ChatGPT Training, Superintelligence, & AI Funding in Vector Databases | E12

In this episode we cover OpenAI's new update to hide chat history or accept they will train using your inputs, we explore doomsday scenarios and cover Max Tegmark's TIME article on "The Don't Look Up Thinking That Could Doom Us All", discuss the explosion in Vector database funding and what it means and learn how large language models get anxious! CHAPTERS: ==== 00:00 - What have we done!? 00:17 - OpenAI Chat History, Privacy & Training GPT-5 09:45 - Where is GPT-4 with Images? & OpenAI Enterprise Deals 12:52 - Max Tegmark's TIME article & Superintelligence, Doomsdaying & Fear 29:23 - Anxiety in AI Models, AI "Emotions" and Motivation 37:33 - AI Distribution: How Much is AI Changing our Lives? AI Hype Cycle 46:01 - Vector Database Funding, Increasing Prompt Sizes, Scaling Transformer to 1M Tokens & Beyond Paper 55:25 - VCs Struggling to Know What & Where to Invest in AI 1:00:39 - Segment Anything in Medical Images & Medical Funding 1:03:20 - Weed Zapping AI Stopping Pesticide Use SOURCES: ==== https://openai.com/blog/new-ways-to-manage-your-data-in-chatgpt https://twitter.com/gdb/status/1651306937991168002?s=46&t=uXHUN4Glah4CaV-g2czc6Q https://time.com/6273743/thinking-that-could-doom-us-with-ai/ https://twitter.com/fchollet/status/1650738654061658112?s=46&t=uXHUN4Glah4CaV-g2czc6Q https://twitter.com/aisolopreneur/status/1646928323363102734?s=46&t=uXHUN4Glah4CaV-g2czc6Q https://twitter.com/sterlingcrispin/status/1650320043107061761?s=46&t=uXHUN4Glah4CaV-g2czc6Q https://www.pinecone.io/learn/series-b/ https://arxiv.org/pdf/2304.11062.pdf https://arxiv.org/pdf/2304.11111.pdf https://arxiv.org/pdf/2304.12306v1.pdf https://twitter.com/Rainmaker1973/status/1649743415549067267 If you like this podcast please consider leaving a review or sharing with a friend.