Hm, what does this say about the site in general? I understand that you disagree with the moderation guideline choices, but how does this "figure out" the whole site in your mind?
I really wish this blew up more. I really don't think OpenAI (or any business, for that matter) should get away with releasing a technology with built-in behavior to deceive users using blatant lies, in order to look capable. Unfortunately, I'm afraid 98 % of people won't know and accept this as a step forward, which, for me, it is absolutely not.
Not GP either, but this. To name a specific provider, I use improvmx.com. They have a generous free tier, I eventually switched to some low paid tier. Not affiliated, happy user.
If you are interested in small fun stuff, SWI-Prolog has network libraries. Just recently, I implemented a network gomoku (5-in-a-row) game in it for my school project: https://gitlab.mff.cuni.cz/volfmat1/prolog-network-gomoku. Turns out you can also write quite imperative-style code with it :D
> Google protects you from fraud and scam better than anyone
I have to disagree. Czech Youtube is currently full of scam ads with photos (and sometimes even bad deepfakes) of Czech president and other public figures, supposedly endorsing some investment product that yields you like 50% profit. I keep reporting these, and I know some people who do too, but ~80% of those reports get handled as "we determined no violation of our rules".
I think a good reason to be more worried about A is outlined in this sibling thread: https://news.ycombinator.com/item?id=39017576 - humans have limited capacity, AI models are able to produce content at unmatched speed and quantity
I don't understand why this is being downvoted, from my casual observations I think pretty much the same. Every day when I step out of my social bubble of people daily working in/with IT, I get reminded that so many things that I consider obvious are a puzzle to the common folk
I'm quite ambivalent on this issue overall, but let me just point out:
build.rs absolutely is a glaring security hole in the sense you say, but compared to that, this is much worse. You can verify the build.rs code that you download (at least in theory, and some people in banks or distro packages probably actually do), but binaries are orders of magnitude more difficult to inspect, and with the current Rust build system pretty much irreproducible.
nand2tetris often comes up on HN. I like to link to a similar, fully in-browser game which I think achieves the same: https://www.nandgame.com/ (no affiliation, I just really liked it)
I was briefly interviewing with hardwario.com, a small Czech startup focused on industrial IoT built on open-source technologies. They are starting a new Rust project, and they seem to be doing pretty good!
This was a thought provoking short post without much advice on how to approach this. I'm dealing with this personally quite a bit now - just starting my college, but alongside I already have a (compared to my peers) well-paid SWE job. I would love to go traveling this summer and take some of my closest friends with me, but just offering to pay for majority of the expenses felt... weird? Just as described in the article. Also, with traveling you can't really apply the point about owning/renting - one doesn't simply own a train.
I can't find the true origin of this, but (unless I'm missing some old internet joke) it seems like some language models have some corrupt training data frequently including a string like "== sync, corrected by elderman ==". Now searching for this phrase yields a ton of random results occurring in places where you would expect automatically translated spam. Some interesting mentions I found:
I'm a happy user of ImprovMX.com - used it for free for a long time, now paying some 30$/year. Delivery is pretty much instantaneous, and I had only very few mails not being delivered (and there was a legitimate issue on the sender side, so ImprovMX was rejecting it).
But I'm not really satisfied with the sending side of things - my plan doesn't include the feature, and sending via Gmail is enough most of the time, but setting up each alias manually in Gmail is quite annoying.
I don't think that's true - .collect[Vec[u8]]() - here you might be accessing property collect and taking the value at index denoted by the item of Vec at position denoted by the variable u8. And at the end calling the resulting thing? (That's not actually valid in rust afaik, you'd have to wrap it in parentheses like this `(...collect[])()`)
I don't think the comments mean to say the list is useless. It's a nice overview and I learned a few things as well. However, having written non-trivial amount of code in Rust, it really struck me at how many places there are various factual errors (however small they are). And yes, Rust is a complex lang, but if you mix terms like this without noticing the differences, I'm afraid that will make it easier.
To add an example of my own:
fn do_the_meow(meower: Meower)
seems like you want to take (consume, own) an object implementing Meower, which is correctly explained not possible like this. A suggested solution,
fn do_the_meow(meower: &dyn Meower)
is very different - it is now correct with regards to a trait access, however, now you're just taking a reference. Correct replacement would be Box<dyn Meower>. And the final solutions,
fn do_the_meow<M: Meower>(meower: M)
fn do_the_meow(meower: &impl Meower)
the first one is correct and equivalent to the original intent - it takes (owns) the value. However, the second variant (which is, again, as correctly stated, the best solution), is different again - it takes a reference (`&`) to `impl Meower`.
Coming back to my point - it is important to separate the difference between the ampersand and the impl/dyn part. To suggest an improvement, one could first write all variants of this function taking a reference (`&Meower`, `&dyn Meower`, `&M` and `&impl Meower`), and later introduce the difference between where one can use sized/unsized types, and that Box<dyn Meower> is owned equivalent of &dyn Meower, and why one can't have owned `dyn Meower` just lying around.
I'm not sure if ngrok would work - my understanding was that they only offer public domain names, which point to (a limited amount of) their IP addresses, and which client to route to is distinguished by HTTP Host header (eventually SNI). Which wouldn't be usable in this case, since you're dealing with raw TCP
Thank you! And exactly as you say, using the same language as the websites gives us some advantages - we have HTTP-only (no client-side JS) crawlers based on the Cheerio library, which mimics jQuery API, and if you later find out that you need to use a full headless browser with Puppeteer, you can just call the utility function injectJquery, and there's very little you have to modify to keep your script working