mashdragon

> For styling you can use this for now https://tailwindcss.com/docs/typography-plugin Thank you, I'll look into it!

@notmd Thank you for your suggestion. Now, those things render much better! But there are now really big margins, and dark mode does not work. For dark mode, I tried...

Keep aware that the Review step will have at least temporary meaning for reviewing Markdown formatting once #1350 / #837 is completed. Eventually with a full fledged Markdown editor we...

> How about asking ChatGPT directly whether it wrote the text or not: "Consider the following text and provide only a percent value describing the likelihood that you wrote the...

@notmd thank you for solving this. The only test case that fails is ordered lists (1., 2., etc.) are not rendering the numbers. The line spacing is also really big...

One other place that needs rendering is in the modals of the ranking task. I did not remember that before.

> If no specific programming language is given after the ```, could the language automatically be detected? Unlike spelling corrections, syntax highlighting should never affect the conversation, so automating this...

I would prefer it if *all* prompts were complex, actually. If Open Assistant can't answer anything difficult, it's not going to be very helpful. The people asking the complex questions...

A few thoughts: * Time spent looking at text, even for thoughtful reviewers, is not necessarily proportional to how much text there is. Let's say the task is formatting or...

Many of these ideas for solutions (restricting accounts or activity per IP address, tracking browser metrics such as the user agent and punishing users who deviate from some standard, using...