philipwhiuk
philipwhiuk on Hacker News
This is to get sites that monetize Facebook as a funnel to pay.
That's not sovereignty, that's redundancy.
Sovereignty is just where stuff is, not how many place it is there.
I'm not sure it's good for any country to depend deeply on the technologic might of another country - be that the US on Chinese AI or the UK on American or any combination.
Just as the US has put up restrictions around export of technology and use of non-American tech in CNI, so should every other country.
It's not really missed though. He does spend a good amount of time saying that it'll be a problem for many. Whether it's most of professional mathematicians is an open question I think.
> Only the first 5 characters of a SHA-1 hash go to HIBP
Some would consider the first five characters being sent over the internet to be 'your secret leaving your system'.
Some would consider phoning home to a remote server to be worrisome. It's home assistant not cloud assistant.
That's not even the original part - it's been done by a car with a driver in it - the McMurtry Spéirling.
Video games is about as big an industry as TV and movies.
The canon doesn't even arrange all the books. It's a somewhat arbitrary selection.
I've never understand the correlation between using Linux and needing to suddenly have everything run in a CLI to the extent of building your own windows in text.
It's like you're not aware Linux has desktop environments.
Why is Mars so close and small?
No I haven't. Because the purpose of a paper is communicating the development in the field. If they aren't doing it, they're bad papers.
How do I know when I read Claude's interpretation of the paper that it's not hallucinating.
Heating the ceiling with the light is an extremely inefficient approach to heating a room.
I'm worried that the frontier lab doesn't read the code they are writing.
It's also hard to believe all these non-Git alternatives aren't merely an attempt to capture developers into a new proprietary solution.
This could have been done as an open-source fork of Git.
There's plenty of highly complex benchmarks for this stuff. Simon doesn't need to do that.
One problem with LLMs in academic contexts is that because you're paying by the token, it's resulting in extremely short papers that don't adequately explain the context. It's the same problem with long-standing short papers.
And this is a fine example of this.
For now, an AI agent is not a 'someone'
SIGBOVIK with extra steps.
By the way, are you really allowed to deeplink to Imgur to host most of your templates?
(I only noticed because your site is not blocked in the UK but most of the templates are.)
The -182 feels highly arbitrary.
> The anti AI slant is hilarious since the whole site is mined for AI.
The whole of everything is mined for AI. The decisions of the site owners to profit from the mining aren't necessarily the views of the contributors.
My guess is that the constant terms are large enough it's not practically useful in most cases.
It's interesting (and unsurprising I guess) to see that bike lanes in the US are much like those in the UK - a line on the road, rather than real segmentation.
This is really the Grant Sanderson position - reframe mathematics from proof-search to understanding-distribution.
Some of this is minor reasonable adjustments, much of it will just be ignored.
Thank you for answering so many detailed questions on this.
It's amusing they carefully refrain from mentioning the battery manufacturer.
But this is a fundamental problem with using it as a 'space is crowded' argument.
> For one thing, anything in LEO would seem to be moving so darn fast that it'd be disorienting.
They take ~90 minutes to orbit - it's not that quick. The satellites are actually moving at the right speed.