Members-Only
Recent Talks & Demos are for members only
You must be an AI Tinkerers active member to view these talks and demos.
Artificial Societies: LLM Agent Simulation
Demonstrates an LLM‑driven simulation of a LinkedIn audience, showing how model‑generated agents predict and explain recent post performance dynamics in real‑time.
Artificial Societies uses LLMs to simulate large groups of people and how they react to information. Here, we demo a simulation of the host Luke Harries’ LinkedIn audience, and see if we can accurately simulate how his recent posts perform.
Optimizes content using simulated audience networks for data-driven insights.
Related projects
The next evolution of AI - Neurosymbolic
London
Demonstration of a lightweight neurosymbolic platform that maps data, lets you define business logic in plain English, and…
AI-Powered Prompt Engineering: Enhancing LLM Performance with PromptLab and IQ
Boston
Live demo shows PromptLab and IQ optimizing Apple Intelligence prompts, demonstrating real-time prompt refinement across multiple LLMs for…
Twitter '95 (and Fireside Chat)
New York City
A prototype of 1995‑style Twitter built with LLM‑generated posts and historically grounded content, demonstrating early‑web aesthetics and AI‑driven…
Human machine
Seattle
This talk explores a code editor and runtime that integrates multiple AIs for prototyping, enabling coding with voice…
SMELL: A Framework For Aligning LLM Evaluators To Human Feedback
London
The talk introduces SMELL, a practical framework that aligns large language model evaluators with human feedback, detailing methodology,…
Conversational AI recruiting
New York City
This talk demonstrates using conversational AI and LLMs to automate candidate phone screens by generating questions, conducting calls,…