Navigating AI's Future: Safety and Governance

Our 123rd episode with a summary and discussion of last week's big AI news!

Read out our text newsletter and comment on the podcast at https://lastweekin.ai/

Email us your questions and feedback at contact@lastweekin.ai

Check out The Gradient here: https://thegradient.pub/

Timestamps:

• (00:00) Intro / Banter
• (04:30) News Preview
• (05:40) Response to listener comments / corrections
• Tools & Apps
  • (07:07) OpenAI rolls out more than 70 ChatGPT plugins including Internet access
  • (11:56) How To Delete Your Data From ChatGPT
  • (17:45) Google is throwing generative AI at everything
  • (23:12) Apple’s new ‘Personal Voice’ feature can create a voice that sounds like you or a loved one in just 15 minutes
  • (25:20) Google develops tool to spot fake AI images after puffer jacket Pope went viral
  • (26:48) Amazon Plans to Add ChatGPT-Style Search to Its Online Store
  • (27:35) W&B Prompts
• Applications & Business 
  • (28:55) Generative AI Breaks The Data Center: Data Center Infrastructure And Operating Costs Projected To Increase To Over $76 Billion By 2028   
  • (31:31) OpenAI CEO Raising $100 Million for Global Crypto Worldcoin 
  • (33:33) OpenAI Confirms ChatGPT Data Breach
  • (35:15) Deal Dive: AI relationship coach Amorai offers more questions than answers
  • (36:50) AI Startup Rewind Gets 170 Offers—and $350 Million Valuation in Unusual Fundraising
  • (38:43) Hippocratic is building a large language model for healthcare
• Projects & Open Source
  • (40:22) EU AI Act To Target US Open Source Software
  • (44:53) Together raises $20M to build open source generative AI models
  • (47:15) Stability AI releases an open source text-to-animation tool
  • (49:19) Listener submission - PandasAI
• Research & Advancements
  • (50:53) PaLM 2 Technical Report
  • (56:48) Microsoft AI Research Introduces Automatic Prompt Optimization (APO): A Simple and General-Purpose Framework for the Automatic Optimization of LLM Prompts
  • (59:51) Threats by artificial intelligence to human health and human existence
  • (01:05:05) Steering GPT-2-XL by adding an activation vector
  • (01:08:45) An Inverse Scaling Law for CLIP Training
  • (01:11:13) Active Retrieval Augmented Generation
  • (01:12:53) Professional Certification Benchmark Dataset: The First 500 Jobs For Large Language Models
  • (01:16:17) FrugalGPT: How to Use Large Language Models While Reducing Cost and Improving Performance
• Policy & Safety 
  • (01:18:37) A.I. ‘controls humanity’ in the worst-case scenario but will probably just find us boring, says Stability AI CEO Emad Mostaque
  • (01:20:37) Hinton issues another AI warning: World needs to find a way to control artificial intelligence
  • (01:26:48) Towards best practices in AGI safety and governance: A survey of expert opinion
  • (01:30:08) The Alan Turing Institute has failed to develop modern AI in the UK
  • (01:32:00) AI voice scams: Report shares 77% of victims lose money, how common it is, and how to protect yourself
  • (01:34:19) Police near Seattle issue warning about AI phone scammers impersonating family members
  • (01:36:01) How to spot AI-generated text
• AI-Generated Media & Art 
  • (01:37:32) Why use of AI is a major sticking point in the ongoing writers' strike
  • (01:39:15) China has its DrakeGPT moment as AI singer goes viral
  • (01:42:16) Hybe uses AI voice technology for Midnatt's multilingual single
  • (01:43:18) Trump shares fake video of Anderson Cooper reacting to CNN town hall
  • (01:46:30) Not even NYT bestsellers are safe from AI cover art
• (01:48:35) Listener Question - more on the writer’s strike(?)
• (01:54:48) Outro