Yes, so I'll disclose it: I was using Opus 5.5, on medium effort, in Claude desktop, which has full access to my entire repo.
Now everyone can officially lambast me for "using the model wrong or using the wrong model," exactly as you say. But I find it quite interesting that one of the commenters here assumed I was using Opus 4.6, because these mistakes sound like that old version! I'm using the version released just four freaking days ago!
I expect some commenters will now say, "Oh, you should have been using Fable, you old boomer." To them I say: "Yeah, well my employer doesn't allow me to use Fable." And, in jest: "Now get off my lawn."
Just today it made three glaring mistakes in one session:
1. It read a file in the wrong directory, because that file had the same name as the file in the right directory. It apologized when I challenged it, promising me that it would remember to "read import statements" in the future.
2. It miscounted the number of times a function was called in my repo. It said 20, while my built-in IDE search accurately showed 17. Again, it apologized when I corrected it.
3. It referred to a variable by name that does not exist anywhere in my code. It apologized, and said it was referring to a variable used internally by one of the third-party packages installed in my repo.
So many apologies.
It's the little things like this that remind me on a regular basis just how little I can trust artificial "intelligence."
https://status.gitlab.com/pages/history/5b36dc6502d06804c083...
I couldn't disagree more. "A problem well stated is a problem half solved."
— James 1:8 (KJV)
1. When I right click, I can't create a new file of any kind. I can create a new folder—sometimes (see the next point).
2. If the folder is full, it's virtually impossible to right-click an empty spot to create a new folder. You have to scroll all the way to the top/bottom and right click the very narrow spot between the folder and the header/footer.
Was the video generated by AI?
1. qualified for by right according to law
2. having a right to certain benefits or privileges
3. having a right or legitimate claim to something
No company is qualified for sales according to law, or has a right to sales, or has a legitimate claim on sales. Sales must be earned by besting the competition.
No company is "entitled" to any sales.
No company is "entitled" to any sales.
1. The v0id_isgood HN user impersonated the actual creator of Wyzer and posted this Show HN.
2. The v0id_isgood HN user then left the credentials to his HN account in the Wyzer discord.
3. Now the real creator of Wyzer is apparently using the v0id_isgood account to answer questions here?!
What's going on? This whole thing feels like a really strange hoax.
Apparently, it takes a lot of time and work to manage all those AI chat bots.
Every single message from him, literally every time he communicates, it's a massive wall of text, overflowing with scope-creep suggestions like nothing I've ever seen before. It's impossible to ask him a simple question and get a simple answer.
And this seems to be a broader trend with AI in general. It's getting more and more difficult to retain some semblance of brevity with AI tools. I'm constantly asking them to "be brief," and "only answer the question I asked," and "don't provide more information than I requested," but they keep churning out more than I want in their responses.
It looks like a subtle attempt to use up more tokens.
The first game I'm building is the card game Phase 10, and I'm done with phases 1-7. After that, I'd like to build Carcassonne, and maybe Jeopardy.
I've created a new issue for this (it's similar to an idea I already had):
(1) When creating the textile, you would add a step to "wait for new clipboard data."
(2) When you run the textile, and it gets to that step, a modal would appear, prompting you to copy the new thing to your clipboard.
(3) Once you've got the new thing on your clipboard, you would go back and click "Continue" on the modal from step 2.
(4) Textile would continue with the new thing on your clipboard.
Is that kind of what you were thinking?
Here's a quick example, and one that I often use with Textile to generate a preview URL based on my current branch:
(1) Start with the output of the command `git branch --show-current` in the `~/code` directory (yielding text like `JIRA-1234/some-feature`).
(2) Replace all `/` characters with `-` (now the text is `JIRA-1234-some-feature`).
(3) Prepend `[preview](https://staging-`, which is the start of a markdown link (now the text is `[preview](https://staging-JIRA-1234-some-feature`).
(4) Append `.example.com)`, which is the end of the markdown link (now the text is `[preview](https://staging-JIRA-1234-some-feature.example.com)`).
(5) Copy the result to my clipboard.
With those steps saved in Textile, I can now click a button to run them over and over again (or use a keyboard shortcut if I assigned one). So no matter which branch I'm on, I'll always get a proper preview link without having to construct it manually myself.
Hi all,I'm excited to show off Textile, a desktop app I recently built.Textile can combine bits of text using various inputs, such as commands on your computer, the contents of your clipboard, and hard-coded strings that you provide. It lets you carefully build up and modify a dynamic string, step by step, until it's exactly how you need it. The saved steps can then be executed on demand, with...
It's a great book, although it's forcing me to admit that I need glasses (the print is so small)!