Tech Brew Ride Home Wed. 01/06 – New GPT-3 Models Can Generate Images From Text
Jan 6, 2021
New AI models are reshaping graphic arts by generating images from text, posing a challenge for artists. Tensions rise as Trump’s tech cold war effort targets Chinese payment apps, while privacy concerns loom over WhatsApp’s updated data-sharing practices. The iPhone 12 mini’s disappointing sales raise questions about the future of compact smartphones. Meanwhile, gaming enthusiasts in EVE Online push boundaries, even breaking the game, and Cameo capitalizes on personalized celebrity messages, showing innovative monetization in the creator economy.
AI Snips
Chapters
Transcript
Episode notes
GPT-3 Advancements
- OpenAI's new GPT-3 models, CLIP and DALL-E, link text and images.
- This helps language models grasp everyday concepts better.
DALL-E's Artistic Prowess
- Sam Altman used DALL-E to create an image of a baby shark wizard with a lightsaber.
- The tool's capabilities sparked discussions about AI's impact on artistic jobs.
Inconsistent China Policy
- The Trump administration's tech policy towards China has been inconsistent.
- This makes it challenging to form an informed opinion on the issue.