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.