Contact us: https://bit.ly/4fAFrVx
Can AI build a text-based adventure game from scratch? In this hands-on session, TJ Gerken, prototype development lead at Cyberdyne Systems, puts three AI agents—ChatGPT, Claude, and “Alex” (spoiler: not human)—to the test.
00:00 | Introduction
00:25 | Meet TJ Gokken and the experiment setup
01:35 | What we’re building: a text-based adventure game
03:03 | Using ChatGPT with .NET and C#
04:09 | ChatGPT struggles with consistency
06:09 | Improving the prompt for better results
09:06 | Switching to Claude for F# development
10:08 | How Claude handles prompts differently
11:30 | Comparing ChatGPT’s and Claude’s code
13:06 | Running and evaluating the Claude solution
19:22 | Surprise reveal: Alex isn’t human
21:24 | TJ compares all 3 results
27:21 | Deep dive: monadic error handling & F# pattern matching
30:33 | Final reflections on prompt quality and AI limitations
32:01 | The fast-changing AI ecosystem
37:04 | Audience Q&A on tooling, efficiency, and senior vs junior devs
He challenges each to create a working game using different levels of prompt guidance and programming languages. Along the way, TJ shares what worked, what failed, and the surprising differences in how these tools behave when left to their own logic. From prompt crafting to debugging strategies, this is a must-watch for developers exploring AI-assisted software development.
About SSW: https://linktr.ee/SSWTV
Links:
LinkedIn: / ssw
Twitter: / ssw_tv
Facebook: / ssw.page
Instagram: / ssw_tv
Tiktok: / ssw_tv
Get in touch: TJ Gokken – Enterprise AI/ML | Quantum Computing | Software Engineering
About: https://tjgokken.com
LinkedIn: / tjgokken
#aiwars#aicoding#fsharp#chatgpt#claudeai#openai#promptengineering#gamedev#dotnet#sswrules#developers
V1 – Bahjat
About the speakers

TJ Gokken
Enterprise AI/ML Integration Engineer building scalable solutions across platforms. Bridging traditional software systems with modern AI. Writing about practical ML/AI integration patterns at tjgokken.com