CEO Space Magic • Developing apps since 2010 • build your MVP quickly and validate faster • Proudly launching Răgaz • BIG DREAMS are for small teams
Okay, so this is happening. This begs the question: In terms of app integrations, will there be any demand for broad-scope custom AI assistants (think VS Code's Copilot), or will the use cases of such tools get more specific, such as NPC's in video games or RoomTube's Virtual DJ? I think the former category will slowly become obsolete as we progress into the your-device-can-see-and-hear-everything-that-you-can era, while the latter will still have a place as a way to increase immersion and improve UX.
ChatGPT desktop: debugging with AI takes a step forward I was impressed by Romain Huet's talk at the AI Engineer World's Fair, where he demonstrated the practical power of ChatGPT's new desktop app as a coding assistant. GPT could "see" what he could. The best part was the use of the desktop app's audio/video capabilities which allowed Romain to get real-time feedback and suggestions from as he worked on his code. It felt like watching a developer collaborate with a teammate, only this teammate was AI. It demonstrated how LLMs + multi-modal support can be a valuable resource for not just brainstorming ideas or generating code, but also for just getting real-time feedback and refining existing code. An advanced rubber duck of sorts. This fits well into the role of AI as an "assistant" for developers. I think the potential applications here are many. Check out the video! #softwareengineering #programming #ai