Omni Prompting with gpt-4o-mini | A Staple In The Future of AI Software?

Поділитися
Вставка
  • Опубліковано 2 лис 2024

КОМЕНТАРІ •

  • @Ms.Robot.
    @Ms.Robot. 3 місяці тому +1

    Another sweet ride through the Midnight Ai Matrix.❤🎉 soo good.

  • @AIPulse118
    @AIPulse118 3 місяці тому

    current research says that you are correct. longer context outperforms fine-tuning quite often.

  • @NightSpyderTech
    @NightSpyderTech 3 місяці тому +2

    It would be great to see you make tools like these into stand alone apps. I love all the cool apps you make, but having them reliant on the browser ads additional steps to getting them launched and is rough for ppl like me with way too many tabs already open in the browser.

    • @AllAboutAI
      @AllAboutAI  3 місяці тому

      Thnx for the feedback. Might look into that :)

  • @lilan1
    @lilan1 3 місяці тому +2

    unsure if this has been answered previously. however, why do you use colon ': ' opposed to full stop ' . ' at the end of each prompt? do you find any difference in output?

    • @AllAboutAI
      @AllAboutAI  3 місяці тому +2

      tbh its a habit, dont think it makes a big diff :)

    • @lilan1
      @lilan1 3 місяці тому +1

      @@AllAboutAI thanks

  • @j0hnc0nn0r-sec
    @j0hnc0nn0r-sec 3 місяці тому +1

    This is absolutely fantastic 👏👏👏

  • @simonpeter9617
    @simonpeter9617 3 місяці тому +1

    super interesting! nice work !

  • @FusionX690
    @FusionX690 3 місяці тому +1

    Amazing job 👍

  • @thenoblerot
    @thenoblerot 3 місяці тому +2

    4o mini is the same price *per image* as 4o, fyi. Mini price only applies to text.

    • @AllAboutAI
      @AllAboutAI  3 місяці тому +1

      Yeah I know, but i still thinks its not too bad. hopefully cheaper going forward :)

  • @24-7gpts
    @24-7gpts 3 місяці тому +1

    Super amazing!

  • @Heisenberg2097
    @Heisenberg2097 3 місяці тому

    NO WAY. I wait for Super-Omni-Prompting Before I waste my time. The current rate of new buzzwords and terms per day is maybe the best about AI.

  • @JNET_Reloaded
    @JNET_Reloaded 3 місяці тому +2

    its not cheap isit its 20 to even get tokens so i stick to local llm ollama thats it! never paying for anything!

    • @24-7gpts
      @24-7gpts 3 місяці тому +1

      don't be sassy about it

    • @ai.universityy
      @ai.universityy 3 місяці тому

      Finally someone who actually know the future of ai software, no matter how "cheap" it's gonna get, llama already kicks its ass, and will do more kicking in the future

    • @24-7gpts
      @24-7gpts 3 місяці тому

      @@ai.universityy Yeah and when 405B version release things are gonna change

    • @khanra17
      @khanra17 3 місяці тому

      I hate paying man.
      But the open source llms output dog poop !
      So I use openrouter!!!

    • @luisfelipe6368
      @luisfelipe6368 3 місяці тому +2

      Someone correct me if I am wrong, but ollama requires at least a mid-level PC for it to be viable. Not everybody has access to that. A mid-level PC are still a lot more expensive than some openAI credits.

  • @jozef-javorsky-dodo
    @jozef-javorsky-dodo 2 місяці тому

    THX stuff be so insightful it is super-great