Let’s build an AI-powered API
summary
Builds a working sentiment-moderation API on top of OpenAI GPT-3 using prompt engineering, the Playground to train examples, and the Playground’s “generate code” button to emit the request snippet; wraps that in a 31-line Flask prototype that calls GPT-3 for each review and returns labels like positive/negative, tests via Postman/curl on localhost, and relies on an uses OPENAI_API_KEY env var (so keys stay out of code); notes limits: cloud-only, per-call latency and cost.
AI-written summary · about
chapters (18)
- 0:00Introduction
- 0:30Welcome
- 3:56Agenda
- 4:33AI vs Machine Learning
- 5:14Art Gallery
- 6:24GPT
- 7:19Prompt Engineering
- 7:46Use Case
- 8:21Sentiment Analysis
- 8:51The API
- 10:00Training the model
- 11:50Generate Code
- 15:29Git Handle
- 16:01OpenAI
- 16:52Electric Minds
- 18:30Conversational chatbots
- 20:52How important is phrasing
- 23:16Future use cases
related talks
01h 00mGoing beyond ChatGPT: introduction to prompt engineering & LLMs
32mIntegrating generative AI into API Platform: Good idea?
32mHow to Make AI Work for You: A Deep Dive Guide into ChatGPT for IT Professionals
45mBringing the Power of AI to Your Application
26mUnleash the Power of AI Craft Your Magic with ChatGPT API!
51mOpenAI and the AI that may overtake humans
21mGPT-3 wrote the description for this talk... Scary or exciting?
24mInfuse Generative AI in your apps using