ascorbic

6 ms·

241.
by ascorbic 1y ago
This doesn't seem much different from Claude's MCP implementation, except it has a lot more warnings and caveats. I haven't managed to actually persuade it to use a tool, so that's one way of making it safe I suppose.
242.
by ascorbic 1y ago
Slack groups the different variants, with counts for the total.
243.
by ascorbic 1y ago
There are generic versions of all of them. All emojis have a base version without skin tone or gender applied. These are mostly displayed with yellow skin and a vaguely gender-neutral appearance. They're combined with modifiers to crea
244.
by ascorbic 1y ago
They were written in French.
245.
by ascorbic 1y ago
> Please don't post shallow dismissals, especially of other people's work.
246.
by ascorbic 1y ago
I wonder how many of the people saying this about copyright infringement today were complaining about the ridiculously harsh enforcement of it 10-15 years ago. They've just paid $1.5B for torrenting.
247.
by ascorbic 1y ago
I think the important part is that the remote person doesn't need to be alert, and make real time decisions within seconds. As I understand it, the remote driver is usually making decisions with the car stationary. I'd imagine tha
248.
by ascorbic 1y ago
And their H100s were diverted to build MechaHitler instead
249.
by ascorbic 1y ago
Why is it particularly bad in Australia? Is it simply that it's the whitest country at that kind of latitude?
250.
by ascorbic 1y ago
It's not the crux of this case. This is a settlement based on the judge's ruling that they books had been illegally downloaded. The same judge said that the training itself was not the problem – it was downloading the pirated book
251.
by ascorbic 1y ago
The settlement was for downloading the pirated books, not training from them. Unless they're paywalled it would be hard to argue the same for a blog.
252.
by ascorbic 1y ago
They're paying $3000 per book. It would've been a lot cheaper to buy the books (which is what they actually did end up doing too).
253.
by ascorbic 1y ago
And the moral of that story is to buy a three pack of Kuhn Rikon peelers.
254.
by ascorbic 1y ago
I was thinking the other day about which websites might still exist in 100 years on whatever the web becomes, and Wikipedia to me seemed the most likely. It manages to be both stable and flexible. Functionally it's almost identical in
255.
by ascorbic 1y ago
Voices, not languages. The "English" one is American though.
256.
by ascorbic 1y ago
And most recently "Golden"
257.
by ascorbic 1y ago
I think it's because it was using the American voice for it. Conversely the female voice in the Mandarin conversation spoke English with a Chinese accent.
258.
by ascorbic 1y ago
Not tried with Zed yet, but in VS Code it's mostly the same except you can view the diffs in the editor, and can give context from the open file. I think it can also see errors from the editor.
259.
by ascorbic 1y ago
According to that link .george is just a proposed TLD. Presumably it's not proposed anymore, because Walmart no longer owns it.
260.
by ascorbic 1y ago
It was introduced to Walmart in 1999 when they bought Asda: https://en.wikipedia.org/wiki/George_(fashion_label)
261.
by ascorbic 1y ago
They're very easy to type on a Mac though (opt+-). I've always used spaced en dashes without realising that that is the more common British style. Unspaced em dashes just look wrong to me.
262.
by ascorbic 1y ago
The point is that the providers aren't burning money by subsidising inference costs. On the contrary, if this article is to believed they're charging healthy margins on it. So there are two answers: for the model providers, it&#
263.
by ascorbic 1y ago
This article is about the model providers' costs, not API users'. Cursor etc have to pay the marked-up inference costs, so it's not surprising they can't make a profit.
264.
by ascorbic 1y ago
Because they're spending it all on training the next model.
265.
by ascorbic 1y ago
You can amortise the training cost across billions of inference requests though. It's the marginal cost for inference that's most interesting here.
266.
by ascorbic 1y ago
I didn't need any tricks. I just asked it in the context of understanding the paper rather than asking for my own use. https://claude.ai/share/bf2a9d7a-bedf-4fbc-af2f-3a6b72f66753
267.
by ascorbic 1y ago
I like that the server includes a feedback_to_devs tool. I tried the flights tool with Claude and it was pretty buggy and limited in what it could do. I asked Claude to give feedback to the devs and it gave a decent bug report and feedback
268.
by ascorbic 1y ago
Snark aside, MCP uses JSON-RPC for messages
269.
by ascorbic 1y ago
Ahem https://www.microsoft.com/en-us/edge/mac
270.
by ascorbic 1y ago
I'm not sure how you can blame counselors when no counselor would have said any of the things that were a problem here. The issue here wasn't that there was examples of counselors in the training data giving practical instructio

More ›