Become a Prompt Engineer, #1 Trick to 10X Your Prompts

Поделиться
HTML-код
  • Опубликовано: 25 ноя 2024

Комментарии • 21

  • @MJFUYT
    @MJFUYT 11 дней назад +1

    Absolutely love this channel! You have a talent for simplifying complex concepts into easy-to-follow instructions. Keep up the fantastic work!
    Like & subscribed

  • @59Marcel
    @59Marcel 16 дней назад +2

    What a simple yet brilliantly clever idea to use web pages to help create prompts especially the Wikipedia site. This will come in very useful. Thank you.

  • @ADancingFlamingo
    @ADancingFlamingo 12 дней назад +1

    This video is fantastic. You are very good at explaining how to create prompts like a pro. Thank you so much, you helped me a lot.

  • @T-Bone54
    @T-Bone54 20 дней назад +3

    Excellent usage of LLM to make AI images and storylines!

  • @bazz1376
    @bazz1376 19 дней назад +2

    Nice. Thanks

  • @Coollsmalls4kTechArtist
    @Coollsmalls4kTechArtist 20 дней назад +2

    Love ❤️ it my brother

  • @wheelsforfeet
    @wheelsforfeet 13 дней назад +1

    what do you use for your prompt database? i started something in notion but im looking for suggestions

    • @metricsmule
      @metricsmule  13 дней назад

      Airtable. I’ve tried everything, including Notion but to me Airtable is best by far. I’m making a video soon demonstrating how I made it so others can do the same.

    • @wheelsforfeet
      @wheelsforfeet 11 дней назад

      @ thanks. Looking forward to the video

  • @PEN32
    @PEN32 14 дней назад +1

    Paywall.

    • @metricsmule
      @metricsmule  14 дней назад

      Nope. Link is in the description to get this prompt, plus others like it, in addition to thousands of other prompts on my website. No paywall for any prompts I share in my videos.

  • @iXzenoS
    @iXzenoS 20 дней назад +1

    I enjoy most of your content but all you're doing is keyword-stuffing the AI generator and calling that "10x". E.g. 4:25 with the car - a similar quality image can easily be achieved with a generic one-liner like "Futuristic cyberpunk sports car driving through the neon-lit night in New York city."
    For your prompt technique to be credible and believable, you need to do a more in-depth analysis, with a before/after comparison, of WHY and HOW your prompt suggestion is "10x"ing the image IMO.

    • @metricsmule
      @metricsmule  20 дней назад +1

      Good point with the before and after. The results are much better when using specific cameras, lenses/settings. Showing the before/after will be beneficial to see the difference. Just adding "Canon EOS 5" in a prompt makes a difference. Have you tried the prompt generator in this video yet? If so, LMK how it worked out. If I tell it "a prompt that generates professional, editorial photos with high end cameras/lenses" the images have been impressive.

    • @iXzenoS
      @iXzenoS 20 дней назад

      @@metricsmule Yup I agree, adding technical details like lens, aperture, style, etc. does seem to make a difference. I was referring more to having LLMs summarize a page, convert that into a descriptive prompt without much technical detail, which usually creates a keyword-stuffed prompt that doesn't seem to make much of a difference. I think your tutorials would be better if you did a before/after showing how the images (using a base prompt) can be drastically improved using your techniques.

    • @metricsmule
      @metricsmule  20 дней назад +1

      @@iXzenoS Good points. I'll do that in my next one and for some AI videos I've been working on.

  • @ScandalousTales24
    @ScandalousTales24 19 дней назад

    Man I just found something out I think that could help us out a lot do you have a email or dm I could message you???

    • @metricsmule
      @metricsmule  14 дней назад

      Yes, email is in description.