The second episode of "The AI Artifacts Podcast" welcomes Thomas Huitric as a guest. Thomas is a data science practitioner with experience in AI infrastructure across advertising, media, gaming, and the metaverse. He joins the podcast to talk about what’s going on behind the screen for users as AI scales up and its resource, infrastructure, and data needs grow.
In addition, co-hosts Brian and Sarah dig into the past week’s big stories about The Beatles, President Biden’s new executive order, and more. Listen in, play this week’s round of "Two truths and l’AI," and share your thoughts on AIArtifacts.net.
Timestamps for this episode:
[0:53] This week’s biggest AI news
[1:00] The Beatles release "Now And Then" track produced using AI
[5:21] U.S. President Joseph Biden’s new executive order on AI
[10:50] "Bletchley Declaration" on AI safety published in Britain
[14:10] Google to invest up to $2 billion in Anthropic
[17:40] Google lays off voice assistant team
[18:59] Microsoft’s new 365 Copilot could generate $10 billion/year in revenue
[19:44] "Two truths and l’AI"
[25:08] Interview with Thomas Huitric
[27:03] What developments made today’s AI technology possible?
[29:27] Cloud vs. local processing for AI applications
[32:35] Apple reportedly spending $1 billion/year on AI to catch up to others
[38:30] Microsoft’s water usage in Iowa while training ChatGPT-4
[43:30] How much data will be required to create the next version of ChatGPT
[49:10] Open-source vs. proprietary AI
[52:48] Meta’s approach to Llama 2
[54:44] Were big tech companies actually surprised by OpenAI’s ChatGPT?
[59:37] Reasons for optimism in AI and LLMs
[1:02:25] Thanks, sign-off, and outro
Links for topics referenced in this episode:
New Beatles song "Now And Then" released (CNBC) https://www.cnbc.com/2023/11/02/a-new-beatles-song-is-set-for-release-after-45-years-with-help-from-ai.html
Official music video for "Now And Then" by The Beatles (YouTube)
President Biden’s new AI executive order (The New York Times): https://www.nytimes.com/2023/10/30/us/politics/biden-artificial-intelligence.html
Text from White House briefing room explaining Biden’s AI executive order (The White House): https://www.whitehouse.gov/briefing-room/presidential-actions/2023/10/30/executive-order-on-the-safe-secure-and-trustworthy-development-and-use-of-artificial-intelligence/
Full text from the Bletchley Declaration (Gov.uk): https://www.gov.uk/government/publications/ai-safety-summit-2023-the-bletchley-declaration/the-bletchley-declaration-by-countries-attending-the-ai-safety-summit-1-2-november-2023
Google invests in OpenAI rival Anthropic (Reuters): https://www.reuters.com/technology/google-agrees-invest-up-2-bln-openai-rival-anthropic-wsj-2023-10-27/
Google lays off 20 data scientists who worked on voice assistant team (San Francisco Chronicle): https://www.sfchronicle.com/tech/article/google-layoffs-california-companies-18465600.php
Microsoft launches 365 Copilot (CNBC): https://www.cnbc.com/2023/11/01/microsoft-365-copilot-becomes-generally-available.html
Joe Biden worried about AI after seeing "Mission: Impossible — Dead Reckoning" (Variety): https://variety.com/2023/film/news/joe-biden-worried-ai-mission-impossible-dead-reckoning-1235775115/
Suspected AI-written review content at Gannett (The Verge): https://www.theverge.com/2023/10/26/23931530/gannett-ai-product-reviews-site-reviewed-union-newsguild
Thomas Huitric (LinkedIn): https://www.linkedin.com/in/thomashuitric/
Apple reportedly "caught flat-footed" by ChatGPT, now spending $1 billion/year to catch up (Bloomberg): https://www.bloomberg.com/news/newsletters/2023-10-22/what-is-apple-doing-in-ai-revamping-siri-search-apple-music-and-other-apps-lo1ffr7p
Apple spending millions of dollars per day on AI training (The Verge): https://www.theverge.com/2023/9/6/23861763/apple-ai-language-models-ajax-gpt-training-spending
Microsoft reportedly used about 11.5 million gallons of water at Iowa data centers in July 2022 (Brian Warmoth): https://warmoth.org/2023/09/10/how-much-water-ai-is-going-to-drink/
Stanford researchers publish Foundation Model Transparency Index: https://crfm.stanford.edu/fmti/
ZenML’s vision for small, in-house AI models (TechCrunch): https://techcrunch.com/2023/10/23/zenml-empowers-companies-to-build-their-own-ai-stack-based-on-open-source-tools/