Mornings in the Lab with Keith, Jon & Friends LIVE
LIVE | When AI Sounds Certain, Why Do We Believe It?
Published · 2 hr 4 min
About this episode
You ask AI a question. A clean answer appears almost instantly. No hesitation. No awkward pause. No obvious uncertainty. It reads like the matter has been settled. So what makes you believe it? The evidence? The wording? The source? Or simply the fact that the answer sounds finished? On Mornings in the Lab, Keith, Jon and Dr. John Petrocelli dig into a growing problem in the way we use artificial intelligence: A confident-sounding AI answer can earn more trust than its evidence deserves. We get into: • Why polished AI answers can feel authoritative • What happens when AI communicates uncertainty • Why people sometimes follow incorrect AI advice • Whether confident wording actually causes overreliance • Why users can trust AI even without exaggerated certainty language • The difference between fluency and accuracy • Why checking one AI with another AI may not be an independent verification • When AI's track record makes trust reasonable • Why people can also become too distrustful after seeing an algorithm make a mistake • How to avoid swinging from blind trust to blanket rejection • When an original source matters more than another generated answer • Why the consequences of being wrong should determine how much verification you do This isn't an argument that AI is useless. AI can save time. It can improve certain kinds of work. And relying on a tool that has repeatedly performed well on a specific task can be completely reasonable. The problem starts when a well-written answer becomes a substitute for evidence. Because fluency is not proof. Confidence is not accuracy. And an answer can sound finished before your checking is finished. A useful starting point is to ask three questions: What exactly is the claim? What supports it? And what happens if it's wrong? If you're brainstorming a name, maybe the cost of a bad answer is almost nothing. If you're making a consequential decision, the standard should be much higher. Before you put your name behind the next important AI-generated claim, find the evidence you'd be comfortable putting your name behind too. 0:00:00 Countdown to LIVE 0:03:32 Cold Open: When Did Formatting Become a Qualification? 0:03:50 Good Morning, Nation: Friday in the Lab 0:04:48 Today's Question: Why Do We Believe AI When It Sounds Certain? 0:07:34 Previously On: Yesterday's CTE Conversation 0:08:31 Trending: The Fort Hood Shooter Faces a Firing Squad 0:11:18 Pick a Side: The Encyclopedia or the Search Bar? 0:13:10 What If You Had to Relive One Year? 0:15:37 Jon Joins the Show 0:17:08 Timing Big Talks With Your Wife 0:19:48 SpaceX Buys Spectrum for Starlink Mobile 0:23:33 A Live-Streamed Military Execution and the Double Standard 0:26:41 Aerosmith Announces a Comeback 0:28:18 Jon's Physique Photo: Strong in Your 50s 0:32:45 Heavy Scrolling, Anger and 3 Hours a Day on Social 0:35:30 Ambient Computing: Watching Hockey Without the Phone 0:38:35 Sleep, Water Cutoffs and Nighttime Bathroom Trips 0:42:19 Why "Intentional" Confuses People 0:44:44 Keith's Shoulder Cap Plan 0:48:02 FDA Talks on Access to Restricted Peptides 0:50:06 Epic Fail: The Slingshot Ride 0:54:43 A Camera Pill You Swallow to Read Your Gut 0:59:16 Peptides, Resting Heart Rate and Going Slow 1:03:12 Eating More Than Ever and Leaner Than Ever 1:05:23 Hour Two: Holley Muraco Postponed by a Hurricane 1:06:13 Why Does AI Sound So Official? 1:07:12 Dr. John Petrocelli Joins 1:10:29 Petro on Dr. Bennet Omalu 1:12:29 Study: AI Summaries Distort Memory of a Car Crash 1:14:46 Elizabeth Loftus: How One Word Plants a False Memory 1:19:09 Who Says What to Whom: Credibility Cues 1:22:09 The Decoy Effect: Does a Third Option Change Your Mind? 1:27:45 Petro's Day Seven on Jon's Plan 1:30:25 The Water Schedule Changes Everything 1:34:46 Energy, Boot Camp and No Crashes 1:39:16 Not Hungry, Just Not Full 1:42:39 For Every Petro, There's a Joe and a Lisa 1:47:12 The Lab as a Mentoring Program 1:50:36 Keith's Brother and the New Pavement 1:51:46 Stop Calling It a Diet 1:54:48 A Week of Guests and a Full October 1:56:12 Weekend Plans and the Undefeated Jets 1:57:52 How Lifestyle Changes Business: Three-Hour Blocks 2:02:06 The Closer 2:03:52 That's the Week ABOUT THE SHOW Mornings in the Lab with Keith & Jon is the morning show for men who refuse to hit snooze on their life. New episodes every weekday at 8 AM ET — real conversations, real stories, no corporate fluff. SUBSCRIBE for daily episodes: https://www.youtube.com/@MorningsintheLab CONNECT WITH KEITH & JON → https://www.youtube.com/@MorningsintheLab #ArtificialIntelligence #CriticalThinking #AI #ChatGPT #MorningsInTheLab
People in this conversation
Read the guest conversations
Explore the available interview transcripts on each guest’s profile.