Custom GPTs for Fun, Profit, and Potential Liability

on YouTube ↗01h 01m

" summary

Shows how to build a working custom GPT two ways: the no-code GPT builder that uploads up to ten files (≈20MB each) and the programmatic Assistants API used with Java, Spring Boot and an OpenAPI action. Demonstrates a live integration: OpenWeatherMap via Spring Boot deployed to Heroku returning current Bangalore weather, DALL·E3 icons and code-interpreter images, and discusses RAG vs large-context tradeoffs (context windows from 16k to 1M tokens), costs ($10 per million tokens) and the stalled GPT-store monetization plus real liability risk from misbehaving chatbots.

AI-written summary · about

" chapters (6)

1. 0:00Introduction and perspective
2. 2:29What are custom GPTs
3. 18:10Uploading files and RAG
4. 32:40Actions and external APIs
5. 51:55Monetization and limitations
6. 57:57Conclusions and liability risks

" related talks

• 01h 09mChatGPT and OpenAI API solutions: 👑 My Favourite ImplementationsAdam Cogan · NDC Oslo 2024
• 39mBuilding Your Own Bot with GPT: Harnessing the Power of AI...Alon Kiriati · GeeCON 2023
• 45mHow to use GPT in your projects: A Case StudyStephan Janssen · Devoxx France 2023
• 38mIntroducing the New Colleague - ChatGPTPeter Sabev · TestCon Europe 2023
• 30mHow To Build Your Own Internal ChatGPTVictoria Albrecht · Techsylvania 2023
• 01h 12mChatGPT and OpenAI API solutions: 👑 My Favourite ImplementationsAdam Cogan · NDC Sydney 2024
• 01h 00mGoing beyond ChatGPT: introduction to prompt engineering & LLMsTies de Kok · PyData Seattle 2023
• 01h 16mChatGPT - and How we Can Benefit from itSanjay Sauldie · Digital Marketing Europe 2023