https://github.com/ixaxaar/VardaGPT/blob/master/STORY.md Skip to content Toggle navigation Sign up * Product + Actions Automate any workflow + Packages Host and manage packages + Security Find and fix vulnerabilities + Codespaces Instant dev environments + Copilot Write better code with AI + Code review Manage code changes + Issues Plan and track work + Discussions Collaborate outside of code Explore + All features + Documentation + GitHub Skills + Blog * Solutions For + Enterprise + Teams + Startups + Education By Solution + CI/CD & Automation + DevOps + DevSecOps Case Studies + Customer Stories + Resources * Open Source + GitHub Sponsors Fund open source developers + The ReadME Project GitHub community articles Repositories + Topics + Trending + Collections * Pricing [ ] * # In this repository All GitHub | Jump to | * No suggested jump to results * # In this repository All GitHub | Jump to | * # In this user All GitHub | Jump to | * # In this repository All GitHub | Jump to | Sign in Sign up {{ message }} ixaxaar / VardaGPT Public * Notifications * Fork 4 * Star 130 * Code * Issues 1 * Pull requests 0 * Actions * Projects 0 * Wiki * Security * Insights More * Code * Issues * Pull requests * Actions * Projects * Wiki * Security * Insights Permalink master Switch branches/tags [ ] Branches Tags Could not load branches Nothing to show {{ refName }} default View all branches Could not load tags Nothing to show {{ refName }} default View all tags Name already in use A tag already exists with the provided branch name. Many Git commands accept both tag and branch names, so creating this branch may cause unexpected behavior. Are you sure you want to create this branch? Cancel Create VardaGPT/STORY.md Go to file * Go to file T * Go to line L * * Copy path * Copy permalink This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository. # ixaxaar add story Latest commit 33ca464 May 6, 2023 History 0 contributors Users who have contributed to this file Story of this project Background Project Idea Process Experience Conclusion Aside 110 lines (77 sloc) 4.53 KB Raw Blame Edit this file E Open in GitHub Desktop * Open with Desktop * View raw * * View blame Story of this project Background With all the hype around ChatGPT, I wondered how much impact ChatGPT really had. I mean, for a programmer, would ChatGPT be like a pair programmer? Like GitHub Copilot++? Or would ChatGPT totally replace programmers so that product managers could tell it what feature to build, and it would just build it! Imagine a bunch of product managers sitting in a sprint planning meeting where, after signing off on the tasks to be done this sprint and starting the sprint, ChatGPT was deployed on those tasks. The sprint lasted for about 2 hours, and everyone met again to do the next day's sprint grooming. Project Idea Now, what the heck should I build to test this? Why not try attaching a memory module to a GPT? I've seen some folks on the internet complain about the "low memory" problem of language models. I've also used FAISS and FLANN before, so I am familiar with how to technically achieve this. Whether it will actually work or not--well, my 1080Ti is on its deathbed with a broken fan, and I don't have the to train this thing on AWS anyway. Let's aim for unit tests to work then. Process Okay. I started with the project plan: start Then I made ChatGPT generate the project foundations, step by step, from creating project directories, Makefile, README, pre-commit, vscode settings for the same tools in pre-commit, setup.cfg, and a GitHub workflow to run tests. In each case, I had to specify exactly what I wanted it to generate. start Yes, I made ChatGPT choose the project name as well. start If I forgot something, I would go back and ask ChatGPT to add it: start In fact, I found it better to let ChatGPT generate a toy-ish version of the code first, then let it add things to it step-by-step. This resulted in much better output than, say, asking ChatGPT to generate production-quality code with all features in the first go. This also gave me a way to break down my requirements and feed them one at a time - as I was also acting as a code-reviewer for the generated output, and so this method was also easier for me to work with. start start start start start Of course, I made ChatGPT write unit tests, and if they failed, I would just copy the pytest output and feed it back into ChatGPT. start start ChatGPT even figured this out!: start The result - I present to you VardaGPT--every inch of this repository was generated by ChatGPT-4! It took a few hours, mostly around 3 weekends, mostly at odd times, to generate this project. Experience It felt neither like a Copilot++ nor like the product manager scenario but rather all at the same time. Sometimes I was amazed at what ChatGPT was able to understand, sometimes I had to stubbornly push it to go in a certain direction, sometimes it generated things I did not think of, sometimes I got super frustrated while making ChatGPT fix the code in a certain way. It was more like handholding a fresh grad who had absorbed all of human knowledge but needed someone to tie various parts of that knowledge to create something useful. Also ChatGPT is bad at dealing with abstractions beyond 2 layers. ChatGPT is definitely a productivity multiplier. I think it is rather a differential productivity multiplier, as it would enhance more the capabilities of those who already know more. If I did not understand deep learning and FAISS, or how projects are structured, I don't think I would have been able to pull this off. On the other hand, it also has some sort of a leveling effect--I have not worked on PyTorch in a while, have no idea of FAISS's new APIs, etc., but these gaps were filled in by ChatGPT. Finally, it was also tiring. Imagine being reduced to giving only instructions and doing code review. Reading and understanding code is tiring! Conclusion It looks like my job is safe this year. Time to generate an elaborate software project and claim supremacy on my ChatGPT usage abilities to hedge against next year. I wonder if by the time ChatGPT-6 comes out, would engineering teams be like, "Hey, let's generate our own Grafana with a purple theme ". Aside I could not resist but add this bit. ChatGPT is great at generating Agda! Maybe this would also be the ultimate tool that can be used to formalize all of pure math? start [ ] Go Footer (c) 2023 GitHub, Inc. Footer navigation * Terms * Privacy * Security * Status * Docs * Contact GitHub * Pricing * API * Training * Blog * About You can't perform that action at this time. You signed in with another tab or window. Reload to refresh your session. You signed out in another tab or window. Reload to refresh your session.