kenmacd
kenmacd on Hacker News
I think it's a significant distinction. The original title implies that the training was illegal when the court actually found that the training was fair use. It implies that Anthropic was required to ask the professors for permission to train on their books.
What the actual ruling found was that the professors should have been paid some tens of dollars for 1 copy of their book.
(Whether you agree with it or not is another matter)
The title is (I believe intentionally) inaccurate. Better would be:
Anthropic trained on illegally acquired Stanford professors' books
The article does not say the training was illegal (although does discuss the professors push for a licensing model).
How would those tool calls work? Would it say what it wants and another model/system act on that and 'read aloud' the document to it?
I guess that's somewhat how we do it, so maybe it would work.
I hear you, but as I said, I hope that buying from Deepseek gets more Deepseek. Maybe they get the same amount (or more) from a Canadian provider, and if so I might reconsider.
The 'clean power' would factor in a well, but I haven't really looked in to it. I know China is installing a ton of renewable and has a carbon market while Canada's building a new pipeline and building Alberta data centres running on natural gas, so I'm not willing to take on faith that switching is cleaner.
> The economic sovereignty of any nation, but especially smaller ones like canada and denmark, is harmed by sending our money out of the country when there are options available in country.
I respectfully disagree. I subscribe to the comparative advantage economic principle, so I'm fine with international trade.
I figure (hope) buying from DeepSeek directly benefits them the most so they can release even more improvements and open-weight models.
Plus only one country has threatened to annex mine and it wasn't China.
We (most of the world) also have to deal with well funded external government organisations constantly pumping out propaganda attempting to make sure we don't work together or have empathy/understanding of each other's views.
Push that through a media filter that optimizes for engagement and it's no wonder these things are happening.
If this is the issue you actually care about then you're in the minority. Most seem to be using it as a proxy for complaining about how the LLM uses the content, not that they didn't pay $10 for the content to start with.
Would you be happy if they bought exactly one copy of each book? That might be a dozen sales between the AI labs, some tens of dollars to the author. Is this really what you're so upset over?
And the solution to avoid this piracy has been to buy up rare editions of old books, cut them up and scan them. Better?
In the Anthropic court case it was determined not to, ie that training was transformative.
It's like if you read a plumbing book and then made YouTube videos on how to fix a sink.
Have you run your numbers recently because a 20+ year break-even isn't common? I wonder if you have some highly subsidized power or something because most places would be under 10.
The writing is so AI-y as to make the otherwise interesting content nearly unreadable.
Subscriptions. From some searching I saw a best case if around 600M tokens as a back of a napkin usage on the Anthropic subscription for $200 and at worst 2.6B for glm-5.3 for $120.
The credit system makes it tricky to compare though so I'm curious how you worked them out to be closer. (K3 does seem much closer, I think)
I don't know about kimi, but my napkin-math shows that qwen and glm give you at least an order of magnitude more usage than anthropic, and that's before you factor in any off-peak discounts.
It doesn't for me (although it does include it in parentheses after the real name), so I'm going to hope this is only an 'only in the US' change.
You are framing this as something they have to do: support it, release documentation, open source it, etc.
It's not. It's something they have to stop doing. Stop adding code that specifically stops third party software from running.
> Android devices are unlocked
<sigh> If only that were true. I have a Samsung S20 with 128gb storage and a adequate battery that is ewaste because of the locked bootloader and now permanently vulnerable kernel.
Not disagreeing with your premise (the same should be done with ISPs — own the connection or the wires, but not both). I'd just take the "hard" install with fewer features over what we have now.
Huh, I'd been using this through Sidebery for so long I thought it was a built in feature.
Learning this has also changed my life, but maybe not for the better. Now every time I see someone I know and their shoes are tied in a granny knot I have to waste a bunch of calories deciding if they'd appreciate me telling them.
Have you moved it anywhere else? I checked codeberg and didn't see it there.
Watch the video. They worked around this by selling lego sets to 10 different people (as it was still owned by the lego owner), then the 10 different people all opened separate $10k suits, which they all won.
Then corporate shut down the location to avoid paying the suits they lost.
Not true of non-US companies. I had also added money to Deepseek, and it was still there (and Z.ai and Moonshot are the same). I'm reasonable though, if it's been 5 years or something I might have understood, but it was 1 year and the account was in use during that time.
Where I live (in Canada) it's actually illegal for gift cards to ever expire, and there's lots available from US companies, so if it's an accounting issue other companies have figured it out.
> scamming from the literal money
That's par the course for Anthropic. I added some money to my account before I really had a use case for product. A year later they said my money had expired and when I contacted support they basically told me to pound sand.
This while they have the audacity to list one of their corporate values as 'Be good to our users'. They'll never get another dollar from me.
There's a 5 hour difference between the replies, and new data that came in, so the posts aren't really in conflict.
Also it doesn't sound like they know "there's a model issue", so opening it now would be premature. Maybe they just read it wrong, do better to let a few others verify first, then reopen.
not to use the cli tool. You can install it and change the settings to point to pretty much any other model.
It's an okay-enough tool, but I don't see a lot of point in using it when open sources tools like Pi and OpenCode exist (or octofriend, or forge, or droid, etc).
@eieio: whatever email protection you're running is triggering on the extension info. For example I see:
> And they’re sent to servers that advertise the availability of the [email protected] extension. What if we just…don’t advertise [email protected]?
If you avoid Chinese-branded devices you should doubly avoid US-branded devices. See National Security Letters and Room 641A as a small subset of why.