#114 - ChatGPT applications, Claude, PALM-E, OpenAI criticism, AI-generated spam
Summary
Hosts Andrei Krenkov and Jeremy Harris discuss the surge of AI integrations in apps like Slack and Snapchat, alongside Anthropic’s $1 billion raise and its novel Constitutional AI safety approach. They analyze OpenAI’s cheap new API, Microsoft’s Bing updates, and Stability AI’s valuation. The episode also covers Google’s PaLM-E robot brain and research into energy-efficient optical neural networks for transformers.
Topics discussed
Sponsors: OutShift by Cisco and Box
AI App Roundup: MyAI, ShortWave, and the 'Technology Overhang'
Anthropic's Claude and Constitutional AI Safety Research
OpenAI Price Cuts, Whisper, and Bing Chat Tone Controls
Stability AI Raises $101M at $1B Valuation
Google's PaLM-E and DeepMind's Gato Multimodal Models
Optical Computing, Neuromorphic Chips, and Biological AI
Scaling Laws, Atari Learning, and AI Reasoning Flaws
OpenAI's Shift to Capped-For-Profit Structure
EU AI Act, Regulation Challenges, and AI Patent Rights
AI Voice Scams and Beauty Filters on Social Media
AI Spam in Publishing, Filmmaking, and AI Seinfeld Spoof
Listen ad-free on Castria