Members-Only
Recent Talks & Demos are for members only
You must be an AI Tinkerers active member to view these talks and demos.
Veteran of the Slop War
This talk demonstrates a Haskell optimizer that finds and refactors duplicated code using e-graphs and anti-unification, with mechanical proof of equivalence.
A HASKELL code optimizer that detects semantically duplicated program structure (scoped equivalence + e-graph saturation + anti-unification) and rewrites it into shared abstractions only when a mechanical proof of equivalence passes. Live, I’ll run it on a real codebase and show the duplication it surfaces, a rewrite firing with its proof seal succeeding, and one where the engine refuses to touch the code. Bam, less slop.
I would prefer not to open source this project, this is still more of a creative and technical experiment.
Generated 6 days ago
Generating a talk summary...
View full transcript
- HaskellAn advanced, purely functional programming language featuring static typing, lazy evaluation, and robust type inference.Haskell is a statically typed, purely functional language built to handle complex software with high mathematical precision. By enforcing purity (functions have no side effects by default) and utilizing lazy evaluation (computations are deferred until necessary), it eliminates entire classes of runtime errors. The Glasgow Haskell Compiler (GHC) serves as its primary engine, delivering high-performance native code alongside advanced concurrency tools like Software Transactional Memory (STM). It is the language of choice for systems requiring absolute correctness, from financial trading engines to compiler design.
- Algebraic Graphs (alga)Alga is a Haskell library for constructing and manipulating graphs using a type-safe, mathematically rigorous algebraic foundation.Traditional graph representations rely on a pair of sets (vertices and edges) that easily lead to partial functions and runtime errors (like an edge pointing to a non-existent vertex). Alga solves this by defining graphs recursively through four core algebraic constructors: Empty, Vertex, Overlay, and Connect. This elegant mathematical framework guarantees that every constructed expression represents a valid graph, allowing developers to apply equational reasoning to prove the correctness of their transformation algorithms. The library handles millions of vertices and billions of edges efficiently, providing a robust, type-safe alternative to low-level graph manipulation.
- MegaparsecAn industrial-strength, highly flexible monadic parser combinator library for Haskell that delivers exceptional error reporting.Megaparsec is a feature-rich Haskell library designed to build robust parsers by combining small, composable functions. It strikes a precise balance between raw speed, flexibility, and the quality of its parse errors (generating highly detailed, customizable error messages out of the box). Developed as a modern, more reliable alternative to the classic Parsec library, Megaparsec supports custom token streams, state tracking, and sophisticated recovery mechanisms. This makes it a go-to choice for production-grade tasks: from parsing custom domain-specific languages and configuration files to handling complex programming language syntaxes.
- SolidJSSolidJS is a declarative, high-performance JavaScript library for building user interfaces using fine-grained reactivity and a compiler instead of a Virtual DOM.SolidJS delivers exceptional performance by compiling JSX templates directly to real DOM nodes, bypassing the overhead of a Virtual DOM entirely. Built on fine-grained reactive primitives (such as Signals, Memos, and Effects), components execute exactly once while updates surgically target only the specific DOM elements that change. With a tiny 7kb minified bundle size and a familiar, React-like API, SolidJS gives developers the ergonomic comfort of modern declarative frameworks alongside the raw speed of direct DOM manipulation.
- CodexCodex is OpenAI's autonomous AI software engineering agent: it executes full development tasks in a sandboxed cloud environment.Codex is the advanced, cloud-based software engineering agent from OpenAI, built on a specialized model like `codex-1` (a fine-tuned version of `o3`). It operates on an asynchronous delegation model, allowing developers to assign complete tasks—not just receive suggestions—via the ChatGPT interface. The agent works independently in a secure, isolated cloud container provisioned with the user's GitHub repository and environment. It reads code, writes new features, fixes bugs, runs tests, and drafts Pull Requests (PRs) for review, significantly accelerating the development lifecycle. Access is provided through ChatGPT Plus, Pro, and Enterprise plans.
Related talks
More from the community
AI Writes the Code. Who Cleans Up the Mess?
San Francisco
See an autonomous agent diagnose tech debt and automatically create prioritized fixes. Witness a live demo of the…
From Abacus to AI: Small Potatoes, Big Results
Hong Kong
See how a non-coder built World of Warcraft addons, a bus app, and a phone ringer using AI…
Building a System That Knows Why It Changed
Nashville
This talk shows a system that learns from failures by committing to expected outcomes and using structured debug…
Coding is Solved. Next Up: Figure Out How to Plan
Seattle
Discover how to manage AI-generated plans and code reviews efficiently. This talk explores building an agent-native Google Docs…
Virtual Machine = Real Collaboration
Seattle
Discover how a shared virtual machine setup for code models enhances team productivity and accuracy, enabling seamless collaboration,…
Dev tools are all you need
Seattle
Explore Maibook, a desktop app using AI agents personalized from your file system. See customized dev tools for…
Compose Email
Loading recent emails...