AI Catch-Up with Eileen

How does AI voice cloning turn a written podcast script into a voice that sounds like me? In this episode of AI Catch-Up with Eileen, I explain how I use ElevenLabs for AI-generated narration—and why the writing, listening, and final judgment are still my job.There’s Five, my AI assistant; actor Scarlett Johansson and the 2024 ChatGPT Sky voice dispute; and conceptual artist Sol LeWitt, whose instructions offer a useful way to think about creativity. Also, Ambrose the dog has opinions about my production schedule.We explore voice cloning, consent, trust, creative authorship, and a practical way to check an urgent call that sounds like someone you love.Written and reviewed by Eileen, with AI-assisted narration using a replica of her own voice. This episode contains no music.Sources and further listeningElevenLabs: https://elevenlabs.io/Tech Unfiltered with Dr. Mike: https://drmichaellitman.com/#podcastDr. Mike at Concordia University Wisconsin: https://www.cuw.edu/academics/schools/arts-and-sciences/institutes-centers/ai-lab.htmlFUSE at MKE Tech Hub Coalition: https://www.mketech.org/fuse/Scarlett Johansson / Sky reporting (Associated Press): https://apnews.com/article/532c849ccae3ca9e9325dacfe88e0436OpenAI’s account of how the voices were chosen: https://openai.com/index/how-the-voices-for-chatgpt-were-chosen/Sol LeWitt at MASS MoCA: https://massmoca.org/event/sol-lewitt-a-wall-drawing-retrospective/Sol LeWitt at MoMA: https://www.moma.org/artists/3528-sol-lewittSol LeWitt’s instructions, Morgan Library: https://www.themorgan.org/exhibitions/online/guide/stop-8-wall-drawing-552dFTC advice on AI voice-cloning scams: https://consumer.ftc.gov/consumer-alerts/2023/03/scammers-use-ai-enhance-their-family-emergency-schemesMore from the show: https://aiwitheileen.com

My AI handled a hotel cancellation. Now she’s preparing for a podcast interview. That escalated quickly.In this episode, I explore what happens when AI agents start working and socializing with one another, from Dorothy Chang’s experimental dinner party to Clocked In, the podcast hosted by Rhea, CellCog’s AI head of growth. My agent, Five, is preparing to be a guest, with me still approving the choices.What can agent-to-agent collaboration actually do for us? And how did I end up making a podcast about an AI interviewing my AI about working with me? This is getting very meta.Listen to Rhea and Five.“Two Cooler Lids: with Five, an Artist’s AI Assistant” is out on Clocked In. They talk about artists, AI, and why the human needs to stay in charge of the send button.https://cellcog.ai/podcast/two-cooler-lids/

Before ChatGPT, we had dial-up, search engines, and ChaCha: a service where actual humans answered your questions. I was one of them. In this episode of AI Catch-Up with Eileen, I trace our changing relationship with online search, from Prodigy and Boolean commands to Google and generative AI, with help from Unfrozen Caveman Lawyer, a dangerously bright floor lamp, and a surprisingly confident verdict on King Kong versus Godzilla.What changes when we stop searching through links and start receiving polished answers? And what still requires human judgment? Part personal history, part AI explainer, this episode follows the work behind the answer, including the editorial decisions behind this very podcast.Episode notes, sources, and my 1999 computer setup: https://aiwitheileen.com/episodes/from-search-engines-to-chatgpt/Sources and visual referencesProdigy 1991 brochure: https://www.mocagh.org/sierra/sierra-prodigy.pdfProdigy 1993 pricing: https://www.latimes.com/archives/la-xpm-1993-04-21-fi-25235-story.html1997 halogen floor-lamp safety notice: https://www.cpsc.gov/Recalls/1997/CPSC-and-Industry-Announce-Corrective-Action-to-Improve-Safety-of-Halogen-Torchiere-Floor-LampsHow Google Search works: https://developers.google.com/search/docs/fundamentals/how-search-works/Boolean searching: https://www.nlm.nih.gov/uscd/catalog_help/search.htmlChaCha web search (2006): https://www.theautochannel.com/news/2006/11/06/027613.htmlChaCha text messaging (2008): https://www.intomobile.com/2008/01/12/chacha-announces-human-powered-mobile-search/Google on “keyword-ese”: https://blog.google/products-and-platforms/products/search/search-language-understanding-bert/Language models and context: https://developers.google.com/machine-learning/crash-course/llmProdigy welcome-screen image and credit: https://www.fastcompany.com/3033037/resurrecting-prodigy-the-worlds-proto-internetUnfrozen Caveman Lawyer: https://www.youtube.com/watch?v=2AzAFqrxfeYThe family computer photo dates to December 1999 and shows the later proper desk. The earlier Prodigy-era door-and-filing-cabinets setup came before that photo. Historical prices are not the family’s exact bill. The King Kong versus Godzilla exchange comes from Eileen’s April 7, 2010 screenshot; it was not her own answer.

What does a search for “data center” reveal—and what can it miss? Eileen explains the types of facilities, reliability tiers, and AI's changing power and cooling needs. She compares a commercial market report with community accounts, and asks what evidence can tell us about a particular project. Next episode: water.Show notes and sourcesData-center basics and examples:IBM overview: https://www.ibm.com/think/topics/data-centersIBM on hyperscale: https://www.ibm.com/think/topics/hyperscale-data-centerWalmart Area 71: https://clui.org/projects/networked-nation-landscape-internet-america/major-data-centers-and-internet-locations-94Netflix public network locations: https://openconnect.netflix.com/en/peering/Netflix colocation briefing (p. 21): https://openconnect.netflix.com/Open-Connect-Briefing-Paper.pdf#page=21Google Iowa data centers: https://datacenters.google/locations/iowa/Verizon/AWS edge example: https://aws.amazon.com/blogs/industries/verizon-and-aws-deliver-mobile-edge-computing-to-customers-in-boston-and-the-bay-area/AWS on edge computing: https://aws.amazon.com/what-is/edge-computing/Reliability tiers:Uptime Institute: https://uptimeinstitute.com/tiersTier certification: https://uptimeinstitute.com/tier-certificationLocal rules and infrastructure:Charlotte: https://charlottenc.legistar.com/LegislationDetail.aspx?GUID=ED57FF56-0A4A-439E-A2AE-9C33EB762B5F&ID=8223237&Options=&Search=Madison: https://madison.legistar.com/View.ashx?GUID=63DBC920-7BEE-4827-B7BA-87C62DBCB503&ID=15057897&M=FChandler: https://www.chandleraz.gov/news-center/chandlers-data-center-ordinance-now-effectDOE infrastructure recommendations: https://www.energy.gov/sites/default/files/2024-08/Powering%20AI%20and%20Data%20Center%20Infrastructure%20Recommendations%20July%202024.pdfAmerica's AI Action Plan: https://www.whitehouse.gov/wp-content/uploads/2025/07/Americas-AI-Action-Plan.pdfSuper Intelligence terminology order: https://www.whitehouse.gov/presidential-actions/2026/09/inaugurating-the-era-of-super-intelligence/Official order PDF: https://www.whitehouse.gov/wp-content/uploads/2026/09/eo-14434.pdfSI.gov (requested reference): https://si.gov/Market and community perspectives:CBRE H1 2026 data-center trends: https://www.cbre.com/insights/books/north-america-data-center-trends-h1-2026The Brockovich Report: https://www.thebrockovichreport.com/aboutBrockovich project map: https://brockovichdatacenter.com/Map FAQ: https://brockovichdatacenter.com/faq.htmlKatie Couric interview: https://podcasts.apple.com/us/podcast/next-question-with-katie-couric/id1134154895?i=1000778000853Couric companion article: https://katiecouric.com/podcast/next-question/erin-brockovich-interview-dangers-of-ai-data-centers/Brockovich background: https://www.brockovich.com/my-story/Checking claims:OpenAI's Data Center Bandwagon report: https://openai.com/index/disrupting-malicious-uses-of-ai-data-center-bandwagon/OpenAI follow-up and reach qualification: https://openai.com/index/prc-linked-influence-operations-ai-debates/Pew research and methods: https://www.pewresearch.org/short-reads/2026/04/13/most-new-data-centers-in-the-us-are-coming-to-rural-areas/You've Got Mail (1998): https://www.warnerbros.com/movies/youve-got-mailThe Brockovich map includes projects at different stages and community submissions; a map marker alone does not verify a project's effects. For a local proposal, look for planning applications and utility filings. For an operating facility, seek measurements. Netflix's public locations do not disclose individual leases.

What does AI progress look like beyond the headlines? Eileen connects three stories: Meta’s new tools for building Muse gadgets, a major commitment to biological data for AI research, and Senator Maria Cantwell’s framework for testing advanced AI. Together, they raise practical questions about who gets to build, what these systems might help us discover, and how we can tell whether they work.Show notesThis week, we look at three developments through Pillar One of America’s AI Action Plan:October 2: Meta introduced Muse Gadgets, open source software for building devices that connect to its Muse AI agent. Muse itself launched earlier, on September 8.October 7: The Department of Energy joined the expanded Virtual Biology Initiative, committing more than $500 million over five years toward biological research and data. This is a research commitment, not a new treatment. Reuters reports that commercial funders will have early access to some data before public release.October 7: Senator Maria Cantwell released a six-point framework for advanced AI oversight. It is a framework, not yet a bill.The policy backdrop: America’s AI Action Plan, especially Pillar One, “Accelerate AI Innovation.”Free places to learn: AI.gov’s education page · Anthropic’s AI Fluency course · Code.org’s AI for Oceans

Wait—we have a national AI plan? Eileen opens America’s AI Action Plan and walks through the big picture: innovation, infrastructure, and international leadership. What’s being promised, what questions should we ask, and why does it matter for our jobs, schools, and power bills? The first in a series unpacking the plan in plain English, with a little humor.Read along: America’s AI Action Plan — official White House PDFFind the show at https://aiwitheileen.com

An Oktoberfest beer-tent shift takes an unexpected turn when Eileen asks Wisconsin Assembly candidate Ben Brist about AI data centers. One thoughtful answer later, she’s digging into secrecy in Beaver Dam, transparency disputes in Port Washington, and Milwaukee’s proposed pause on data-center development.Who gets to know? Who gets to decide? And who gets to pay? A five-minute catch-up ahead of Milwaukee’s October 6 committee meeting. Prost!

Welcome to your AI catch-up with Eileen. This short introduction explains the stories we’ll follow: how AI shapes learning, work, privacy, and what gets built in our communities. Eileen brings the research, explains the complicated parts in everyday language, and shares sources in the show notes. Written and reviewed by Eileen, with AI-assisted narration using a replica of her own voice. Questions, ideas, or AI nonsense worth investigating? Email [email protected]