I have a question regarding handling data that exceeds the context window limit. I once worked on an application to generate unit tests. As you know, writing a good unit test for a function requires access to the definitions of all dependencies used by the function. While I, as a human, can access the entire codebase, this is not possible with an LLM. If I don’t provide the definitions of all the dependencies, the LLM will make assumptions, which can result in incorrect mocks for other classes or external tools used by the function I’m trying to test. Do you have any suggestions for addressing this issue? Summarization or advanced chunking solutions won't work in this case because the LLM must have full access to all dependencies to generate accurate unit tests.

Hey Guys! Seeking AI/software dev partner to build Media Franchise creator platform. Mentored by Chair of Marvel. Rep'd by Chair of Cooley. Excited to share this opportunity with you superstars.

Let me start with a little background about the problem. When I made my first Instagram Page, I knew what content to post for my audience but I just couldn't be consistent. I managed to get my first 3k followers within the first few months. It was ok, but I was not able to post consistently, something that many people struggle with. I used to follow a fixed pattern when creating carousel posts 1. Gather the text that I need to write 2. Open up Canva 3. Chose beautiful background images from Pexels 4. Write the text on the background image 5. Add a little bit of branding and post Fast forward a few years and ChatGPT is launched. I realized (being a programmer) that all of this can now be AUTOMATED with AI. So, I began working on the platform for creating Carousel posts for IG. And it works exactly in the same way I created posts manually, except that it uses AI. It uses OpenAI APIs to create the content and fetches images via Pexels API and combines them in an HTML Canvas element which can then be downloaded as an Image. But why Carousel posts? Because Carousel posts have the highest engagement rate (excluding reels) on Instagram as per Later[dot]com And now, with this software I created, I am able to create a Carousel post in less than a minute's time. Sounds cool right? This platform solves the real problem of creating meaningful content quickly and helps stay consistent on social media. This is going to help freelancers who create content for their clients, small to mid-size businesses trying to level up their social media, or even individuals looking to grow their social media. I am yet to release it publicly as it's still in beta and might contain a few bugs. I am also polishing the platform and making it more presentable and user-friendly. If you need free access to the software to test it out, just send me a DM and I'll give you free access to it. In return, please help me improve the platform with your feedback. I will post here once I launch it publicly. I am sure it'll help many.

Hello, I saw a notice in the course saying that we should use nextJs 14 instead of 15. Can anyone explain why? I just installed nextJs and it assumed I wanted the most recent version. I'm happy to work through the course with version 15, but if it's going to cause me a ton of issues, I'd rather take the time now to go back to 14.

Hi, all. I have an idea for a desktop app that requires AI functionalities. The app has huge potential in my field of work and potentially many others. I'm stunned that such an app doesn't already exist. I'm looking for a London-based AI developer with strong experience in Python, especially libraries like PyMuPDF, PyPDF2, or Adobe Acrobat SDK; familiarity with NLP tools like spaCy or Hugging Face; GUI development expertise with PyQt, Tkinter, or Electron.js; and experience with file processing and annotation tools. The project will need full-time attention for several weeks. We can discuss payment in person. Let me know if interested.