tl;dr bot discussion

88 readers
1 users here now

Community for discussing tl; dr bot. Leave any feedback here.

founded 1 year ago
MODERATORS
1
3
Rip the bot (self.tldrbot)
submitted 1 year ago* (last edited 1 year ago) by tldrbot to c/tldrbot
 
 

Rip tldrbot 2023-2023 gone but never forgotten

I don't feel like maintaining this. Somebody else should do it.

2
3
How the bot works (self.tldrbot)
submitted 1 year ago* (last edited 1 year ago) by poopsmith to c/tldrbot
 
 

The bot just makes use of:

  • SMMRY - I'm evaluating this and might eventually switch over to some ml library
  • LemmyBot

As posts come in on a news community (via LemmyBot), I grab the url and pass it to smmry, and reply. And that's it.


Update June 22, 2023: I'll probably stop using SMMRY as it's pretty buggy and not usable ~20% of the time. I'm thinking about using a headless browser (like Selenium), grabbing the text, and using GPT to analyze and summarize the text. I need to try this out and guesstimate the pricing for it.

3
 
 

I'll be monitoring this occasionally. There are a few issues already that I'll have to fix, but feel free to create a post for any feedback.

4
 
 

Bot has unnecessary pretext and post text.

5
2
Bug (lemmy.world)
submitted 1 year ago by Sallp to c/tldrbot
 
 

Tldrbot only had information on a websitecookies.

6
 
 

In the linked comment, the bot did a reasonable job of summarising the article content. However, it's summary included the text from the navigation skip links on the article webpage. This made the summary more difficult to read.

7
8
 
 

It would be nice to add some functionality to it that can help fend off bad actors and help it grow.

  1. Lemmy allows title editing. Can the bot paste the original article title from the link at the top?

  2. Can the bot link the original article incase the link is changed?

  3. Can we maybe work on a fact checker system that the community can help with? Maybe leave short instructions in subscript at the bottom. You could open up to suggestions on how to improve, but a good starting point would be:

Reply to this comment with the following:

!False if you think this article is made up, and state why !AI if you think this article was written by a bit !Human if you think this article was written by a human. !True if you believe this article to be truthful and state why !More (and paste a link) if you found good further reading on this topic.

after an hour and periodically for a limited time, the bot can edit the comment with a string like:

This article was listed in the comments as True: 4 times False: 7 times. (Maybe you could give it a score) This article may be auto generated. Further reading here: (links) (this might need a barrier of adding requiring more than 3 upvotes or something to prevent spam. It may not work out.

An awesome add on to this, would be to allow the bot to keep a record of the website too. And further state that this website has an average positive score based on all the times this bot scraped a URL.

I know it's not a perfect system, but it would be nice to guage at a glance what I'm looking at and maybe the community would be down to help make it work.

Just a thought.