The world's first AI programmer and debugger, that isn't just a re-skin of ChatGPT. Tenet AI is trained for hyper-specialization on specific programming languages and implements an actual feedback loop, via the output terminal and memory storage.
Important
Tenet AI is currently in very early stages of development, as more of a proof-of-concept application. Community contributions are very much welcome to help accelerate our progress!
If you can see this checklist, that means that this project is in its very early stages, and likely incomplete.
| Status | Icon |
|---|---|
| Not Started | ⬜️ |
| In Development | 🟨 |
| Finished | ✅ |
- GUI Program
- 🟨 Code Editor (supporting all programs & editable)
- ⬜️ Internet Browsing Capabilities
- ⬜️ In-built Browser
- 🟨 Output Terminal
- 🟨 To do List
- 🟨 Settings
- LLM & Internet APIs
- 🟨 OpenAI
- 🟨 Anthropic
- ⬜️ Gemini
- ⬜️ Mistral
- ⬜️ Groq
- ⬜️ Google Search Engine
- ⬜️ Bing Search
- Programming Languages
- 🟨 Python (CUI only)
- ⬜️ Java (CUI only)
Innovation is fast. Five years ago, if you would've told me that OpenAI would release a LLM that would input and output text to barely resemble a conversation, I would've been scratching my head over that. But, this is reality.
Here's the catch - Adaptability is slow. Companies will not simply fire their core staff because AI can replace them. No, I don't make this claim because I "have faith in the humanity of people". I say so because of all the little edge cases that can make-or-break an entire industry, if attempted with such carelessness.
Companies need to worry about, how will they be able to substitute mass outages of AI bots, how will we provide these AI bots with the necessary computing power, how do we manage those AI bots to make sure they don't screw up as fast, etc.
The core purpose of this AI tool is to create a whole new environment to assist existing programmers and engineers. Good product is made by understanding what your users want, not by how clean and documented your code is. We are not trying to be Devin, we're trying to be better.
The purpose of Tenet AI is to be able to not only be able to write AI generated code, but also to help developers with structuring their project files, manage external dependencies, optimize for CPU and memory consumption, and more.
On top of all, I want Tenet to help us, instead of replace us.
- Tenet understands technology.
Instead of immediately acting upon a poorly written prompt, on a poorly described project, Tenet will actually ask the user for more specifications, research the internet for viable ways to accomplish certain tasks, and create a robust To-do-List for it to be able to adhere to. - Tenet understands feedback.
Instead of presenting a half-assed program to the user, with silly errors like calling variables that never existed, or intentionally causing a runtime error, Tenet will use elaborate program outputs, and check the memory stack, to implement faster and/or less resourceful alternatives, and fix needless errors within the code. - Tenet understands your budget.
Instead of eating through your API credits, Tenet help you recommend a strong budget for your project. The budget can be easily expanded and contracted based on further refining of the project idea. On top, instead of tracking using tokens, we track using US Dollars ($), along with 5 other currencies.
If you're a newcomer into the project, or if you're simple confused as to how is our project useful, here are the recommended use cases for Tenet AI. Do note that we do not want to be limited by these cases alone, so feedback would be appreciated here as a feature request.