dminik

8 ms·

151.
by dminik 5mo ago
The answer is that C (and by extension Zig, C++) code goes through a hardening process. New code in these languages tends to be unsafe. But bugs and vulnerabilities get squashed over time. Bun gets updated fast and so has a lot of new unsaf
152.
by dminik 5mo ago
Yeah. Context size matters a lot. With OpenCode dumping like 10k tokens in the system prompt it takes like 4 rounds before it had to compact at say 64k. It's not really worth it to run at anything below 100k and even then the models ar
153.
by dminik 5mo ago
Because they have their own drawbacks. To make them really useful, you need a resizable stack. Something that's a no-go for a runtime-less language like Rust. You may also need to setup a large stack frame for each C FFI call.
154.
by dminik 5mo ago
And where does one store and version this harness?
155.
by dminik 5mo ago
This was posted before and it seems that Firefox randomizes the extension URLs.
156.
by dminik 5mo ago
I mean, how do we qualify which companies get punished for which crimes? Do we punish gun manufacturers for someone being shot? Kitchen utensil companies for someone being stabbed? Car manufacturers for car crashes? Road construction compan
157.
by dminik 6mo ago
Is Anthropic speedrunning their fall from grace? Their "stand" against the US government, but not really, happened roughly two months ago. Yet they've been doing something stupid every week since. Who is running this company?
158.
by dminik 6mo ago
I've recently built a script that periodically (every 25 minutes) fetches the latest merged PRs to check for some potential rule violations. I'm not an admin and couldn't get the events API working, so I just resorted to poll
159.
by dminik 6mo ago
It feels a bit like a "better is better" language hitting all of the quirks of a "worse is better" environment.
160.
by dminik 6mo ago
I feel like this is pretty pointless. Rather than trying to convince the model to do all of this, why not just run the tests automatically?
161.
by dminik 6mo ago
I mean, not one "bleeding edge" lab has stated they are profitable. They don't publish financials aside from revenue. And in Anthropic's case, they fuck with pricing every week. Clearly something is wrong here.
162.
by dminik 6mo ago
I mean, the financials just don't look great either way. Their main product is part VSCode, which is a market that's almost impossible to make money in, and part reselling already expensive LLM tokens. You can look at more paramet
163.
by dminik 6mo ago
Each tool call requires a new request no? The harness receives a response, has to parse out the tool call, execute it and then start a new request with the tool call result.
164.
by dminik 6mo ago
Even more than that possibly. I don't imagine cursor has any deals in place with Anthropic. They likely pay API pricing like everyone else.
165.
by dminik 6mo ago
I mean, that still feels overpriced. Biden won in 2020 with just ~$2 billion. Though he didn't get all the branches. Neither did Obama, who won two terms for <$2 billion. Are the house and courts really so expensive?
166.
by dminik 6mo ago
If Whatsapp is burning through say ~$1B yearly with zero revenue and Cursor is burning through say ~$2B with a ~$1B revenue, they're both still in the hole. I wish people would stop talking about just revenue. It's mostly meaningl
167.
by dminik 6mo ago
Yes. But unlike cursor, Whatsapp had the following advantages: 1. It's cheap to run. 2. It has clear advantages over existing technology (SMS). 3. My mom uses it. She's never going to use cursor. Whatsapp had a huge userbase in Eu
168.
by dminik 6mo ago
Revenue without expenses is meaningless. Annualized revenue is even worse. It's like a gambler bragging that they spin through $20,000 a month. Yeah, but for how long? If you give me a billion, I can do an annualized revenue run rate o
169.
by dminik 6mo ago
Wow. Tech CEOs and investors have completely lost touch with what money really is worth. How is a VSCode fork and a open weight LLM fine-tune worth $60B? One would think Elon would learn his lesson after overpaying for Twitter and then havi
170.
by dminik 6mo ago
Well, if there's one way to show that you're not profitable on inference, this would be it.
171.
by dminik 6mo ago
Interesting. So the 50%+ number that's been floating about isn't even real. It's just an example.
172.
by dminik 6mo ago
What "advanced features" are there to speak of in C? What does the syntax of C being similar to JS matter? This comment makes no sense.
173.
by dminik 6mo ago
I'm technically waiting until my vision stabilizes after an operation to get contacts. So it's not quite that bad yet. I've read that other people prefer dark themes, but for me that doesn't quite work. Something about t
174.
by dminik 6mo ago
It's funny, but I have keratoconus and at this point I can't read white text on a dark background. If an app doesn't have a light theme, it's very hard for me to use.
175.
by dminik 6mo ago
Just plugging in a networking library isn't going to get you a what you need for a CS like game. The players here expect perfection. You will also need to figure out client/server side prediction, rollback/reconciliation, lag
176.
by dminik 6mo ago
If I owned two lines both selling the same thing (preumably here Coke is a stand-in for compute), I would throttle the $2 dollar line. People without a choice might move to the $3 dollar line. Unfortunately, back in the real world, Anthropi
177.
by dminik 6mo ago
If they had such a high margin, they wouldn't need to fuck around with token usage/pricing every three days. I have no data to support this, but I think they just about break even on API usage and take overall loss on subscription
178.
by dminik 6mo ago
Anthropic intruduces fake tool calls to prevent distillation of their models. Others still distill. Anthropic distils third party models. Claude now hallucinates tools. Brilliant.
179.
by dminik 6mo ago
Tbf I don't think that it's just this one reason. While I'm not a subscriber to any LLM provider, the general feeling I get from reading comments online is that the models have a long history of getting worse over time. Of co
180.
by dminik 6mo ago
Why do stores increase prices before a sale?

More ›