Newsroom Robots Simon Willison: OpenAI's New Features & Security Risks of Large Language Models (Part 1)
4 snips
Nov 24, 2023 
Simon Willison, creator of Datasette, discusses OpenAI's new features and the security risks of large language models. Topics include recent events at OpenAI, concerns about feature rollout and reputational risk, the uncertainty of OpenAI and the evolution of generative AI, OpenAI's dev day conference and new features, challenges with creating chatbots using custom GPT, and exploring the use of GPTs for newsrooms.
AI Snips
Chapters
Transcript
Episode notes
Run Models Locally For Sensitive Data
- Try local LLMs for sensitive data by running models on your own laptop or phone to avoid cloud exposure.
- Use tools like Ollama, GPT4All or local libraries to download and run models offline where feasible.
Bigger Context Enables New Tasks
- GPT-4 Turbo cut cost and raised token limits dramatically, enabling processing of much longer documents.
- Vast token context unlocks new workflows like multi-article synthesis and long transcript analysis.
Transcribe Locally With Whisper
- Use Whisper for high-quality local transcription and run it on your own machine for interview and meeting workflows.
- Transcribe audio locally to speed reporting and preserve control over source recordings.