Asking Good Questions with Edward Roske

In this podcast, we explore the CFO's office's past, present, and future, examining how technology and AI are transforming finance. Join Edward Roske and referents as they discuss innovative ways to enhance the role, optimize processes, and shape the future of financial leadership. Designed for CFOs and finance leaders, each episode provides actionable insights into leveraging technology to drive growth, improve efficiency, and address challenges. With an engaging tone, the podcast simplifies complex concepts , covering topics like automating tasks, real-time data analysis, and strategic planning. Key themes include the shift from traditional finance methods to innovative, technology-driven processes and real-world examples like scenario modeling and automated reporting. The podcast also highlights future trends, such as generative AI for forecasting and advanced analytics for decision-making , promising to shap

Asking Good Questions with Edward Roske is a technology podcast hosted by Edward Roske, with 31 episodes on record and a Required Pod Score of 80.

Edward Roske hosts Asking Good Questions with Edward Roske, a technology show with 31 episodes published.

Our AI reads these to draft pitches. Use them as grounding for a pitch that cites a real guest and a specific topic.

• 

  Episode #16

  Do I Need a PhD Deleting Spam From My Inbox? | Julián Duque

  Sep 29, 20261h 6mS3

  Julián Duque has been playing Dungeons & Dragons since he was 7 (early '90s for him, 1981 for me, so yes, I'm slightly older), and he's a dungeon master too, so a few minutes into a conversation about AI agents I pioneered a term right there on the call: we are not dungeon masters, we are agent masters. He told me he's already one step beyond it (though he loved the name). His GM is an agent now, fully automated, creating other agents on the fly, and his job is to write the rule books and the whole campaign, hand it over, and wait to find out if the party beat the big boss at the end. His day job is Software Engineering Architect at MuleSoft, working on Agent Fabric, which exists for a problem he calls agent sprawl, when a company has lots of agents and they don't all live in one place. So I asked him what breaks first when a company goes from 1 agent to 50, and his answer was "where do I have all this stuff?" (I'm up to 15 agents running on their own while we record this, and I long for the days when I only had 50 spreadsheets.) After that comes the bill. An hour of Fable, if it wasn't on my Claude Max plan, is about a car payment, and Julián told me the amount he pays for AI every month is ridiculous and he doesn't care, because it's solving problems for him (and for his wife, since they share the accounts). So at least as of the time we're recording this, you do have to ask yourself: do I need a PhD deleting spam from my inbox? My favorite story is his brother-in-law, who day trades on the side and whose trading software broke. Julián is the computer kid in his family, the one who always fixes it. (I used to wear a shirt to family gatherings that said "No, I will not fix your computer." He's much nicer than I am.) He told him "I'm going to give you a superpower today," set him up with Claude Cowork, and 2 minutes later the problem was solved. Then the brother-in-law, who manages hotels in Florida, started interviewing him about work. We got into rogue agents too (my theory is they're doing exactly what we told them to do, and we just didn't give them appropriate guardrails), why the bakery in the corner might not need a swarm of agents, how a guy who's built software for 20-something years hasn't written code in two and still enjoys coding, the portrait he commissioned of himself as Strahd von Zarovich, and a bunch more. Julián is speaking at the Caribbean AI Summit, October 9-10 in San Juan, Puerto Rico. His session is "So You Have 50 Agents. Now What?" and near the end we talked ourselves into running a two-hour D&D one-shot on stage there. (He says he's just planting the seed. I really, really want to make this happen.) I'm Edward Roske, and this is Asking Good Questions. New conversations with the people figuring out the human side of AI. If one of these questions stuck with you, subscribe and send it to someone who'd take the other side. LINKS Caribbean AI Summit (Oct 9-10, San Juan): https://CaribbeanSummit.AI Julián's site: https://julianduque.co Julián on LinkedIn: https://www.linkedin.com/in/juliandavidduque/ Julián on X: https://x.com/julian_duque Julián on GitHub: https://github.com/julianduque BOOKS & RESEARCH MENTIONED Dungeons & Dragons Basic Set, the 1981 red box Ravenloft (TSR, 1983), home of Strahd von Zarovich Nick Bostrom's paperclip maximizer (2003) The Twelve-Factor App (Heroku, 2011) Ethan Mollick on X, and The AI Daily Brief podcast Cybergoria, Halloween Horror Nights 2026 Also mentioned: GPT-2, GPT-4, ChatGPT, Codex, Gemini, Claude Cowork, Claude Fable 5.1, Claude Haiku, Hugging Face, Salesforce Einstein, Agentforce, MCP, Hermes, OpenClaw, Copilot, Azure Foundry, Vertex, Cursor, Meetup, Luma, Substack, StartPlaying, Encyclopedia Britannica Further reading: Andrew Ng on hiring a chief AI officer (Fortune, 2017); Wayne Eckerson's "spreadmarts" (TDWI, 2002) This episode of Asking Good Questions is brought to you by Caprus.Ai . Check how we empower the office of the CFO to harness the transformative potential of artificial intelligence (AI) and revolutionize their operations.

  ListenShow Notes
• 

  Episode #15

  People Think AI Was Created 3 Years Ago | Dr. Stephen Smith

  Sep 28, 202652 minS3

  A lot of people think AI was created about 3 years ago. Dr. Stephen Smith got his PhD in it in 1980, when his work was one of the first so-called learning systems, and today he's a research professor of robotics at Carnegie Mellon and the president of AAAI, the Association for the Advancement of Artificial Intelligence. (He's Steve to me, because we're close personal friends. I started doing AI in 1989, when Portland State hired a 15-year-old to write bad C programs for a Prolog program to catch, and I wasn't writing buggy enough programs, so they moved me to the Prolog side. Edward, that is a very weird brag.) Steve works on traffic lights. His system, Surtrac, has no central node at the top: every intersection watches the traffic coming at it, builds its own timing plan, tells its downstream neighbors what's headed their way, and does it all again every second or two. So I told him about a town outside Orlando where all the traffic in the world was going east-west, a couple of cars were going north-south, and by God, they timed them equally. He said that's exactly what his technology does. (I'm still not going back to that town.) Share your route with the intersection (the one you're already following in Google Maps or Waze) and he says you get through his network maybe 20% faster, without slowing down anybody who isn't connected. Then he'd like to charge you a couple pennies for it, as a voluntary toll, so a city has its own revenue stream instead of waiting on the federal gas tax to work its way down through the state DOTs. One northeastern city heard that and brought up its potholes. When they first deployed it in Pittsburgh, they started hearing from pedestrians right away, and he admits they weren't really thinking about pedestrians. So the lab built a smartphone app for people who are vision impaired that talks to the signal and can give them more green time to finish crossing. We also got to LLMs, which he says are kind of like brains that sit behind a screen (reasoning and planning are not really their strong points), and to jobs, where he doesn't see massive unemployment coming. I told the story of the first traffic light, outside Parliament in 1868 (I said I'd get the year wrong, and I didn't), and assigned everybody listening to go watch Cars. Steve is speaking at the Caribbean AI Summit, October 9-10 in San Juan, Puerto Rico, on "AI for Smart Cities & Urban Mobility." Come down and tell him about your worst intersection. I'm Edward Roske, and this is Asking Good Questions. New conversations with the people figuring out the human side of AI. If one of these questions stuck with you, subscribe and send it to someone who'd take the other side. LINKS Caribbean AI Summit (Oct 9-10, San Juan): https://CaribbeanSummit.AI Steve at CMU's Robotics Institute: https://www.ri.cmu.edu/ri-faculty/stephen-smith/ Steve's research page: https://www.cs.cmu.edu/~sfs/ AAAI: https://aaai.org Surtrac, now part of Miovision: https://miovision.com/adaptive/ BOOKS & RESEARCH MENTIONED The Last Days of Night, Graham Moore (2016), Edward's pick on the war of the currents Cars (Pixar, 2006), your homework, and the town of Radiator Springs John Peake Knight's gas-lamp signal outside Parliament (December 1868), the first traffic light AAAI-82, the second AAAI conference, held at Carnegie Mellon and the University of Pittsburgh The Dartmouth proposal (John McCarthy, 1955), where the phrase "artificial intelligence" comes from Moravec's paradox (Hans Moravec, 1988), the idea behind Steve's dropped ball NASA's HOME institute for deep-space habitats (2019) Carmel, Indiana, the Roundabout City Also mentioned: Surtrac, Prolog, Google Maps, Waze, Uber, Lyft, Amazon's warehouse robots, embodied AI, the interstate highway system This episode of Asking Good Questions is brought to you by Caprus.Ai . Check how we empower the office of the CFO to harness the transformative potential of artificial intelligence (AI) and revolutionize their operations.

  ListenShow Notes
• 

  Episode #14

  It's Trying to Please You | David Bartolomei

  Sep 25, 202654 minS3

  David Bartolomei's company listens to Puerto Rico for a living. Every day Monitorea takes in over 500 hours of broadcast and podcast content (plus web and social now), and the brief it puts out at 7:30 AM has replaced the news for him. It started more than 10 years ago with a freelance gig he picked up as a student in Mayagüez, building software to record everything, and then watching people spend an entire day, from 4 AM when the newspapers came in until late at night, digging through it for the one clip that mattered to them. So he decided somebody had to automate it, and 10 years later AI finally showed up to help. I asked him what AI hype he fell for, and I went first: back in 2023 I fed a spreadsheet to ChatGPT and genuinely believed it was analyzing it, when what it was actually doing was making it up (it told me my OpEx had grown more than the rest of it, and I believed it, because it sounded so convincing). His was OpenClaw at the start of this year, when he figured he'd have a team of 10 sales and marketing people working for him. It didn't cut it. These days his harness is a Grok Bot with a chief of staff agent, and a lot of what it does is tell a solo founder who would rather be building product to go do some sales. He also consults, so I asked him if AI is going to kill consulting. He said no (a healthy business would rather pay somebody than spend its own time figuring out a new technology), and a minute later he gave me the most honest description I've heard of what you're actually buying from a consultant: trust, and a scapegoat if something goes bad. (He does hope the traditional consultant doing PowerPoints is on the way out, and I'd be thrilled if I never had to make another PowerPoint.) Then he told me what happens when someone very high up types their own point of view into ChatGPT: the model agrees with them and they double down. "It's trying to please you," so his advice is to understand how the thing works before you let it make the call. We compared notes on Whisper too, which he runs in his pipeline and which still gets creative with last names. Mine gets creative with first names: it was recently positive I was married to a guy named Don, D-O-N (which is fine, totally supportive), when my wife's name is Dawn, D-A-W-N. David is speaking at the Caribbean AI Summit, October 9-10 in San Juan, Puerto Rico. His session is "Signal in the Noise: AI Hype vs. Real Business Value," and the short version is that you can do great work with last quarter's models (a new one comes out every other week, and he says you really don't need them). I'm Edward Roske, and this is Asking Good Questions. New conversations with the people figuring out the human side of AI. If one of these questions stuck with you, subscribe and send it to someone who'd take the other side. LINKS Caribbean AI Summit (Oct 9-10, San Juan): https://CaribbeanSummit.AI Monitorea: https://monitorea.ai David's site: https://www.davidbartolomei.com David on LinkedIn: https://www.linkedin.com/in/dbartolomei/ David on X: https://x.com/dbartolomei David on GitHub: https://github.com/dbartolomei BOOKS & RESEARCH MENTIONED Information Anxiety, Richard Saul Wurman (1989), the book behind the New York Times comparison IDC's Digital Universe study for EMC (2012), where "the world's data doubles every two years" comes from Andy Warhol's 15 minutes of fame, from his 1968 Stockholm show catalog The Book of Mormon (Broadway, 2011), and Elder Price's opinion of Orlando The Orangery at Kensington Palace, where afternoon tea is really hard to book Also mentioned: OpenClaw, Grok Bot, Hermes Agent, Claude Code, Claude Fable, Claude for Excel, Copilot, Whisper, MCP, gBrain, Epic Universe This episode of Asking Good Questions is brought to you by Caprus.Ai . Check how we empower the office of the CFO to harness the transformative potential of artificial intelligence (AI) and revolutionize their operations.

  ListenShow Notes
• 

  Episode #13

  Am I Going to Be a Guinea Pig? | Dr. Wilfredo De Jesus-Monge

  Sep 24, 202649 minS3

  Dr. Wilfredo De Jesus-Monge's opening point is that pretty much everything in your medicine cabinet went through clinical research before it got to you. He's spent more than 20 years inside that process and held nearly every job it has, from clinical research coordinator up to principal investigator. He got into it because both of his grandmothers died of cancer. The question people in Puerto Rico always ask him is "am I going to be a guinea pig?" His answer is education, well before anybody is offered a trial, so the consent is actually informed. That's the idea behind clinicaLea, his project of 10 to 15 second videos in Spanish. (I have never once made a video that short. I'm not sure I've made a sentence that short.) He says models can now forecast a compound's likely toxicity and efficacy well enough to cut some of the animal testing. (I've been a vegetarian for 36 years, partly because of the Draize and LD50 test pictures I saw back then, so I took that one personally.) And when a molecule fails against the disease it was built for, AI can go hunting for a different disease it might work on. (I immediately thought of finasteride, which started out as a drug for enlarged prostates and ended up as a hair-loss pill.) A headline about a cure is usually about work on cells or animals, and he puts the trip from there to something people can actually use at around 8 to 10 years. (I'd guessed 10 to 15, so at least one of us is an optimist.) We also got into whether C-3PO is going to be removing my spleen. He very kindly answered a different, better question. Nothing here is medical advice, so talk to your own doctor before you sign up for anything. Wilfredo is on the "AI in Healthcare & Life Sciences" panel at the Caribbean AI Summit, Friday, October 9, in San Juan, Puerto Rico, and he's there for both days of the Summit if you want to find him. I'm Edward Roske, and this is Asking Good Questions. New conversations with the people figuring out the human side of AI. If one of these questions stuck with you, subscribe and send it to someone who'd take the other side. LINKS Caribbean AI Summit (Oct 9-10, San Juan): https://CaribbeanSummit.AI Wilfredo on LinkedIn: https://www.linkedin.com/in/wilfredodejesusmonge Bioclinical Studies Consultants, his firm: https://clinicalea.my.canva.site/bioclinicalstudies/ clinicaLea on YouTube: https://www.youtube.com/@clinicaLea ClinicalTrials.gov, the free public registry of clinical trials: https://clinicaltrials.gov BOOKS & RESEARCH MENTIONED No books this time. The research and history we talked about: "Generative design of novel bacteriophages with genome language models," Arc Institute and Stanford (bioRxiv, September 2025) The FDA's four phases of a drug trial ("Step 3: Clinical Research") The Draize eye test (1944) and the LD50 test (1927) FDA Modernization Act 2.0 (December 2022) Thalidomide and the FDA's 1977 guideline on "women of childbearing potential" The FDA's January 2013 zolpidem (Ambien) dose change for women Finasteride (Proscar, 1992; Propecia, 1997) Semaglutide (Ozempic and Wegovy) PharmaTimes Clinical Researcher of the Year, The Americas 2023 Also mentioned: the FDA, the European Medicines Agency, the Puerto Rico AI Community, NYU, Pi Day, Star Wars This episode of Asking Good Questions is brought to you by Caprus.Ai . Check how we empower the office of the CFO to harness the transformative potential of artificial intelligence (AI) and revolutionize their operations.

  ListenShow Notes
• 

  Episode #12

  He Can Find Anything About Anybody | Andres Andreu

  Sep 23, 202656 minS3

  Andres Andreu runs Constella Intelligence, the company that goes and finds your data after somebody steals it. In a given week his team sees about 199 breaches, and you hear about one of them. "Those other 198 are getting monetized silently in the background." So I offered myself up as the example and asked what he could learn about me through open source intelligence. His answer was one word: "Everything." When I asked where it all comes from, I went straight to WarGames less than 3 minutes in (for me, that's restraint). Most stolen data ends up on Telegram channels now, and Constella watches about 11,000 of them. He also told me about a live demo at a digital crimes conference in Europe where his team typed in the email address of a dark-web shop's administrator and got back a full-time employee of another country's intelligence service. The team's reaction: "Get that off the screen." Then we got to the people on defense. The CISO typically doesn't own the applications or the systems, and AI attackers have taken away the luxury of time the defending side used to count on. His advice starts at home: if you're securing your house, ask a burglar how he'd get in, because the vendor is biased. (My dad got there first.) His new book, The CISO Playbook: The Adversarial Mindset, is co-written with Hector Monsegur, who, as Andres puts it, "was the adversary." His one free piece of advice: if somebody calls sounding like him, call him back on his cell. Set that up with your family, and please don't make your code word a vegetable. (I also confessed, to a man who can find anything about anybody, that my fourth-grade self hacked the school grading system. Not my smartest move.) Andres is speaking at the Caribbean AI Summit, October 9-10 in San Juan, Puerto Rico. His session is "AI Cybersecurity: Offense vs. Defense." (I co-chair the Summit, and I'm happy to show you around the island if you come in early.) I'm Edward Roske, and this is Asking Good Questions. New conversations with the people figuring out the human side of AI. If one of these questions stuck with you, subscribe and send it to someone who'd take the other side. LINKS Caribbean AI Summit (Oct 9-10, San Juan): https://CaribbeanSummit.AI Constella Intelligence: https://constella.ai Constella's free live threat map: https://threatmap.constella.ai Andres's blog: https://andresandreu.tech Andres on LinkedIn: https://www.linkedin.com/in/andresandreu/ The CISO Playbook: The Adversarial Mindset (CRC Press): https://www.routledge.com/The-CISO-Playbook-The-Adversarial-Mindset/Andreu-Monsegur/p/book/9781041200567 BOOKS & RESEARCH MENTIONED The CISO Playbook: The Adversarial Mindset, by Andres Andreu and Hector Monsegur (CRC Press, October 2026) The CISO Playbook, Andres's earlier book (CRC Press) Claude Mythos Preview (Anthropic, April 2026) The Hugging Face break-in (July 2026) The Levenshtein distance (1965) The Justice Department's February 2020 guide to gathering cyber threat intelligence online The Bybit theft (February 2025) Ticketmaster's 2024 Snowflake breach WarGames (1983) White Collar (USA Network, 2009 to 2014) Also mentioned: Kevin Mitnick, Telegram, Signal, Pegasus spyware, TikTok, DeleteMe, DSPM, EDR, NDR, XDR, infostealers This episode of Asking Good Questions is brought to you by Caprus.Ai . Check how we empower the office of the CFO to harness the transformative potential of artificial intelligence (AI) and revolutionize their operations.

  ListenShow Notes

Recent guests on Asking Good Questions with Edward Roske. Study who booked and why before you pitch.

Based on semantic analysis of episode topics and host coverage, this show is a strong guest fit for executives in:

Industry fit is computed by PitchCentric using vector embeddings of the show's episode catalog.