cdrini

54 ms·

331.
by cdrini 4y ago
You're free to speculate, but I would respectfully disagree. Extrapolating from how effective GPT3/Copilot are at interpreting text and generating code, I think it has more of a chance of working well than any other tech has ever
332.
by cdrini 4y ago
What you're describing is more like dictation. What you'd probably say is "export the button component", and it would determine the syntax.
333.
by cdrini 4y ago
Main things that have changed are: 1) Improvements in speech to text, as others have mentioned 2) Improvements in language models (and model size) allowing for more flexible interpretation of speech. This isn't dictation anymore. It&#x
334.
by cdrini 4y ago
This looks to be much more heavily using GPT3/Codex/Copilot, which I've found to be eerily effective. It basically feels like a voice interface to Copilot. The main difference between these and something like Google Home is h
335.
by cdrini 4y ago
I can't speak for others, but I've personally been quite impressed by the dalle output. It creates things that would take me hours (if not days) to create, which no other tech I've tried has been able to generate. It feels li
336.
by cdrini 4y ago
I'm always surprised to see posts like this, and no one mentioning testcafe. It's a super solid e2e testing library! I've been using it for quite a while and it runs reliably and consistently. The API design can be a little m
337.
by cdrini 4y ago
I wasn't really trying to address whether the argument is valid, I was just noting the representation of the other side here is reductive to the point of being in bad faith. I find that kind of rhetoric a little frustrating since it&#x
338.
by cdrini 4y ago
I haven't heard anyone saying that copilot is legal "just because it's AI." That's a pretty bad faith, reductive, and disingenuous representation. The core argument I've seen is that the output is sufficiently
339.
by cdrini 4y ago
Spreadsheets are used for _waaaay_ more than just finances. I don't think it's anywhere near 50% finances. I can't recall where, but I saw a study from I think the 90s saying most of the spreadsheets they found were being us
340.
by cdrini 4y ago
It could generate child pornography.
341.
by cdrini 4y ago
I think the internet is in a bit of a risky place right now. People don't know what to trust online, and scammers and manipulators are taking advantage of it. People are believing wild conspiracy theories and finding ample "eviden
342.
by cdrini 4y ago
Correct, it doesn't hold up generally. But it doesn't need to. It holds up here. We do not try things when there is exceptional risk or cost in the trying. Here there is no cost to trying, so it does not make sense not to try. I b
343.
by cdrini 4y ago
I believe the argument being made is that in _actual, real-world_ use of copilot, no copyright infringement happens. In order to make an informed decision as to whether you agree with that, you can try copilot to reach an informed conclusio
344.
by cdrini 4y ago
I believe the argument is that you shouldn't accept arguments against the use of copilot from people unless they have tried to use it. In a realistic context. That seems reasonable to me. It's the bare minimum to make an informe
345.
by cdrini 4y ago
Iconify? I love this tool! https://iconify.design/
346.
by cdrini 4y ago
Fun fact: not only is the iteration order of sets in python unordered, it's also undeterministic! A set of the same values, iterated in two python program, could iterate in a different order. I learned this when a unit test that iterat
347.
by cdrini 4y ago
That's a super interesting observation; you've convinced me that ML algorithms could generate novel things.
348.
by cdrini 4y ago
I think the biggest misconception with a lot of the comments here is that they are assuming that the code generated by copilot is final. Copilot, as the name suggests, is a human-in-the-loop system. It requires a developer to give it prom
349.
by cdrini 4y ago
I would argue a programmer who doesn't notice the encoding issue when looking at this code would be the same kind of programmer who would write this kind of encoding issue. You definitely need to manage copilot's generated code to
350.
by cdrini 4y ago
Unfortunately I couldn't find a good package either, so ended up rolling my own script, FnToCLI https://github.com/internetarchive/openlibrary/blob/ccf10246... . It wasn't too hard to get the bones
351.
by cdrini 4y ago
Oh wow! You are correct, I am very interested in it! Thanks for sharing! Honestly on most of my personal projects, I only use typescript via the jsdoc interface. I just never felt the need to get to 100% type safety for most projects. And I
352.
by cdrini 4y ago
For me the remaining pieces for Python are much more attainable than the missing pieces for typescript. I think basically all the helper TS things (eg KeyOf, etc) are addable to Python. Runtime types are not addable to TS. It goes against s
353.
by cdrini 4y ago
I love that Python has types available at runtime! The ability to have those type-generated CLIs or APIs is fantastic. I learned from this article about functools.singledispatch, and can't wait to use it. I've always wanted someth
354.
by cdrini 4y ago
Just as a note that when the parent comment was posted, the comment section looked very different. It was almost all vitriol. It was genuinely discouraging and disheartening to read, and ugly was the correct adjective. I really resonated se
355.
by cdrini 4y ago
Hmm I'm not sure I understand. Here's what I'm assuming: You currently have some number of org repos; eg: my-project/my-project, foobar/foobar And you have your personal account: buro9. You use your personal account
356.
by cdrini 4y ago
I'm not entirely sure, but I think if your personal account is marked as a maintainer for the repo, it doesn't matter if the repo is under its own org? I feel like most projects create org accounts, so would be kinda surprised if
357.
by cdrini 4y ago
Oh as of this week copilot is open to the public: https://github.com/features/copilot/ . You can get a trial, or it is free for students and open source maintainers.
358.
by cdrini 4y ago
Oh yeah those numbers were hogwash, I was just looking for a number less than 3 :P I was thinking like small mom and pop shop that has like a small Dev team for some reason might be able to reduce it down. Vox hosted a fantastic discussion
359.
by cdrini 4y ago
The parent comment seemed to me to be implying that humans aren't capable of copying. In my experience I've had more issues with human developers plagiarising than with copilot plagiarising. You could create such a system. I agree
360.
by cdrini 4y ago
I definitely agree copilot isn't understanding the problem space of your code. But I also don't believe it's simply remixing code samples from its training set. It's somewhere in between. I don't know the internals,

More ›