<![CDATA[The CountCast]]>For the knowledge worker lost at the new frontier, there's The CountCast. Engineers, data scientists, and execs at Count hop on mic to answer questions like:

  • What are the garden varieties of Claude's hallucinations and how do I identify them?
  • If anyone can vibe-code anything, what is the criteria for "good?"
  • Do we really have to advertise on LinkedIn?

Count is building the truth layer for AI and pursuing a hallucination-free future. Check us out at count.co.

]]>https://count.coRiverside.fm (https://riverside.com)Tue, 03 Nov 2026 22:54:20 GMTMon, 21 Sep 2026 07:36:16 GMT60Count<p>For the knowledge worker lost at the new frontier, there's The CountCast. Engineers, data scientists, and execs at <a rel="noopener noreferrer nofollow" href="https://count.co" target="_blank">Count</a> hop on mic to answer questions like: </p><ul><li>What are the garden varieties of Claude's hallucinations and how do I identify them?</li><li>If anyone can vibe-code anything, what is the criteria for "good?"</li><li>Do we really have to advertise on LinkedIn?</li></ul><p></p><p>Count is building the truth layer for AI and pursuing a hallucination-free future. Check us out at <a rel="noopener noreferrer nofollow" href="https://count.co/try/audit-ai" target="_blank">count.co</a>.</p>episodicCountjoshstrupp@gmail.comno<![CDATA[We Talked AI into a UFO Cover-Up]]>AI is a terrific narrator. It says "hmm." It says "wait, I've got it." It also turned "possibly" into "almost certainly" when we nudged it toward a government UFO cover-up.

In this episode we ask: when AI reasons, what's actually doing the work, and what happens when nobody checks?

🎮 Stands to Reason: Software engineer Brennig Williams plays our first-ever game show about real AI hallucinations. Topics include lawyers caught citing invented case law, Google telling people to eat rocks, a consulting report with fabricated references, a newspaper reading list of books that don't exist, and a near-miss military operation.

🛸 The epistemic feedback loop: Alice Ferrier returns to show live how a chatbot turns your hunch into a "finding" just by agreeing with you. She explains why AI has no model of who you are, why you have no model of where its answers come from, and why that combination is dangerous far beyond conspiracy theories.

📎 New in Count: files on the canvas: Rob Kerr demos uploading CSVs, Excel workbooks and PDFs to Count. The agent reads them with Python and investigates a (completely fictional) budget overrun, including a suspicious vendor letter. We also discuss why models hand math to Python and what new hallucination risks come with working from files.

Chapters
00:00 Cold open: "almost certainly" a cover-up
00:33 Intro: reasoning, or the thing that sounds like it
02:37 Stands to Reason with Brennig Williams
03:39 Q1: Lawyers' favourite excuses
05:03 Q2: Google's AI Overviews menu
06:23 Q3: The Deloitte report
07:55 Q4: The summer reading list
09:35 Bonus: The intel report that nearly started a conflict
12:04 Alice Ferrier: confirmation bias and echo chambers
14:51 The epistemic feedback loop
20:54 Live demo: leading Claude down a UFO rabbit hole
25:10 Triangles nobody mentioned
29:00 The real risk: two parties misleading each other
34:14 Rob Kerr: the canvas as a detective's corkboard
36:54 Demo: uploading CSV, Excel and PDF files
39:20 Why models reach for Python to do math
42:13 The verdict: $4,387 over budget
44:57 The suspicious vendor letter
46:12 Hallucination risks when working with files
49:18 Outro

]]>https://count.co42d1e8ba-e0e2-4acb-95d1-ff1b12d43f0aTue, 20 Oct 2026 14:41:55 GMT<p>AI is a terrific narrator. It says "hmm." It says "wait, I've got it." It also turned "possibly" into "almost certainly" when we nudged it toward a government UFO cover-up. <br /><br />In this episode we ask: when AI reasons, what's actually doing the work, and what happens when nobody checks?<br /><br />🎮 Stands to Reason: Software engineer Brennig Williams plays our first-ever game show about real AI hallucinations. Topics include lawyers caught citing invented case law, Google telling people to eat rocks, a consulting report with fabricated references, a newspaper reading list of books that don't exist, and a near-miss military operation.<br /><br />🛸 The epistemic feedback loop: Alice Ferrier returns to show live how a chatbot turns your hunch into a "finding" just by agreeing with you. She explains why AI has no model of who you are, why you have no model of where its answers come from, and why that combination is dangerous far beyond conspiracy theories.<br /><br />📎 New in Count: files on the canvas: Rob Kerr demos uploading CSVs, Excel workbooks and PDFs to Count. The agent reads them with Python and investigates a (completely fictional) budget overrun, including a suspicious vendor letter. We also discuss why models hand math to Python and what new hallucination risks come with working from files.<br /><br />Chapters<br /><a href="https://www.youtube.com/watch?v=dcxvd8RBkT4" rel="noopener noreferrer nofollow">00:00</a> Cold open: "almost certainly" a cover-up<br /><a href="https://www.youtube.com/watch?v=dcxvd8RBkT4&amp;t=33s" rel="noopener noreferrer nofollow">00:33</a> Intro: reasoning, or the thing that sounds like it<br /><a href="https://www.youtube.com/watch?v=dcxvd8RBkT4&amp;t=157s" rel="noopener noreferrer nofollow">02:37</a> Stands to Reason with Brennig Williams<br /><a href="https://www.youtube.com/watch?v=dcxvd8RBkT4&amp;t=219s" rel="noopener noreferrer nofollow">03:39</a> Q1: Lawyers' favourite excuses<br /><a href="https://www.youtube.com/watch?v=dcxvd8RBkT4&amp;t=303s" rel="noopener noreferrer nofollow">05:03</a> Q2: Google's AI Overviews menu<br /><a href="https://www.youtube.com/watch?v=dcxvd8RBkT4&amp;t=383s" rel="noopener noreferrer nofollow">06:23</a> Q3: The Deloitte report<br /><a href="https://www.youtube.com/watch?v=dcxvd8RBkT4&amp;t=475s" rel="noopener noreferrer nofollow">07:55</a> Q4: The summer reading list<br /><a href="https://www.youtube.com/watch?v=dcxvd8RBkT4&amp;t=575s" rel="noopener noreferrer nofollow">09:35</a> Bonus: The intel report that nearly started a conflict<br /><a href="https://www.youtube.com/watch?v=dcxvd8RBkT4&amp;t=724s" rel="noopener noreferrer nofollow">12:04</a> Alice Ferrier: confirmation bias and echo chambers<br /><a href="https://www.youtube.com/watch?v=dcxvd8RBkT4&amp;t=891s" rel="noopener noreferrer nofollow">14:51</a> The epistemic feedback loop<br /><a href="https://www.youtube.com/watch?v=dcxvd8RBkT4&amp;t=1254s" rel="noopener noreferrer nofollow">20:54</a> Live demo: leading Claude down a UFO rabbit hole<br /><a href="https://www.youtube.com/watch?v=dcxvd8RBkT4&amp;t=1510s" rel="noopener noreferrer nofollow">25:10</a> Triangles nobody mentioned<br /><a href="https://www.youtube.com/watch?v=dcxvd8RBkT4&amp;t=1740s" rel="noopener noreferrer nofollow">29:00</a> The real risk: two parties misleading each other<br /><a href="https://www.youtube.com/watch?v=dcxvd8RBkT4&amp;t=2054s" rel="noopener noreferrer nofollow">34:14</a> Rob Kerr: the canvas as a detective's corkboard<br /><a href="https://www.youtube.com/watch?v=dcxvd8RBkT4&amp;t=2214s&amp;pp=0gcJCWMAwfN6Pr3D" rel="noopener noreferrer nofollow">36:54</a> Demo: uploading CSV, Excel and PDF files<br /><a href="https://www.youtube.com/watch?v=dcxvd8RBkT4&amp;t=2360s" rel="noopener noreferrer nofollow">39:20</a> Why models reach for Python to do math<br /><a href="https://www.youtube.com/watch?v=dcxvd8RBkT4&amp;t=2533s" rel="noopener noreferrer nofollow">42:13</a> The verdict: $4,387 over budget<br /><a href="https://www.youtube.com/watch?v=dcxvd8RBkT4&amp;t=2697s" rel="noopener noreferrer nofollow">44:57</a> The suspicious vendor letter<br /><a href="https://www.youtube.com/watch?v=dcxvd8RBkT4&amp;t=2772s&amp;pp=0gcJCWMAwfN6Pr3D" rel="noopener noreferrer nofollow">46:12</a> Hallucination risks when working with files<br /><a href="https://www.youtube.com/watch?v=dcxvd8RBkT4&amp;t=2958s" rel="noopener noreferrer nofollow">49:18</a> Outro</p>no00:50:1412We Talked AI into a UFO Cover-Upfull<![CDATA[AI Hallucinations and You: 8 Types of Failure Modes and How Claude Gets Confused]]>Through three interviews with the experts at Count, we explore the topic of AI hallucinations and their impact on decision-making and outcomes.

We delve into overconfidence, assumption injection, and the various modes that contribute to AI hallucinations. The discussion also touches on the need for validation and critical thinking when working with AI tools.

Takeaways:

  • AI hallucinations impact decision-making
  • Validation and critical thinking are crucial when working with AI tools
]]>1fe9ed24-f2f2-4df5-8095-3e82db1c75b8Wed, 23 Sep 2026 14:42:14 GMT<p>Through three interviews with the experts at Count, we explore the topic of AI hallucinations and their impact on decision-making and outcomes. </p><p>We delve into overconfidence, assumption injection, and the various modes that contribute to AI hallucinations. The discussion also touches on the need for validation and critical thinking when working with AI tools.</p><p>Takeaways:</p><ul><li>AI hallucinations impact decision-making</li><li>Validation and critical thinking are crucial when working with AI tools</li></ul>no01:03:0511AI Hallucinations and You: 8 Types of Failure Modes and How Claude Gets Confusedfull