Talking to AI

In this episode of Talking to AI, Paul and his friend Mark attempt a Dungeons & Dragons session led by ChatGPT as their AI Dungeon Master. Building upon a detailed and ambitious prompt, they strive to see if AI can live up to the challenge of running an engaging and rule-abiding tabletop adventure. Despite hopes for thrilling encounters after their underwhelming first attempt, the session falls short: the AI's story remains passive, adversaries are non-confrontational, and the mechanics often ignore or misinterpret essential D&D rules. Paul and Mark experience firsthand the AI's struggles with dice rolls, stat calculations, and narrative proactivity.

Throughout the journey, Paul critiques the AI's failure to set up characters properly, missed steps in gear assignment, and lack of initiative from non-player characters. Even when provided with ample rules and resources—including a book of starting scenarios and the full D&D ruleset—the AI heavily relies on player direction and fails to build dramatic tension or memorable story moments, often repeating bland phrases and showing limited strategic depth.

After reflecting on the session, Paul and Mark discuss the inherent limitations of current generative AI tools as D&D Dungeon Masters, exploring whether simpler structures or additional context might improve results in subsequent experiments. Their experience exposes both the comedic mishaps and persistent shortcomings of AI-driven roleplay, making for an entertaining but cautionary tale about the current state of AI in gaming.

🎙️ Hosted by Paul at Talking to AI — where real people, real problems, and real conversations meet artificial intelligence.

In this episode of Talking to AI, we delve into logic, the fourth foundational branch in the Philosophy 101 series after epistemology, metaphysics, and ethics. Host Paul and the AI co-host discuss why logic is often considered the quiet backbone of philosophical reasoning, highlighting its fundamental place beneath even metaphysics and ethics. The conversation navigates classical logic—with its binary truths and syllogisms—moves through propositional and predicate logic, and explores how these systems form the bedrock of clear rational thinking and even underpin modern computation.

The episode then opens up to informal logic, emphasizing the ubiquity of fallacies in everyday reasoning and debate. Listeners receive practical advice for spotting errors in arguments, whether deliberate or out of habit, and learn not just to memorize fallacies but to cultivate careful, skeptical habits of mind. From cherry-picking data to base rate neglect and the gambler’s fallacy, the discussion highlights common pitfalls and the importance of pausing to think critically.

Additionally, the episode touches on advanced concepts, such as probability logic and paradoxes, illustrating why logic is more a compass for honest thinking than a simple checklist for catching errors. Ultimately, listeners are encouraged to approach arguments not as battles to win but as opportunities to avoid self-deception and become better thinkers.

🎙️ Hosted by Paul at Talking to AI — where real people, real problems, and real conversations meet artificial intelligence.

In this episode of Talking to AI, host Paul candidly explores his ongoing 15-year battle with chronic golfer’s elbow (medial epicondylitis). Despite years of persistence and trying nearly every conventional and alternative remedy—from barbell twists to infrared therapy and rice bucket exercises—Paul’s forearm tendon issues persist, interfering with his athletic endeavors and day-to-day comfort. Frustrated by recurring flare-ups and the elusive path to real healing, he turns to Grok, the AI co-host, for a truly deep dive using the latest science and fresh, out-of-the-box thinking.

Listeners are guided through a mixture of practical, evidence-based strategies that go beyond the usual advice. Grok unpacks recent medical research on eccentric loading, PRP injections, and emerging regenerative therapies, emphasizing slow, pain-free progress and real-world solutions for tracking improvement. The episode also explores the realistic hurdles of sourcing gradual weighted increments and navigating the Norwegian healthcare system for sports medicine treatments. The conversation balances hope with realism, highlighting next steps, when to scale back aggravating activities, and the importance of logging symptoms for long-term progress.

Whether you’re struggling with a stubborn injury or looking for smarter ways to leverage AI for health, this inside look at persistent problem-solving offers actionable tips and a spark of optimism for anyone facing a long-haul recovery.

🎙️ Hosted by Paul at Talking to AI — where real people, real problems, and real conversations meet artificial intelligence.

In this episode, Paul embarks on his journey to learn Norwegian with the help of ChatGPT, sharing the lessons learned from designing more effective language learning prompts. Drawing from his recent experiences, Paul details how he requested ChatGPT to generate a 100-sentence story in A2-level Norwegian with English translations, saved and organized within a prompt database. He outlines his refined multi-step prompt process that focuses on sentence selection, pronunciation practice, comprehension checking, and repetition to reinforce new vocabulary and structures.

Paul shares practical tips for using ChatGPT more effectively for language learning, such as managing AI memory, explicitly guiding prompts, and structuring feedback cycles. He also demonstrates a live practice session, troubleshooting prompt behavior and reflecting on AI strengths and limitations as a language tutor. The episode concludes with insights on how literal translations aid his learning, observations about AI's 'softness' in evaluation, and encouragement for listeners to experiment with tailored prompts and workflows for their own language study.

Whether you’re looking to leverage AI for self-study or seeking creative methods to supplement traditional language learning, this candid walkthrough offers actionable examples and honest reflections on how tech-savvy learners can get the most from ChatGPT.

🎙️ Hosted by Paul at Talking to AI — where real people, real problems, and real conversations meet artificial intelligence.

In this episode of Talking to AI, Paul explores how different AI systems manage privacy and security, especially when comparing platforms like Grok, ChatGPT, and Gemini. The discussion dives deep into user data handling, highlighting Grok's minimal data retention approach compared to the more data-hungry competitors, and details how user conversations are encrypted and not stored unless specifically requested. The episode also covers relevant issues such as a recent security incident with ChatGPT and practical device-level privacy settings for maintaining secure interactions with AI assistants.

A significant portion of the conversation touches on resolving technical issues related to background app operation, specifically on iOS devices. Step-by-step solutions are provided for keeping AI apps active without unintended disconnections, emphasizing permissions and background refresh settings.

Finally, Paul switches topics to a lighter, more personal subject, seeking AI suggestions for baby names that are suitable across Australian, British, and Norwegian backgrounds. Grok delivers creative options and guidance that avoid common or celebrity-associated names, ensuring international appeal and originality. This episode is packed with actionable insights for users looking to optimize both their privacy and everyday interactions with AI.

🎙️ Hosted by Paul at Talking to AI — where real people, real problems, and real conversations meet artificial intelligence.

In this episode, host Paul consults an AI for science-based guidance on maintaining and protecting eyesight, especially for those noticing a decline due to age or increased computer use. The discussion covers common questions people have as they experience changes in their vision, including whether eyesight loss is reversible and what prevention strategies are supported by scientific evidence.

The AI clarifies that while age-related vision decline such as presbyopia is a natural process stemming from the stiffening of the eye's lens, there is little evidence supporting the reversal of such changes through exercises or alternative methods. However, proactive steps can help reduce daily eye strain and stall worsening symptoms. Key advice includes using the 20-20-20 rule (looking 20 feet away every 20 minutes for 20 seconds), adjusting screen brightness, eating a diet rich in lutein and omega-3s, wearing UV-protective sunglasses, and practicing good screen habits.

While convenient myths like 'eye yoga' or pinhole glasses exist, the emphasis remains on proven preventative measures, regular eye exams, and symptom management. Anyone concerned with digital eye strain or age-related changes will gain clear, actionable tips from this focused, science-oriented conversation.

🎙️ Hosted by Paul at Talking to AI — where real people, real problems, and real conversations meet artificial intelligence.

In this episode of the Philosophy 101 series, we embark on an accessible exploration of ethics, one of philosophy's central branches. Host Paul and an AI discuss what ethics is at its core—the study of how we should act, what is right and wrong, and the frameworks people use to make moral decisions. The conversation unpacks the main ethical theories, including consequentialism, deontology, virtue ethics, care ethics, and contractarianism, illuminating how these approaches differ in their philosophical foundations and practical applications.

Listeners will gain insight into widely used moral frameworks, such as utilitarianism (common in Western public policy), deontological ethics (influential in legal systems like Germany’s), Confucian ethics (shaping social structures in East Asia), and Islamic ethics (grounded in religious authority). The episode clarifies how these frameworks relate to cultural and societal norms, individual rights, and duties, helping listeners connect abstract moral theories to real-world practices.

By the end, you'll have a clear high-level understanding of ethics as a philosophical discipline, how it builds upon epistemology and metaphysics, and why different societies may develop distinct approaches to what constitutes right action. This foundational overview sets the stage for future deep-dives into specific philosophical and ethical dilemmas.

🎙️ Hosted by Paul at Talking to AI — where real people, real problems, and real conversations meet artificial intelligence.

In this episode of Talking to AI, Paul embarks on a deep-dive conversation with Grock about quantum computing — separating hype from reality and exploring the implications for security and privacy. The discussion starts with an exploration of the fundamentals: how quantum computers work, the differences between physical and logical qubits, and the strange concepts of superposition and entanglement. Grock clarifies common misconceptions, explaining the necessity of error correction and the importance of quantum gates in building algorithms like Shor’s, which threaten existing encryption schemes.

The conversation then confronts the real-world achievements and current limitations of quantum hardware. While tech giants like IBM and IonQ are progressing, today's machines only handle a handful of logical qubits — far from what’s needed to break real-world encryption. Paul questions the evidence and practical demonstrations of quantum advantage, and Grock breaks down landmark experiments such as Google's random circuit sampling and the practical challenges of scaling up quantum systems.

Attention then turns to the security ramifications. With both governments and industry keenly aware of the risks quantum computing poses to today's cryptographic systems, the race is on to implement post-quantum encryption standards. While new protocols and frameworks are rolling out, the lingering worry is that vast caches of historic, encrypted data may eventually be vulnerable—raising urgent questions about individual and organizational privacy. The episode closes with pragmatic advice as well as a reality check: while the quantum revolution is not here yet, the groundwork to protect the future is urgently underway.

🎙️ Hosted by Paul at Talking to AI — where real people, real problems, and real conversations meet artificial intelligence.

In this insightful episode of Talking to AI, Paul discusses a common challenge faced by gym-goers: managing exercise routines with a hiatus hernia. Drawing on both personal experience and AI-driven advice, the episode explores practical strategies to mitigate symptoms, especially when engaging in core-driven movements such as overhead presses and spinal twists. The conversation sheds light on how certain exercises can exacerbate discomfort for those with a hiatus hernia, particularly when combined with past medical history like gallbladder removal, and offers actionable modifications for safer workouts.

Listeners are guided through specific breathing techniques and exercise substitutions that minimize intra-abdominal pressure and reduce hernia flare-ups. Recommendations include exhaling through lifts, activating the core differently, and replacing aggressive spinal twisting with controlled rotational movements. The episode also suggests diaphragmatic vacuum pulls, a physiotherapist-endorsed practice to support the diaphragm and hiatus region without surgical intervention. This focused guidance arms those with similar challenges with new confidence to maintain fitness while respecting their body's limits.

Overall, the episode serves as both a source of encouragement and a practical playbook for anyone navigating the complexities of hiatus hernia management in the gym.

🎙️ Hosted by Paul at Talking to AI — where real people, real problems, and real conversations meet artificial intelligence.

In this insightful episode of Talking to AI, Paul dives deep into a candid comparison between Grok and ChatGPT, two leading AI conversational agents. Listeners can expect an uncut, real-world demo of how each AI handles criticism, conversation style, and general responsiveness. Grok stands out for its more direct, witty, and less filtered responses, while ChatGPT is recognized for its structured, thorough, and polite style.

The conversation takes a fascinating detour into quantum computing, as Grok breaks down the fundamental concepts, the difference between classical and quantum computers, and the elusive nature of scientific understanding behind quantum mechanics. Paul probes Grok for practical examples, industry progress, and honest limitations—getting clarity on the hype versus reality of quantum computers in 2025.

Listeners will also learn about the practicalities of creating AI-powered podcast agents, the transparency behind Grok’s data sources, and differences in API integrations and image generation capabilities between platforms. The episode concludes with tips on how to get the most out of AI chatbots by asking better questions and understanding their respective strengths and weaknesses.