

EP 279: Google’s New AI Updates from I/O: the good, the bad, and the WTF
May 23, 2024
Google's latest I/O conference was a whirlwind of AI updates and new products. From the impressive advancements of Project Astra to the upgraded Gemini models, the impact on users is significant. Concerns arise over feature accessibility and marketing transparency. The hosts dive into doubts about Google's Gemini claims and critique the buzzword-heavy approach from executives. Meanwhile, innovations like GEMS promise to change interactions with AI, but not without raising questions about usability for everyday folks.
AI Snips
Chapters
Transcript
Episode notes
Deceptive Marketing
- Google's Gemini demo in December was misleading.
- It presented real-time interaction, but was actually pre-recorded with human input.
Project Astra Demo
- Project Astra demo showed real-time interaction with objects and code.
- A user asked questions about a speaker, crayons, and code displayed on screen.
Astra vs. Omni
- Project Astra's real-time capabilities look impressive, potentially exceeding OpenAI's GPT-4 Live Omni.
- Google's history of misleading demos warrants skepticism.