This Will Change How You Use ChatGPT Forever!

Author Avatar

athompson

Joined: Mar 2024
Spread the love

This Will Change How You Use ChatGPT Forever!


ChatGPT rolled out a new feature that helps you better leverage custom GPTs. Check it out!

Discover More From Me:
🛠️ Explore hundreds of AI Tools: https://futuretools.io/
📰 Weekly Newsletter: https://www.futuretools.io/newsletter
😊 Discord Community: https://futuretools.io/discord
🐤 Follow me on Twitter:…

source

Reviews

0 %

User Score

0 ratings
Rate This

Sharing

Leave your comment

Your email address will not be published. Required fields are marked *

34 Comments

  1. People are so stupid. You're showing off a new feature, have the title and thumbnail saying it's changed forever, and someone called it awful clickbait with 303 other morons liking the comment? How does a new feature not change it forever?

    If, for example, when Pokémon GO released the Buddy feature, someone made a video called "Pokémon GO is changed forever," would that be wrong? No!

    Keep doing what you're doing, Matt. You don't have terrible clickbait. Sometimes the titles get repetitive, but it's hard to come up with new titles. Don't let the morons get you down.

    If I've learned anything in my life it's that there are a lot of dumbasses on Planet Earth, and we now live in a society where it's okay to be stupid and if you call someone out for being an imbecile, people get mad at you. If someone spells something wrong and you correct them, you'll get someone screaming "LaNgUaGe EvOLvEs!" at you! I'm so tired of morons.

  2. Like the "idea" of a GPT Store but I'm not sure or am missing if you can create a GPT and make it available to other mobile users without a GPT Plus account, if you can't the store does seem very closed or limited in uses,

  3. 🎯 Key Takeaways for quick navigation:

    00:00 🤖 ChatGPT Plus feature introduction and demonstration
    – ChatGPT Plus allows users to call upon multiple GPTs within one chat.
    – Introduction to the feature and demonstration of its capabilities.
    – Examples include generating images, creating PDFs, making PowerPoint slides, explaining concepts, creating diagrams, and producing videos.
    02:02 🎨 Image generation, PDF conversion, and PowerPoint slide creation
    – Using DALL-E to generate an image.
    – Converting the image into a PDF using a GPT called "Convert Anything."
    – Creating a PowerPoint slide with a story based on the image using "PowerPoint Presentation Maker."
    04:29 💡 Explaining concepts, diagram creation, and video production
    – Requesting an explanation of a three-way circuit from ChatGPT.
    – Creating a diagram of the circuit using a GPT called "Diagrams Show Me."
    – Producing a video about the topic using "Video Maker by Nvidia Ai."
    06:59 📚 Research, diagram generation, and HTML conversion
    – Using the "Consensus" GPT to find the top five papers about artificial intelligence.
    – Creating a diagram based on the papers using "Diagrams."
    – Attempting to convert the diagram into HTML using "Screenshot to Code GPT."
    09:04 🔄 Integration of GPTs and future possibilities
    – Discussion on the potential of GPT integration and improvement over time.
    – ChatGPT Plus as a tool for streamlining tasks and acting as a personal assistant.
    – Expectation of improved functionality as more GPTs are designed to work together.

    Made with HARPA AI

  4. I'm looking for a prompt to generate a table for different actions for my prompts (such as analyze , rewrite , etc)

    But I also want a prompt for thinking-outside-the-box about generating tables for those classes

  5. 🎯 Key Takeaways for quick navigation:

    00:00 ChatGPT artık birden fazla GPT'yi bir sohbette çağırma özelliğine sahip, bu özellik yalnızca ChatGPT Plus üyeleri için geçerli.
    00:27 Belirli bir GPT ile doğrudan sohbet etmek için, daha önce kullanılmış olması gerekiyor.
    02:15 Dolly 3 kullanılarak kurtun bir taco yediği bir resim oluşturuldu ve sonra "convert anything" GPT'si ile bu resim PDF'e dönüştürüldü.
    03:25 PowerPoint sunumu yapıcı GPT, kurdun taco yeme hikayesini içeren bir slayt oluşturdu, ancak resmi slayta ekleyemedi.
    04:29 ChatGPT, üç yönlü bir devrenin nasıl çalıştığını açıkladı ve ardından bir diyagram oluşturmak için başka bir GPT çağrıldı.
    05:23 Nvidia video yapıcı GPT, üç yönlü aydınlatma devresinin nasıl çalıştığını özetleyen bir video oluşturdu.
    07:14 Yapay zeka hakkındaki en iyi beş makale hakkında bilgi verildi ve bu bilgilerle bir diyagram oluşturuldu.
    08:35 Bir GPT, diyagramı HTML web sitesi koduna dönüştürmeye çalıştı, ancak tam olarak başarılı olamadı.
    09:18 Bu yeni özellik, ChatGPT'yi daha çok tek bir sohbette birçok işlemi gerçekleştirebilen kişisel bir asistana dönüştürmeyi amaçlıyor.
    11:08 Yapay zeka araçları hakkında güncel bilgiler ve en yeni haberler için futur tools. web sitesine bakılması öneriliyor.

    Made with HARPA AI

  6. I didn’t know I had to type @. During a conversation it seems to pick up on when I ask for an image just from the way I give a prompt for an image. I’ll try this.

  7. Awesome feature! Create a team of employees and have them work together. It sure does feel like the jobs that don't require a lot of brain power and creativity will be gone very soon. Self-education will be crucial for decent survival in the future.

  8. Hey there! I think it would be awesome if you could grab the presentation and that cool wolf image. Once you have them, hop on over to Canva and add that wolf image to the presentation. It's going to look amazing!

  9. I could see this being helpful for authors. Use a GPT to help you brainstorm, and another to organize the data, and another to generate character artwork for reference material, etc.

  10. The reason you were having trouble with images in each of your demos is that you misunderstood how images are handled by the LLM. The LLM only deals with text. When the LLM sends text to Dalle3, Dalle3 can drop binary format image files into /mnt/data inside the python vm, which is offline-only. ChatGPT can only do a few things regarding these image files: (1) show a link to one such that it renders or is downloadable but only to the present user not the public, (2) process the file using python which it writes and runs in the offline virtual machine, or (3) since we technically have the "vision" model it can in theory write a text description of the image. It CANNOT pass the image to an external API such as the slideshow or video services you showed. The custom GPT actions (defined by a schema) only accept TEXT which is sent to a remote server. The only way the LLM could conceivably pass an image to those services (for example to appear in a slide deck or video) would be for you to browbeat it into painstakingly encoding the image data into TEXT (base64 etc) and chunking it through the API to bypass datasize/token limits – and the API would need to be engineered to work with this. The token limits for output are even smaller than for input and as of the time of writing this there is no reasonable solution. Basically you can't do what you are trying to do. If you look back at your demonstrations with this perspective you will immediately understand why things failed in the ways they did. BTW, once image exfiltration is possible, the Canva GPT and many others with similar goals (get user to go buy products/services and use generate images on external sites) will be doing it. You can watch them as a sort of canary. If they aren't doing it, it can't be done yet.

  11. Hi Matt, that looks really cool! And of course, the next step/question is – can you create a Jarvis GPT that would just seek out and ping the respective GPT's and coordinate the end result, kind of like you can build with agents via the OpenAI api. That will be cool – like you could select 10 GPT's in a checkbox list as your 'agents' that the central Jarvis GPT could employ to coordinate, interact with and provide feedback with until the entire start-to-end task is done.

  12. It doesn’t matter what people (negatively) say about Matt…
    I subscribed to his channel when he had around 50K subs, look at him now! This says it all.