I would have agreed with you before they pointed out that "frozen water" gets a word: ice. Honestly, I think it's reasonable: people deal with frozen water far more than they do boiling water, but it changes it from a case of "what are they talking about?" to "okay, where do we draw the line?" for me.
But water that has boiled into gas also gets a word: steam.
As far as I'm aware, there is no separate word for freezing water -- i.e. water that is very cold and will, if it continues to get colder (and has something to crystallise around), turn into ice.
So the symmetry seems complete: ice -> freezing water -> water -> boiling water -> steam.
Freezing water is already at or below 0, it doesn't need to get "colder" to turn into ice, it simply needs to exchange the energy with the environment and rearrange in crystals.
Basically as it gets colder water exchanges energy with the environment and gets colder.
But once it reaches freezing temperature, it can no longer get colder and all the energy is used for the formation of crystals.
“Water normally freezes at 273.15 K (0.0 °C; 32 °F), but it can be "supercooled" at standard pressure down to its crystal homogeneous nucleation at almost 224.8 K (−48.3 °C; −55.0 °F).”
So, I got the physics wrong. Apologies and thanks for the correction.
But the semantic point still stands. Boiling water is still water -- in the specific sense of H2O in its liquid state -- while ice is not. The complaint that frozen water has a single-word synonym while boiling water does not is making a false equivalence.
Yes, is that not the same with boiling water? It doesn't need to get "hotter" to turn to steam, it needs to exchange the energy with the environment to gasify
Frozen water represents a state change and that different state commonly gets its own word: ice/water/steam equates to solid/liquid/gas
Boiling/freezing water represents the state of the liquid, not the transition. Its descriptive. Water boils away into steam, or freezes into ice.
Should we consider luke-warm water also singular? What about body-temperature water? cool water? It makes sense not to treat adjectives/descriptive words combined with the subject as singular because the definition already exists in the root of the words (meaning of adjective word + meaning of subject word). Blue clay is another example, why would that be a singular?
It really only makes sense to me in the rare cases where the combination words represent something different or non obvious than the combined meanings of the two words (i.e to 'give up')
Ice, slush, sleet, snow, graupel, hail... And within there is a subtype "black ice", a compound noun that isn't really just a description (it's not black, it's nearly invisible - a similar sense as another one, "black hole", which you'd never figure out from the components alone).
We have a lot of words for "frozen water" because it takes a lot of forms. As far as I know "boiling water" is only one thing so we've never needed additional words to distinguish it.
Says you get 3 submission attempts, but doesn't say how many attempts you've made. It doesn't say why you were rejected, just that you were. Microsoft Support's phone lines literally hear the word "Office" and say that all support for office is handled online, give you the URL, and hang up on you. And they black list your phone number for the day: any calls after that go straight to a canned message and immediate hang-up, no alternative available.
Multiple people from MS support have called me back over the month I've been at this, and each of them says, in summary, 1. I can't help you 2. Call this number 3. Give another number like the above that does not actually help.
But the crime in the human instance is the reproduction, not the storage. So the crime in the AI circumstance would not be in the training, but in prompting the output.
And of course AIs are excellent at taking direction, so:
If I prompt it with "Harry Potter, but Voldemort wins: dark, and Hermione is a sex slave to Draco Malfoy" and get "Manacled," that's copyright infringement, and on me, not on the LLM/training.
If I prompt it with "Harry Potter, but Voldemort wins: dark, and Hermione is a sex slave to Draco Malfoy, and change enough to avoid infringing copyright," and get "Alchemised," then that should be fine. I doubt the legal world agrees with me though.
> But the crime in the human instance is the reproduction, not the storage. So the crime in the AI circumstance would not be in the training, but in prompting the output.
I wouldn't be so sure, at least under US law. 17 USC 101 defines a "copy" as:
[...] material objects, other than phonorecords, in which a work is fixed by any method now known or later developed, and from which the work can be perceived, reproduced, or otherwise communicated, either directly or with the aid of a machine or device.
If I memorize a work what ends up in my brain is not a copy according to that definition because with current technology there is no machine or device which can be used to perceive, reproduce, or otherwise communicate it. The work can only be perceived, reproduced, or otherwise communicated by using my brain which is not a machine or device.
No copy in my brain means that memorizing the work cannot infringe the copyright owner's exclusive right to reproduce the work in copies.
An LLM, unlike my brain, is a machine or device which can be used to perceive, reproduce, or otherwise communicate the work and so the work stored in the LLM is a copy.
Training an LLM then, unlike a brain memorizing a work, makes a copy and so would be covered by the copyright owner's exclusive right to make copies.
That's going to need to be justified, probably by arguing fair use.
I'd argue your brain is that "machine or device" -- the fact that the storage and the playback mechanism are one and the same is irrelevant. The fact that you have to be willing/induced to replay the content back just makes you a worse machine :-)
Interesting argument but not likely to go far. As far as I can tell US copyright law has never been taken to include brains as machines or devices.
This is actually relevant in some real cases, namely improvised works. Attempts to claim copyright on improvised works that were not recorded have generally failed. If brains counted as machines or devices than the work inside the performer's head would be a recording and the work would have copyright.
That is one of the reasons it is usually recommended that musicians should record their live performances. That gets them copyright on anything they improvise during the show. Also it gets them copyright on that particular performance of their music, which helps them go after anyone who makes an unauthorized recording of the show. (Copyright is only automatic upon recording when the recording is by or under the authority of the creator).
This speaks very much to the idea that LLMs are in some sense a ridiculously effective, somewhat lossy, compression algorithm that has been applied to the whole internet.
It's a good way to frame base models that have only been pretrained.
However, modern frontier models have undergone rounds of fine-tuning, RLHF (reinforcement learning from human feedback), and RLVR (RL from verifiable rewards) that turn them into something else. The compressed internet is still in there, but it's wrapped in problem-solving and people-pleasing circuitry.
Facebook knows I'm male, and I see things like this very rarely -- on the order of one or two a month. Maybe that's FB ('s algorithm) testing my "defenses": they/it show me something like that as an experiment, and if I ever clicked on it, the floodgates would open.
But I don't, so it doesn't. Or maybe FB knows I'm happily married and that won't work on me in the first place.
Or maybe FB knows I'm a sucker for chess and go puzzles, so they're my equivalent of this?
It works fairly well because your PM and King aren't complete loons. At the end of the process there has to be someone making decisions, and when that person is a narcissistic 8-year-old in an 80-year-old's body, bad things are going to happen no matter how the system is written.
To the extent this is true, the entire thing could literally be a grift (I'm not saying Trump is smart enough to come up with this, just that people around him are, and he's grifty enough to go along with it):
1. Trump enacts the tariff, despite knowing it will be struck down.
2. The tariff extracts hundreds of billions from the economy.
3. Finance firms buy the potential refund for pennies on the dollar, knowing that Trump has no plan to defend the tariff.
4. The Supreme Court strikes down the tariff, as planned.
5. The finance firms profit on the refunds.
6. We are all poorer, Trump's cronies are richer.
Trump has been obsessed with tariffs for decades. I fully believe he thought this was a great idea. Lutnick, on the other hand, quite obviously forsaw this eventuality (as did anyone who understands how power actually flows in the United States) and encouraged it while preparing to profit massively himself. It's an obvious play, good on him for getting away with it. It's clear at this point that this administration has utterly collapsed the idea of the rule of law, though. 15 years ago this would have been a scandal that would have led to firings and possibly impeachment
The team/person responsible for Woot sold it to Amazon, and then launched Meh the day their non-compete ended, along with a manifesto explaining how badly they thought Amazon had handled Woot.
Depends, I love it and am happy to pay for it out of privacy concerns and supporting a non-monopolist. It's got some neat features that I use all the time that google doesn't have. Is it's search results better than google? Maybe. Maybe not. I do know when I can't find something on kagi, google doesn't either.
I have no clue where I read it, that was back when meh.com launched eleven-ish years ago. I didn't find it in a hot minute of searching either. I did find these, some of which talk about the circumstances obliquely:
reply