Members-Only
Recent Talks & Demos are for members only
You must be an AI Tinkerers active member to view these talks and demos.
AI powered reading buddy - what Kindle should be, but isn't
A live demo of a web‑and‑Docker tool that reads academic PDFs line‑by‑line, adds context via graph search, and organizes personal research libraries.
I’ve been working on a side project. ResearchwithAI (https://researchwith.ai) - a reading assistant to help read Arxiv papers (and eventually PDFs and other docs). Inspired by Andrej Karpathy’s tweet (
https://x.com/karpathy/status/1866902647804203363), I set out to build a tool you can access via the web or run locally (via Docker and Ollama) that can help you read papers and explain concepts line by line instead of uploading a whole PDF and then asking questions - Like a smarter Kindle.
I also plan to 1) integrate graph search to include other relevant research as context to help answer the questions better and 2) To provide a way to store and access your past-read papers via the UI. You can view an early version on the web using the URL I shared earlier. After adding a few more features, I plan to make the codebase public (hopefully before the presentation).
During the presentation, I’ll talk about the architecture, my approach to building it as someone who just picked up Web development earlier this year, my (bad) experience with Vibe coding, and how I plan on improving the tool over time.
researchwith.ai uses AI to process arXiv articles via ID.
X.com's styled 404 error page with JavaScript disabled warning.
- ChatGPT APIIntegrate OpenAI's flagship LLMs (GPT-4o, GPT-3.5-turbo) into your applications via a robust, scalable REST API for text generation and conversational AI.The ChatGPT API (accessed through the Chat Completions endpoint) is your direct pipeline to deploy OpenAI's most advanced large language models, including GPT-4o and the highly efficient GPT-3.5-turbo. This powerful, pay-as-you-go service allows developers to build sophisticated conversational experiences, generate diverse content (code, structured JSON, human-like prose), and execute complex reasoning tasks at scale. Authentication is simple: use your secret API key for Bearer token authorization in your standard HTTP requests. It is the proven, production-ready foundation for next-generation AI features.
- ReactReact is an open-source JavaScript library for building dynamic user interfaces (UIs).React is a component-based JavaScript library, developed by Meta (Facebook), engineered for building fast, declarative UIs. It mandates a one-way data flow and utilizes a Virtual DOM mechanism to ensure efficient, predictable updates to the user interface. Developers construct complex UIs by composing small, encapsulated components; this architecture promotes code reusability and simplifies state management across large applications. The library employs JSX (a syntax extension) to integrate HTML-like markup directly within JavaScript logic, supporting development for both web (React DOM) and native mobile platforms (React Native).
- NextNext.js is the full-stack React framework: it delivers high-performance web applications via hybrid rendering and powerful, Rust-based tooling.This is the React Framework for production: Next.js enables you to build full-stack web applications with zero configuration and maximum efficiency. It supports a hybrid rendering approach (Server-Side Rendering, Static Site Generation, and Incremental Static Regeneration) for optimal speed and SEO performance. Key features include React Server Components, Server Actions for running server code directly, and the App Router for advanced routing and nested layouts. Developed by Vercel, it leverages Rust-based tools like Turbopack and the Speedy Web Compiler for the fastest possible builds and a superior developer experience.
- CursorThe AI-native code editor designed for high-velocity development through deep LLM integration.Cursor is a fork of VS Code that embeds AI directly into the development workflow while maintaining full extension compatibility. It leverages models like Claude 3.5 Sonnet and GPT-4o to power features such as Cmd+K for inline edits and Cmd+L for codebase-wide chat. By indexing local files, Cursor provides precise context for its predictive 'Tab' completions and multi-file 'Composer' mode. This setup allows engineers to move from high-level intent to functional code without leaving the editor or losing context.
- AiderAider is the open-source, terminal-native AI pair programmer: it executes multi-file code edits and maintains a clean Git history with automatic, descriptive commits.This is your command-line AI coding partner: Aider connects your local Git repository to powerful LLMs (GPT-4o, Claude 3.7 Sonnet, DeepSeek) for direct, conversational code modification. It handles complex, multi-file changes across 100+ languages (Python, Rust, JavaScript) without forcing you to leave the terminal. The key differentiator is its Git-native workflow: Aider automatically stages and commits every AI-generated change with a clear, contextual message, giving you a clean, reviewable history and full control. Typical feature costs run low: expect $0.01–0.10 per feature with cloud models.
Compose Email
Loading recent emails...