22
submitted 11 months ago by Maebbie@lemmy.world to c/videos@lemmy.world

You know how sometimes you see an intriguing title and just want to know the answer, only for the youtuber to drag it out for 5 minutes until he actual answers the thing.

Not anymore with this prompt, I used it so much that I end up being less on youtube. This method will optimize 10 minutes of inefficiently placed information down to its core of a few sentences.

you are viewing a single comment's thread
view the rest of the comments
[-] CurlyMoustache@lemmy.world 20 points 11 months ago

Yeah, I'm not watching this two minute video. Humans of the Internet, what is the prompt?

[-] littlebluespark@lemmy.world 12 points 11 months ago

I'm dead. 🤣 This species is so fucked. 🤌🏼

[-] Maebbie@lemmy.world 9 points 11 months ago* (last edited 11 months ago)

lmao, based. The video is designed to show the method to people who never used a LM before in their life, just in case. The prompt is (its also in the description of the video):

Please read the following transcript of a youtube video and answer this: "(insert title between these quotation marks)"

This is the transcript of the video:
(insert transcript here)

make sure to turn off timestamps in the transcript before copying it.

[-] otter@lemmy.ca 3 points 11 months ago

Watched it

It's not a specific prompt, it's about how to get the transcript through the UI so you can paste it in for the summary. Most of the video is about other things

I personally prefer using the highlights in SponsorBlock (which to be fair this person recommendeds as well)

[-] Maebbie@lemmy.world 1 points 11 months ago

I'd just like to note that you should indeed use the specific prompt in the description, or in my other reply here, to get a proper response from the LM. The prompt is designed to have the LM answer based only on the transcript and with a hard focus on the title, that made you click the video in the first place. That way your curiosity is sated and you can move on without having to deal with 10 minutes of bloat.

It takes the power of an LM being able to process language and the information assembled by a human. That way the LM cant drift into ambiguity or guessing.

[-] otter@lemmy.ca 2 points 11 months ago

Oh that's a fair point thank you :)

Now all we need is a way to drop the thumbnail in and we'd be golden on sating the curiosity lol

this post was submitted on 18 Jan 2024
22 points (100.0% liked)

Videos

14417 readers
261 users here now

For sharing interesting videos from around the Web!

Rules

  1. Videos only
  2. Follow the global Mastodon.World rules and the Lemmy.World TOS while posting and commenting.
  3. Don't be a jerk
  4. No advertising
  5. No political videos, post those to !politicalvideos@lemmy.world instead.
  6. Avoid clickbait titles. (Tip: Use dearrow)
  7. Link directly to the video source and not for example an embedded video in an article or tracked sharing link.
  8. Duplicate posts may be removed

Note: bans may apply to both !videos@lemmy.world and !politicalvideos@lemmy.world

founded 2 years ago
MODERATORS