If you need help with this or any other Bubble-related build --> you can grab a coaching slot here (I have a few per week): buy.stripe.com/aEUcQD8V3bp7h1ueV2 If you want to learn how to build a SaaS app in 7 days using Bubble and AI, check out www.avalanlabs.co
Hey. Im wondering how come you have not shared the whole prompt you used? I mean including the beinning part that give the outline for it being a prompt engineer. Would be appreciated, as i am a lazy piece of trash who cant be bothered to type the whole thing manually but somehow bothered to type all this.
@@avalanlabs Ideally a tutorial to build a simple but concrete application of the new custom GPTs in Bubble. For instance, having 2 GPTs, "discussing" with one another for several exchanges? Where we could select on GPT who is more in favor of A on a topic, and another GPT more in favor of B, each of them corresponding to a different GPT. I am very curious about how to put together the chat, history, database, when manipulating different concepts and ChatGPT "entities". I do not know how complex this would be in Bubble though... What is interesting is how to manage all those different ChatGPT api calls "efficiently" in Bubble. What to store in database (front end or backend?) what not to store and use custom states instead maybe, etc.
Great questions! GPTs are essentially just the Assistant API but with chatGPT as a front-end + you are able to build custom GPTs by just prompting it in the chat. You can then share GPTs with other chatGPT users. But the foundation is the same, so they have similar capabilities.
thank you for the tutorial! what we're missing now, is the instructions on how to make the assistant usable within bubble. You showed us how to connect it ti the api, but now we have to connect it to a chat within bubble. how can we do this? thank you once again
CAUTION! FYI there is a failure of direct retrieval with GPT-4 using the new OpenAI Assistant API. GPT tokenizes text and creates its own vector embeddings based on its specific training data. The new terms and sequences may not connect well to the pretrained knowledge in GPT's weight tensors. There was no semantic similarity between the new API terms and GPT's existing vector space. This is a fundamental issue with retrieval augmentation systems like Rag - external knowledge is not truly integrated into the model's learned weights. Adding more vector stores cannot solve this core problem. The solution is to have multiple learned "knowledge planes" with trained weight tensors for specific tasks that can be switched in. This is better than just retrieving separate vector representations.
Great toturial, thanks! I was trying to continue and adding the thread and message API calls, so I can create the full cycle UX but couldn't find the way. Are you using only this API call to create a complete UX process on the app?
To implement the Assistants API from OpenAI in Bubble, you would use Bubble's API Connector. This plugin allows you to integrate with any service with an API. Below are the steps and corresponding code snippets you might use to set up the API calls within the Bubble API Connector. Keep in mind that you'll need to replace placeholders with your actual API keys and relevant information. Step 1: Set up the API Connector in Bubble Go to your Bubble editor and navigate to the Plugins tab. Add a new plugin and search for the API Connector. Once you've added the API Connector, you'll start setting up the API calls. Step 2: Configure the API Calls Create an Assistant In the API Connector, create a new API call. Name it Create Assistant. Set the type to Data (this is a call to get or send data). Method should be POST. The URL will be api.openai.com/v1/beta/assistants. Add a header with the key Authorization and the value Bearer your_openai_api_key. Add another header with the key OpenAI-Beta and the value assistants=v1. In the Body, you'll define your assistant, something like this: json Copy code { "name": "Math Tutor", "instructions": "You are a personal math tutor. Write and run code to answer math questions.", "tools": [{"type": "code_interpreter"}], "model": "gpt-4-1106-preview" } Create a Thread Create another API call named Create Thread. Method: POST. URL: api.openai.com/v1/beta/threads. Headers: Same as above. Add a Message to a Thread Create an API call named Add Message. Method: POST. URL: api.openai.com/v1/beta/threads/{thread_id}/messages. URL Params: Define thread_id as a dynamic parameter to be passed with each request. Headers: Same as above. Body: json Copy code { "role": "user", "content": "I need to solve the equation `3x + 11 = 14`. Can you help me?" } Run the Assistant Create an API call Run Assistant. Method: POST. URL: api.openai.com/v1/beta/threads/{thread_id}/runs. Headers: Same as above. Body: json Copy code { "assistant_id": "{assistant_id}", "instructions": "Please address the user as Jane Doe. The user has a premium account." } Display the Assistant's Response Create an API call Retrieve Messages. Method: GET. URL: api.openai.com/v1/beta/threads/{thread_id}/messages. Headers: Same as above. Step 3: Use the API in Your Bubble Application Now, you can use these API calls in your application. For example, when a user submits a question: Run Create Thread to initiate a new conversation. Use Add Message to send the user's query to the thread. Call Run Assistant to get the assistant to process the query. Retrieve the response with Retrieve Messages and display it to the user. You'll need to manage the workflow and dynamic data in your Bubble application to ensure that you're passing the correct thread IDs and assistant IDs to each API call. This will involve storing these IDs in your Bubble database or as temporary states in your application. This is a high-level guide and assumes familiarity with the Bubble platform and working with APIs. You may need to adjust these instructions to fit the specifics of your application and handle any additional logic or error checking as needed.
🎯 Key Takeaways for quick navigation: 00:00 🚀 *OpenAI has released multiple significant updates, including the Assistant API.* 00:29 📈 *GP4 Turbo introduced, offering faster processing, lower cost, and a much larger context window (128,000 tokens).* 01:11 🤖 *The Assistant API allows the creation of purpose-specific assistants that can reference external materials like PDFs and code.* 01:38 🌐 *Assistant API includes a code interpreter and can call external APIs for task automation.* 02:05 🌟 *Potential for assistants to create their own assistants, hinting at a hierarchical AI system.* 02:44 🛠️ *Introduction to building an assistant in the OpenAI Playground, including custom functions and data retrieval.* 03:38 📚 *Limitations in file uploads for assistants: maximum 20 files, up to 512 MB each.* 04:19 📝 *Emphasis on clear and specific instructions for creating effective AI assistants.* 05:14 💡 *Effective utilization of AI models requires detailed and well-thought-out instructions.* 06:12 🚀 *Using NASA's biomimicry design assistant prompt as a template for creating detailed assistant instructions.* 07:07 🎨 *Concept of "meta prompts" introduced - prompts that generate other prompts.* 08:16 🧠 *Demonstrating the creation of a specific assistant (SaaS pricing expert) using a detailed prompt.* 09:37 ⚙️ *Process of integrating the Assistant API with Bubble, a no-code platform.* 11:04 🖥️ *Tutorial on setting up and testing a custom assistant using the Assistant API.* 15:23 🌐 *Integration of the Assistant API with Bubble to enable no-code development of AI-powered applications.* 16:59 🤔 *Challenges and solutions in translating OpenAI documentation into Bubble API connector setup.* 19:08 🔄 *Iterative process of setting up API calls in Bubble based on OpenAI's Assistant API documentation.* 24:20 🛠️ *Debugging and correcting the API integration process in Bubble.* 25:44 📚 *Plans to share step-by-step instructions for implementing the Assistant API in Bubble.* 26:10 👋 *Conclusion and invitation to subscribe for more tutorials on AI and no-code development.* Made with HARPA AI
°\\• I believe we are meant to be like Jesus in our hearts and not in our flesh. But be careful of AI, for it is just our flesh and that is it. It knows only things of the flesh (our fleshly desires) and cannot comprehend things of the spirit such as peace of heart (which comes from obeying God's Word). Whereas we are a spirit and we have a soul but live in the body (in the flesh). When you go to bed it is your flesh that sleeps but your spirit never sleeps (otherwise you have died physically) that is why you have dreams. More so, true love that endures and last is a thing of the heart (when I say 'heart', I mean 'spirit'). But fake love, pretentious love, love with expectations, love for classic reasons, love for material reasons and love for selfish reasons that is a thing of our flesh. In the beginning God said let us make man in our own image, according to our likeness. Take note, God is Spirit and God is Love. As Love He is the source of it. We also know that God is Omnipotent, for He creates out of nothing and He has no beginning and has no end. That means, our love is but a shadow of God's Love. True love looks around to see who is in need of your help, your smile, your possessions, your money, your strength, your quality time. Love forgives and forgets. Love wants for others what it wants for itself. Take note, true love works in conjunction with other spiritual forces such as patience and faith (in the finished work of our Lord and Savior, Jesus Christ, rather than in what man has done such as science, technology and organizations which won't last forever). To avoid sin and error which leads to the death of our body and also our spirit in hell fire, we should let the Word of God be the standard of our lives not AI. If not, God will let us face AI on our own and it will cast the truth down to the ground, it will be the cause of so much destruction like never seen before, it will deceive many and take many captive in order to enslave them into worshipping it and abiding in lawlessness. We can only destroy ourselves but with God all things are possible. God knows us better because He is our Creater and He knows our beginning and our end. Our prove text is taken from the book of John 5:31-44, 2 Thessalonians 2:1-12, Daniel 7-9, Revelation 13-15, Matthew 24-25 and Luke 21. Let us watch and pray... God bless you as you share this message to others.
If you need help with this or any other Bubble-related build --> you can grab a coaching slot here (I have a few per week): buy.stripe.com/aEUcQD8V3bp7h1ueV2
If you want to learn how to build a SaaS app in 7 days using Bubble and AI, check out www.avalanlabs.co
How I can ask for your service to help building an MVP? I mean a contact to reach you. On the web site there is no way to get your contact
tally.so/r/3NXR5B
Hey. Im wondering how come you have not shared the whole prompt you used? I mean including the beinning part that give the outline for it being a prompt engineer. Would be appreciated, as i am a lazy piece of trash who cant be bothered to type the whole thing manually but somehow bothered to type all this.
Really great explanation , thanks for not adopting a scam or too heavy hype tone
This is really great! Are you planning for more Bubble / ChatGPT API tutorials in the next few weeks? Thanks!
Sure! What would you like to see?
@@avalanlabs Ideally a tutorial to build a simple but concrete application of the new custom GPTs in Bubble. For instance, having 2 GPTs, "discussing" with one another for several exchanges? Where we could select on GPT who is more in favor of A on a topic, and another GPT more in favor of B, each of them corresponding to a different GPT. I am very curious about how to put together the chat, history, database, when manipulating different concepts and ChatGPT "entities". I do not know how complex this would be in Bubble though... What is interesting is how to manage all those different ChatGPT api calls "efficiently" in Bubble. What to store in database (front end or backend?) what not to store and use custom states instead maybe, etc.
is it possible to connect via API on custom GPT? just like using Assistant API
According to OpenAI developer forum: ”At the moment it is not possible to connect to custom GPTs via API”
Hi! What differences are there between GPTs and Assistants? When is it adequate to use one or another?
Great questions! GPTs are essentially just the Assistant API but with chatGPT as a front-end + you are able to build custom GPTs by just prompting it in the chat. You can then share GPTs with other chatGPT users. But the foundation is the same, so they have similar capabilities.
It's like the difference between chatGPT and the regular gpt-4 api
Only thing missing: your meta prompt.
thank you for the tutorial! what we're missing now, is the instructions on how to make the assistant usable within bubble. You showed us how to connect it ti the api, but now we have to connect it to a chat within bubble. how can we do this? thank you once again
Great video thank you. Would you be able share link to the metaprompting NASA site you mention please
Here you go: github.com/nasa-petal/discord_bot#bidara--bio-inspired-design-and-research-assistant
Much appreciated, thank you.
CAUTION! FYI there is a failure of direct retrieval with GPT-4 using the new OpenAI Assistant API. GPT tokenizes text and creates its own vector embeddings based on its specific training data. The new terms and sequences may not connect well to the pretrained knowledge in GPT's weight tensors.
There was no semantic similarity between the new API terms and GPT's existing vector space. This is a fundamental issue with retrieval augmentation systems like Rag - external knowledge is not truly integrated into the model's learned weights. Adding more vector stores cannot solve this core problem.
The solution is to have multiple learned "knowledge planes" with trained weight tensors for specific tasks that can be switched in. This is better than just retrieving separate vector representations.
How would you implement into your assistant for instance if the answer requires a graph or anything else besides text?
Great toturial, thanks! I was trying to continue and adding the thread and message API calls, so I can create the full cycle UX but couldn't find the way. Are you using only this API call to create a complete UX process on the app?
Epic video bro - love it!
how can I replicate an action from custom got into assistants? I got the open api schema
Instructions are missing!
To implement the Assistants API from OpenAI in Bubble, you would use Bubble's API Connector. This plugin allows you to integrate with any service with an API. Below are the steps and corresponding code snippets you might use to set up the API calls within the Bubble API Connector. Keep in mind that you'll need to replace placeholders with your actual API keys and relevant information.
Step 1: Set up the API Connector in Bubble
Go to your Bubble editor and navigate to the Plugins tab.
Add a new plugin and search for the API Connector.
Once you've added the API Connector, you'll start setting up the API calls.
Step 2: Configure the API Calls
Create an Assistant
In the API Connector, create a new API call.
Name it Create Assistant.
Set the type to Data (this is a call to get or send data).
Method should be POST.
The URL will be api.openai.com/v1/beta/assistants.
Add a header with the key Authorization and the value Bearer your_openai_api_key.
Add another header with the key OpenAI-Beta and the value assistants=v1.
In the Body, you'll define your assistant, something like this:
json
Copy code
{
"name": "Math Tutor",
"instructions": "You are a personal math tutor. Write and run code to answer math questions.",
"tools": [{"type": "code_interpreter"}],
"model": "gpt-4-1106-preview"
}
Create a Thread
Create another API call named Create Thread.
Method: POST.
URL: api.openai.com/v1/beta/threads.
Headers: Same as above.
Add a Message to a Thread
Create an API call named Add Message.
Method: POST.
URL: api.openai.com/v1/beta/threads/{thread_id}/messages.
URL Params: Define thread_id as a dynamic parameter to be passed with each request.
Headers: Same as above.
Body:
json
Copy code
{
"role": "user",
"content": "I need to solve the equation `3x + 11 = 14`. Can you help me?"
}
Run the Assistant
Create an API call Run Assistant.
Method: POST.
URL: api.openai.com/v1/beta/threads/{thread_id}/runs.
Headers: Same as above.
Body:
json
Copy code
{
"assistant_id": "{assistant_id}",
"instructions": "Please address the user as Jane Doe. The user has a premium account."
}
Display the Assistant's Response
Create an API call Retrieve Messages.
Method: GET.
URL: api.openai.com/v1/beta/threads/{thread_id}/messages.
Headers: Same as above.
Step 3: Use the API in Your Bubble Application
Now, you can use these API calls in your application. For example, when a user submits a question:
Run Create Thread to initiate a new conversation.
Use Add Message to send the user's query to the thread.
Call Run Assistant to get the assistant to process the query.
Retrieve the response with Retrieve Messages and display it to the user.
You'll need to manage the workflow and dynamic data in your Bubble application to ensure that you're passing the correct thread IDs and assistant IDs to each API call. This will involve storing these IDs in your Bubble database or as temporary states in your application.
This is a high-level guide and assumes familiarity with the Bubble platform and working with APIs. You may need to adjust these instructions to fit the specifics of your application and handle any additional logic or error checking as needed.
@@avalanlabs hello. Can you please a another video for a beginner to understnd this concept?
hey great video. can you publish this as GPT so that we can use it? thanks
Hi Man, can you make a complete tutorial video on how to make a AI SaaS using OpenAI's Assistants API & Bubble also link it with Stripe? Thank You
I already subscribed bytheway
Awesome
🎯 Key Takeaways for quick navigation:
00:00 🚀 *OpenAI has released multiple significant updates, including the Assistant API.*
00:29 📈 *GP4 Turbo introduced, offering faster processing, lower cost, and a much larger context window (128,000 tokens).*
01:11 🤖 *The Assistant API allows the creation of purpose-specific assistants that can reference external materials like PDFs and code.*
01:38 🌐 *Assistant API includes a code interpreter and can call external APIs for task automation.*
02:05 🌟 *Potential for assistants to create their own assistants, hinting at a hierarchical AI system.*
02:44 🛠️ *Introduction to building an assistant in the OpenAI Playground, including custom functions and data retrieval.*
03:38 📚 *Limitations in file uploads for assistants: maximum 20 files, up to 512 MB each.*
04:19 📝 *Emphasis on clear and specific instructions for creating effective AI assistants.*
05:14 💡 *Effective utilization of AI models requires detailed and well-thought-out instructions.*
06:12 🚀 *Using NASA's biomimicry design assistant prompt as a template for creating detailed assistant instructions.*
07:07 🎨 *Concept of "meta prompts" introduced - prompts that generate other prompts.*
08:16 🧠 *Demonstrating the creation of a specific assistant (SaaS pricing expert) using a detailed prompt.*
09:37 ⚙️ *Process of integrating the Assistant API with Bubble, a no-code platform.*
11:04 🖥️ *Tutorial on setting up and testing a custom assistant using the Assistant API.*
15:23 🌐 *Integration of the Assistant API with Bubble to enable no-code development of AI-powered applications.*
16:59 🤔 *Challenges and solutions in translating OpenAI documentation into Bubble API connector setup.*
19:08 🔄 *Iterative process of setting up API calls in Bubble based on OpenAI's Assistant API documentation.*
24:20 🛠️ *Debugging and correcting the API integration process in Bubble.*
25:44 📚 *Plans to share step-by-step instructions for implementing the Assistant API in Bubble.*
26:10 👋 *Conclusion and invitation to subscribe for more tutorials on AI and no-code development.*
Made with HARPA AI
°\\• I believe we are meant to be like Jesus in our hearts and not in our flesh. But be careful of AI, for it is just our flesh and that is it. It knows only things of the flesh (our fleshly desires) and cannot comprehend things of the spirit such as peace of heart (which comes from obeying God's Word). Whereas we are a spirit and we have a soul but live in the body (in the flesh). When you go to bed it is your flesh that sleeps but your spirit never sleeps (otherwise you have died physically) that is why you have dreams. More so, true love that endures and last is a thing of the heart (when I say 'heart', I mean 'spirit'). But fake love, pretentious love, love with expectations, love for classic reasons, love for material reasons and love for selfish reasons that is a thing of our flesh. In the beginning God said let us make man in our own image, according to our likeness. Take note, God is Spirit and God is Love. As Love He is the source of it. We also know that God is Omnipotent, for He creates out of nothing and He has no beginning and has no end. That means, our love is but a shadow of God's Love. True love looks around to see who is in need of your help, your smile, your possessions, your money, your strength, your quality time. Love forgives and forgets. Love wants for others what it wants for itself. Take note, true love works in conjunction with other spiritual forces such as patience and faith (in the finished work of our Lord and Savior, Jesus Christ, rather than in what man has done such as science, technology and organizations which won't last forever). To avoid sin and error which leads to the death of our body and also our spirit in hell fire, we should let the Word of God be the standard of our lives not AI. If not, God will let us face AI on our own and it will cast the truth down to the ground, it will be the cause of so much destruction like never seen before, it will deceive many and take many captive in order to enslave them into worshipping it and abiding in lawlessness. We can only destroy ourselves but with God all things are possible. God knows us better because He is our Creater and He knows our beginning and our end. Our prove text is taken from the book of John 5:31-44, 2 Thessalonians 2:1-12, Daniel 7-9, Revelation 13-15, Matthew 24-25 and Luke 21. Let us watch and pray... God bless you as you share this message to others.