This week I gave a presentation, demo and participated in Live Q&A at the Embarcadero CodeRage 2025 online developer conference. My session showed how to use AI tools including ChatGPT v5, Claude Sonnet v4.5, Gemini v3 to build the project and code for a Multi-Device Solar System Text Adventure game using Embarcadero Delphi 13 and the FireMonkey (FMX) cross platform framework. Using FireMonkey allows you to create applications for Windows, macOS, iOS, Android and Linux using the same project code.
My session was at CodeRage 2025 Day 2 – Tuesday December 2, 2025 at 15:00 CST (21:00 GMT). The session titled is “Using Modern AI tooling to Create a Working Delphi Multi-Device Application”. The session description is as follows:
“Wherever developers look they can read about modern AI’s impact on software development. How much time should a developer spend trying to separate the ground breaking potential, confusing hype and rapid advances in the usefulness and quality of development tools? The only way I know, for myself, is to keep up to date and try to build applications from scratch using as many tools, APIs, libraries, components and 50+ years of experience with computer science and software engineering. In this talk, I will use one (or more) of the (currently) top modern AI coding solutions to build a working Delphi multi-device application using RAD Studio 13. Join me for my development (Mister Toad’s) wild ride and travel down the (Alice’s) rabbit hole.”
The YouTube URLs and GitHub Respository for my session
Here are the links for CodeRage 2025, my session and my code repository. Note: my session is part of YouTube replay is for Day 2 and I provide the URLs with YouTube time offsets for me session.
- CodeRage 2025 with Ian Barker YouTube Playlist for all days:
https://www.youtube.com/playlist?list=PLwUPJvR9mZHhSM9scEmVLcmrRUqucl6BA - Solar System Text Adventure GitHub Repository containing code, chat transcripts, slides, etc.:
https://github.com/davidi99/SolarSystemTextAdventure - Start of my session YouTube URL with time offset:
https://youtu.be/Nw0JVNVNGeo?list=PLwUPJvR9mZHhSM9scEmVLcmrRUqucl6BA&t=17969 - Presentation/Demos starts at YouTube URL with time offset
https://youtu.be/Nw0JVNVNGeo?list=PLwUPJvR9mZHhSM9scEmVLcmrRUqucl6BA&t=18051
Note: presentation/demos runs from time 5:00:49 to 5:49:08 - Live Q&A starts at YouTube URL with time offset
https://youtu.be/Nw0JVNVNGeo?list=PLwUPJvR9mZHhSM9scEmVLcmrRUqucl6BA&t=20413
Note: Live Q&A runs from time 5:40:25 to 6:14:30
Building the Solar System Text Adventure project
For the presentation I built and tested the generated projects and code using a licensed version of Delphi 13 for the Windows platform. I only used projects that were generated completely by ChatGPT v5, Claude Sonnet x4.5 and Gemini v3. I will also update the GitHub repository after I complete deployment and testing on additional platforms.
Disclaimer – I created this presentation only using the LLM chat interface for for each modern AI tool. Normally, I use the Embarcadero RAD Studio / Delphi IDE which has plug-ins and components that let you do project development and code completion. For this presentation I wanted to see if I could use only the LLMs themselves to create the application. The GitHub repo includes the prompts and complete conversations I had with each of the AI tools.
For additional information about how AI is integrated into the RAD Studio / Delphi IDE you can find additional information in the blog post at https://blogs.embarcadero.com/introducing-the-smartcore-ai-components-pack/