I'm not sure how to be upset over this. For decades copyright has been extended and extended. Meanwhile, the ease and speed of spreading published works across the globe have increased massively. Don't forget that when copyright was first created, it could take multiple years for first editions to make it across the globe. In that environment, multiple decades of copyright makes sense. Whereas today that would actually stifle innovation and creativity rather than incentivizing it. I mean, because of the length of copyright we've gotten all of these live action remakes of Disney films or superhero movies rehashing the same stories.
I find it a little hard to be upset about AI. Supposedly stealing copyrighted works when the vast majority of those works. Probably should have been in the public domain to begin with. I have a faint hope that this scuffle between the AI companies and the publishing industry will result in more reasonable copyright laws, but I think it's more likely that exceptions will be made and AI will be treated as a special case.
I'm upset because I have to pay to use it or else be hit with onerous rate limiting for stuff that "should have been in the public domain to begin with."
If "god" is synonymous with the absolute truth of how the universe works then it can't address moral issues at all, since you can't prove an "ought" like "people shouldn't stab people" from an "is" like "stabbing hurts people and may kill them". You need to add extra axioms like "hurting people ought to be minimized".
The universe (and its intrinsic functions) holds no special value for people or trees or life in general. The best we can say is that it seems to have a slight preference for matter over antimatter for some reason.
All our moral codes and social norms we live by are purely human invention, derived from things that happen to work to produce a somewhat functioning society.
> All our moral codes and social norms we live by are purely human invention
I'd say much of it is selected for by evolution. e.g. Most animals don't fight or kill more than they need to since this is not optimal - risk of personal injury and depleting a food resource (there are exceptions of course such as the fox in the henhouse), and this is so universal it seems it has to have a genetic basis.
>All our moral codes and social norms we live by are purely human invention, derived from things that happen to work to produce a somewhat functioning society.
Religious people would disagree, and the evidences are all to see, hear and ponder from their sacred holy books. But the main questions are that which holy books are truly sacred, meaning that purely and utterly God words. Otherwise it's human innovation/exaggeration or something in between (i.e corrupted God words).
Based on these holy books, religious people adhere and follow their prophets teachings and actions to the best of their ability since the prophets are their fellow human beings and not angels send from the heavens.
This is actually a really common form of fingerprinting. At one point in time, it could generally tell the website what operating system and CPU architecture you're running on. I know this type of fingerprinting was fixed in Chrome so it always gives the same answers regardless of platform. I think it was also fixed in Firefox and Safari, but I don't follow their releases as closely.
Of course, even though it's probably useless now, things like that hang around because it costs more for trackers to remove the code than it does to keep it in.
Increased funding did improve educational outcomes. It just didn't improve it enough to be considered worthwhile.
The fact is that educational outcomes are caused by a large number of variables. Areas with (historically) poorer schools have less competent (or more corrupt) administrations so money doesn't go as far. Even with good leadership at the top (superintendent or public school committee), if the source of that money is not something you can rely on consistently enough to pay salaries with then you are not going to see significant changes.
Pre-K education is also critical to determining later school performance, as knowing the alphabet and numbers are practically assumed by kindergarten curriculums as they are covered far too far to be more than a review. But we have not tried that yet, so we'll see if that pans out...
Not being able to exhaustively test simple problem absolutely means you can't exhaustively test massively more complex problem. For a monotonically increasing function f, which I think we can agree a function mapping from the size of the input set to the runtime of an exhaustive execution of all elements in the set qualifies as, if |small| < |large| then f(small) < f(large). Therefore if f(small) is too big then f(large) is definitely too big.
> Not being able to exhaustively test simple problem absolutely means you can't exhaustively test massively more complex problem.
"Simple"/"complex" doesn't necessarily imply "small"/"large", respectively. Testing whether the Collatz conjecture holds is "simple", but its search space is (potentially) infinitely large. Testing correctness of concurrent data structures is (probably) a more "complex" problem, and yet it's still possible to perform exhaustive tests of their functionality [0].
In the end, the practicality of exhausting the 64-bit integer search space is irrelevant since it says nothing about the size of the write + checkpoint search space. I could just as easily say that it's easy to exhaust the 16-bit integer search space, but that doesn't mean exhaustively testing for this particular bug is easy.
They didn't share the prompt, but they did share two problematic training tasks where the AI went overboard. They also have examples from the AI's reasoning train of thought showing the AI knew it was sound something unintended.
right, but the cache retention time is very short for Anthropics LLMs.
5 minutes or 1 hour (with additional costs).
So you have to prompt basically non stop to not get a cache eviction.
A const prompt across all of Anthropic's subscribers could draw from a global cache rather than per-user?
Although saying that out loud makes me question it - each per-user chat and growing cache would need eventually to own its own ~contiguous memory block.
If there are no changes to the system message, yes this is possible and also likely done by Anthropic.
When there are additional local MCPs, reusability will be lower.
I think that Anthropic will bill you in any case :D
There are schemes where you don't need key pairs for each user (assuming the government has some way of authenticating users). Private State Tokens use blinded tokens for this.
It doesn't prevent tokens from being stolen or sold, but the token issuer only accepts each token once and can limit the rate that tokens are issued and control how fast they expire, giving decent control over how practical using stolen or sold tokens are.
I find it a little hard to be upset about AI. Supposedly stealing copyrighted works when the vast majority of those works. Probably should have been in the public domain to begin with. I have a faint hope that this scuffle between the AI companies and the publishing industry will result in more reasonable copyright laws, but I think it's more likely that exceptions will be made and AI will be treated as a special case.
reply