# GPT-4 Projects at AI Tinkerers

> Canonical HTML: https://aitinkerers.org/technologies/gpt-4?page=19
> Markdown URL: https://aitinkerers.org/technologies/gpt-4.md?page=19
> Technology record last updated: 2026-02-22T10:50:14Z
> Generated: 2026-09-21T08:53:08Z

GPT-4 is OpenAI’s large multimodal model: it processes both text and image inputs, delivering human-level performance on complex professional and academic benchmarks.

This is OpenAI’s latest milestone in scaling deep learning: a large multimodal model accepting both text and image inputs. It demonstrates a significant capability leap over its predecessor, scoring in the top 10% on a simulated bar exam (GPT-3.5 scored in the bottom 10%). The model handles nuanced instructions and long-form content, supporting context windows up to 32,768 tokens (32K model). This capacity allows processing up to 25,000 words in a single, complex prompt. GPT-4 is engineered for enhanced reliability, steerability, and advanced reasoning across diverse tasks.

- Official technology site: https://platform.openai.com/docs/models/gpt-4
- Public AI Tinkerers demos and talks: 529
- Result page: 19 of 23

## Recent Public Talks and Demos

### [Invention Cards enhanced by AI](https://seattle.aitinkerers.org/talks/rsvp_swW8heNcmQI)

I used GPT4's multimodal features to accelerate and improve my recently rebooted Visual Chronology of Science & Discovery project now hosted at https://invention.cards. First I used image-to-text models to extract and structure content from pages of Asimov's encyclopedia for accelerated data entry. Then I used text-to-image models to generate visually consistent imagery for each invention and discovery card. This post describes both workflows and my findings. Relevant source code is available in my asimov-gpt repository.

- Event context: AI Tinkerers Seattle - July 2024 Meetup — 2024-07-12 — Seattle
- Public talk page: https://seattle.aitinkerers.org/talks/rsvp_swW8heNcmQI

### [ExamGPT - An AI tool to help you prepare for your exam](https://seattle.aitinkerers.org/talks/rsvp_BvyjJi6DNdk)

Using flash cards to prepare for an exam is an incredibly useful technique, but creating flash cards is an absolute pain! So I created this application that uses AI to do it for you. It uses openai to generate the flash cards after chunking the content. The biggest problem was weeding out chunks that contain content that don't make for good flash cards (eg table of content, or page containing information about the exam itself). PDF parsing still seems very rudimentary in python. The live demo would be similar to the youtube video I linked to.

- Event context: AI Tinkerers Seattle - July 2024 Meetup — 2024-07-12 — Seattle
- Public talk page: https://seattle.aitinkerers.org/talks/rsvp_BvyjJi6DNdk

### [BAML: Using algorithms to guarantee JSON](https://seattle.aitinkerers.org/talks/rsvp_vCVU8mkZnms)

When using an LLM, its important to be able to parse the response reliably, many times JSON.parse doesn't work due to missing quotes, new lines, or when it does work, the exact schema may not be followed. You could have wanted an array of numbers, but gotten just 1. Using knowledge of your schema, BAML fixes those errors for you and provide a layer of guarantee, it will either be your schema or an exception will be raised. (0 additional LLM calls, runs locally in the package, built in rust) During the talk we'll see: 1. a few examples of BAML fixing the errors 2. walkthrough parts of the algorithm 3. Talk about edge cases like markdown and missing commas and how we address them

- Event context: AI Tinkerers Seattle - July 2024 Meetup — 2024-07-12 — Seattle
- Public talk page: https://seattle.aitinkerers.org/talks/rsvp_vCVU8mkZnms

### [Tinkering and recreating parts of Claude Artifacts through Claude 3.5 playground](https://bengaluru.aitinkerers.org/talks/rsvp_jcmUaDP8GD0)

The recent launch of Claude 3.5 had this cool interaction playground that ran react code, ran html, markdown, python output. In this demo, I want to show using the react code running part and fully doing most of the coding through the Claude 3.5 alone.

- Event context: AI Tinkerers Bangalore - July - RSVP REQUIRED — 2024-07-05 — Bengaluru
- Public talk page: https://bengaluru.aitinkerers.org/talks/rsvp_jcmUaDP8GD0

### [Collaborating Agents](https://toronto.aitinkerers.org/talks/rsvp_MrhTOEptmUo)

In a world where AI agents increasingly drive decision-making, trust is paramount. This demo explores teams of collaborating agents, and how blockchain technology can address the trust problem in AI. By leveraging blockchain's decentralized and immutable ledger, we can ensure transparency, accountability, and security in AI operations. Discover how these technologies converge to create a future where AI decisions are transparent and verifiable, fostering trust between technology and humans.

- Event context: AI Tinkerers - Toronto - June 2024 Meetup — 2024-06-27 — Toronto
- Public talk page: https://toronto.aitinkerers.org/talks/rsvp_MrhTOEptmUo

### [Unlocking Insights from Tabular Data with LLMs](https://toronto.aitinkerers.org/talks/rsvp_EVqtcfX0ujE)

This application demonstrates how we can leverage Large Language Models (LLMs) to extract valuable insights directly from your company's tabular data. Unlike unstructured data (PDFs, Docs etc.) that require Retrieval-Augmented Generation (RAG) techniques, this approach allows LLMs to interact with structured data in a more efficient way. Here's how it works: Natural Language Understanding with LLM: The user submits a question in plain text. The LLM parses this question to understand its intent and the relevant data points needed. SQL Query Generation: Based on the understanding, the LLM generates a corresponding SQL query to retrieve the desired information from your company's data warehouse. Data Retrieval and Summarization: The generated SQL query is executed, fetching the relevant data points. Finally, the LLM summarizes the retrieved data in a clear and concise manner, providing a comprehensive answer to the user's initial question.

- Event context: AI Tinkerers - Toronto - June 2024 Meetup — 2024-06-27 — Toronto
- Public talk page: https://toronto.aitinkerers.org/talks/rsvp_EVqtcfX0ujE

### [OpenSesame Demo](https://toronto.aitinkerers.org/talks/rsvp_7C5CZ6quT8M)

OpenSesame is a plug-and-play API that detects AI hallucinations and improves response accuracy for companies using LLMs. Our platform leverages context-based information to detect hallucinations in LLMs with precision. By ensuring accurate and error-free text generation, we provide a reliable solution for enhancing AI outcomes.

- Event context: AI Tinkerers - Toronto - June 2024 Meetup — 2024-06-27 — Toronto
- Public talk page: https://toronto.aitinkerers.org/talks/rsvp_7C5CZ6quT8M

### [AI-powered Legal Contract Analysis POC](https://kuala-lumpur.aitinkerers.org/talks/rsvp_SRNn5lVdOsI)

We are using LLMs to help do legal contract analysis in the following areas: 1. Clause Identification (and custom Key clause identification) 2. Clause Recommendation 3. RAG over Contract

- Event context: AI Tinkerers Kuala Lumpur - June 27th — 2024-06-27 — Kuala Lumpur
- Public talk page: https://kuala-lumpur.aitinkerers.org/talks/rsvp_SRNn5lVdOsI

### [Chess Predict](https://la.aitinkerers.org/talks/rsvp_JloIyZD7310)

Chess Predict is a software the predicts the best move given a digital screenshot of a chess game. Improvements to the software include an additional smart contract feature which combats cheating. I will demo this functionality.

- Event context: June 25th - LA AI Tinkerers Meetup & Demos — 2024-06-26 — Los Angeles
- Public talk page: https://la.aitinkerers.org/talks/rsvp_JloIyZD7310

### [Anatomy of a Thinking Machine](https://la.aitinkerers.org/talks/rsvp_MjNJg6eHsLw)

We all know that AI has a hardware problem. - What happens at a hardware level during inference? - What exactly are all these tools in the inference ecosystems from Nvidia, AMD...? - An early preview of Cortex, an open source tool that runs LLMs across multiple platforms Here's my cofounder Dan doing a similar talk, but I'm planning for this demo to be **shorter & purely technical**: https://www.youtube.com/watch?v=orcPcUzSbOw&ab_channel=HackerHouseTW

- Event context: June 25th - LA AI Tinkerers Meetup & Demos — 2024-06-26 — Los Angeles
- Public talk page: https://la.aitinkerers.org/talks/rsvp_MjNJg6eHsLw

### [Scripter Studio -- Founders University Cohort 8](https://la.aitinkerers.org/talks/rsvp_tSKu0qwQ1JA)

An Ai powered tool that assists the Hollywood industry in script scheduling, coverage, analysis

- Event context: June 25th - LA AI Tinkerers Meetup & Demos — 2024-06-26 — Los Angeles
- Public talk page: https://la.aitinkerers.org/talks/rsvp_tSKu0qwQ1JA

### [LLM-powered Knowledge Graph Generation](https://boston.aitinkerers.org/talks/rsvp_DVbe4XH70-Q)

We'll be running code to test how well LLMs can build knowledge graphs from scratch starting from a single node.

- Event context: June 2024 Meetup at C10 Labs — 2024-06-10 — Boston
- Public talk page: https://boston.aitinkerers.org/talks/rsvp_DVbe4XH70-Q

### [WorldEnder.ai](https://seattle.aitinkerers.org/talks/rsvp_OHedJydXeXo)

I am developing a text-based adventure RPG that uses LLM as the interactive "engine" for the game. Builds upon other research in the field and develops completely new ideas about how to use structured output from LLMs in game development.

- Event context: AI Tinkerers Seattle - June 2024 — 2024-06-07 — Seattle
- Public talk page: https://seattle.aitinkerers.org/talks/rsvp_OHedJydXeXo

### [Pick-Em's Bot](https://seattle.aitinkerers.org/talks/rsvp_H6MqDp0e2zM)

This is an LLM agent built that scrapes sports articles and team stats to predict winners for ESPN's Pick-Em's game. For the 2023 NFL season, this agent won my local group's pick-em's competition and came in the top 15% across all of ESPN's users (~one million users played). The top player across all of ESPN for 2023 correctly chose 195 games (~71%). This agent correctly chose 163 (~60%).

- Event context: AI Tinkerers Seattle - June 2024 — 2024-06-07 — Seattle
- Public talk page: https://seattle.aitinkerers.org/talks/rsvp_H6MqDp0e2zM

### [LLM-powered Knowledge Graph Generation](https://nyc.aitinkerers.org/talks/rsvp_hauOLk3d7qc)

Generalized LLMs have made tasks like entity tagging and relationship extraction from unstructured text much easier. This enables developers to create structured networks that are important for relational insight out of unstructured data.

- Event context: Omakase Founder Dinner | NY #TechWeek — 2024-06-04 — New York City
- Public talk page: https://nyc.aitinkerers.org/talks/rsvp_hauOLk3d7qc

### [Ellipsis: Fixing build failures on pull requests](https://nyc.aitinkerers.org/talks/rsvp_Amwi91giatU)

You just opened a pull request, but you forgot an import and the linter is failing. What now? Let's see how GPT4 can aid developer productivity but automatically fixing a build issue on a GitHub pull request. Background: In August 2023, I demo'd an "AI junior software engineer" at the first AI Tinkeres x NYC meet up. It made a great demo, but I quickly realized that state-of-the-art LLM's aren't good enough to reliably make nontrivial changes to production-scale codebases. So I scaled down the problem. Today, Ellipsis is a developer tool that reviews pull requests and fixes bugs on them. One of our most requested features is having Ellipsis automatically resolve CICD failures.

- Event context: Omakase Founder Dinner | NY #TechWeek — 2024-06-04 — New York City
- Public talk page: https://nyc.aitinkerers.org/talks/rsvp_Amwi91giatU

### [AI to redefine surgery](https://nyc.aitinkerers.org/talks/rsvp_WJWUC5ydMz8)

We are developing AI to track, analyze and improve surgery to prevent a billion surgical complications over the next 10 years.

- Event context: Omakase Founder Dinner | NY #TechWeek — 2024-06-04 — New York City
- Public talk page: https://nyc.aitinkerers.org/talks/rsvp_WJWUC5ydMz8

### [momentum - ship your backend with confidence](https://bengaluru.aitinkerers.org/talks/rsvp_ZFaNolQ0otk)

momentum is a code auditor that analyses the necessary code behavior and tests it at every git push to ensure the code is ready for production.

- Event context: AI Tinkerers - Bangalore Inaugural - RSVP REQUIRED — 2024-06-02 — Bengaluru
- Public talk page: https://bengaluru.aitinkerers.org/talks/rsvp_ZFaNolQ0otk

### [THub - A Low Code No Code Platform for building GenAI apps](https://bengaluru.aitinkerers.org/talks/rsvp_0HPA8fw99z4)

THub is a low code no code for building Generative AI apps like agents, retrieval augmented generation, recommender system, retrieval augmented fine tunning, Q&A Chatbot etc with simple drag and drop features. THub supports all the major Large Language Model like OpenAI GPT's 3.5/4.4-o, Gemini 1.5 pro, Anthropic Claude 3, Llama 3 etc, support various embedding models, supports vector stores like Pinecone, Qdrant, orchestration & development framework like langchain & llamaindex etc

- Event context: AI Tinkerers - Bangalore Inaugural - RSVP REQUIRED — 2024-06-02 — Bengaluru
- Public talk page: https://bengaluru.aitinkerers.org/talks/rsvp_0HPA8fw99z4

### [App de Cuentos para la Hora de Dormir usando IA](https://bogota.aitinkerers.org/talks/rsvp_qnSvEzeFN08)

Utilizando tecnologías basadas en IA como OpenAI (ChatGPT 4.0), StableDiffusion, Deepgram, entre otras, se procede a la creación de un cuento para dormir que pueden disfrutar niños entre los 2 y 10 años.

- Event context: Primer AI Tinkerers Bogotá — 2024-05-30 — Bogotá
- Public talk page: https://bogota.aitinkerers.org/talks/rsvp_qnSvEzeFN08

### [KathaBook](https://toronto.aitinkerers.org/talks/rsvp_nskL0vxWd0g)

KathaBook is an app to create personalized storybooks in the form of narrative videos. These can be used as gifts for loved ones, visual bedtime stories for children, and to archive family histories for future generations.

- Event context: AI Tinkerers - Toronto - May 2024 Meetup — 2024-05-30 — Toronto
- Public talk page: https://toronto.aitinkerers.org/talks/rsvp_nskL0vxWd0g

### [Burgundy - Content Idea Copilot](https://toronto.aitinkerers.org/talks/rsvp_mLxs1evmaKo)

AskBurgundy.com also features an advanced AI infrastructure that provides content ideas and facilitates writing up to 20-30 times faster than conventional systems. This capability significantly enhances the speed and efficiency of content generation and marketing strategies.

- Event context: AI Tinkerers - Toronto - May 2024 Meetup — 2024-05-30 — Toronto
- Public talk page: https://toronto.aitinkerers.org/talks/rsvp_mLxs1evmaKo

### [Reliable AI Agents](https://nyc.aitinkerers.org/talks/rsvp_TE3Go21sxYY)

AI agent demos based on LLMs are so common, but as Tinkerers we know they have many pitfalls around accuracy and hallucination. If we alter the prompt just slightly, the output completely changes! Grounded in years of dialogue systems research, I will be presenting the latest features in an agent built using classical approaches to Conversational AI. Rather than LLMs+RAG, this agent has an special NLU module (natural language understanding) that explicitly predicts and tracks specific named entities in the conversation, allowing for unprecedented control and reliability.

- Event context: AI Tinkerers NYC May 2024 Meetup — 2024-05-22 — New York City
- Public talk page: https://nyc.aitinkerers.org/talks/rsvp_TE3Go21sxYY

### [Twitter '95 (and Fireside Chat)](https://nyc.aitinkerers.org/talks/rsvp_wlGwEKyF4lw)

This project imagines what Twitter would be like if it had been invented in 1995, using LLMs to mimic social media participants and historical information to inject grounding.

- Event context: AI Tinkerers NYC May 2024 Meetup — 2024-05-22 — New York City
- Public talk page: https://nyc.aitinkerers.org/talks/rsvp_wlGwEKyF4lw

## Related Technologies

- [LangChain](https://aitinkerers.org/technologies/langchain) ([Markdown](https://aitinkerers.org/technologies/langchain.md)) — 445 public demos
- [OpenAI API](https://aitinkerers.org/technologies/openai-api) ([Markdown](https://aitinkerers.org/technologies/openai-api.md)) — 520 public demos
- [Llama-2](https://aitinkerers.org/technologies/llama-2) ([Markdown](https://aitinkerers.org/technologies/llama-2.md)) — 227 public demos
- [GPT-3](https://aitinkerers.org/technologies/gpt-3) ([Markdown](https://aitinkerers.org/technologies/gpt-3.md)) — 191 public demos
- [PyTorch](https://aitinkerers.org/technologies/pytorch) ([Markdown](https://aitinkerers.org/technologies/pytorch.md)) — 273 public demos
- [BERT](https://aitinkerers.org/technologies/bert) ([Markdown](https://aitinkerers.org/technologies/bert.md)) — 179 public demos
- [Python](https://aitinkerers.org/technologies/python) ([Markdown](https://aitinkerers.org/technologies/python.md)) — 662 public demos
- [RoBERTa](https://aitinkerers.org/technologies/roberta) ([Markdown](https://aitinkerers.org/technologies/roberta.md)) — 118 public demos
- [BLOOM](https://aitinkerers.org/technologies/bloom) ([Markdown](https://aitinkerers.org/technologies/bloom.md)) — 115 public demos
- [PaLM 2](https://aitinkerers.org/technologies/palm-2) ([Markdown](https://aitinkerers.org/technologies/palm-2.md)) — 116 public demos
- [Claude-3](https://aitinkerers.org/technologies/claude-3) ([Markdown](https://aitinkerers.org/technologies/claude-3.md)) — 110 public demos
- [Transformers](https://aitinkerers.org/technologies/transformers) ([Markdown](https://aitinkerers.org/technologies/transformers.md)) — 148 public demos
- [TensorFlow](https://aitinkerers.org/technologies/tensorflow) ([Markdown](https://aitinkerers.org/technologies/tensorflow.md)) — 90 public demos
- [scikit-learn](https://aitinkerers.org/technologies/scikit-learn) ([Markdown](https://aitinkerers.org/technologies/scikit-learn.md)) — 84 public demos
- [Keras](https://aitinkerers.org/technologies/keras) ([Markdown](https://aitinkerers.org/technologies/keras.md)) — 74 public demos
- [ONNX](https://aitinkerers.org/technologies/onnx) ([Markdown](https://aitinkerers.org/technologies/onnx.md)) — 83 public demos
- [Streamlit](https://aitinkerers.org/technologies/streamlit) ([Markdown](https://aitinkerers.org/technologies/streamlit.md)) — 89 public demos
- [RAG](https://aitinkerers.org/technologies/rag) ([Markdown](https://aitinkerers.org/technologies/rag.md)) — 147 public demos

## More Results

- Previous: https://aitinkerers.org/technologies/gpt-4.md?page=18
- Next: https://aitinkerers.org/technologies/gpt-4.md?page=20