertical integration, they're putting together the hardware, the software and the OS, of course you'll be getting a better experience than pulling together random stuff. Also game devs will be optimizing games specifically for this hardware combo, just like consoles.
Most people thought Fable had more 'taste' than Opus, there was certainly a better quality of writing that felt more 'smart human' and not 'stochastic parrot stringing sentences together'.
1. Sometimes you need a request body.
2. POST cannot be guaranteed to be safe if re-sent.
3. This is GET with a request body, guaranteed* to be safe if re-sent.
* With the caveat that it's only guaranteed if the server is following the RFC correctly.
I am jokingly agreeing with you: Bill Gates famously claimed 640 kilobytes of RAM was sufficient for everyone, which felt like a funny parallel to people telling you "open source models are sufficient for everyone"
There's obviously a competitive advantage to having better tools, and I find it amusing/absurd that some people don't get that.
