Can artificial intelligence actually be funny, or is humor still a human stronghold? We explore that question with Eric Doyle, co-founder of ComedyBytes, a Brooklyn-based multimedia comedy show where AI and humans face off in roast battles, dating games, and other interactive formats. Doyle combines the craft of stand-up with the tools of generative AI, building AI characters like “AI Kanye West” or “AI Sarah Silverman” that deliver pre-scripted jokes in real time.

In this episode of AI-Curious, we dig into:

• [0:52] The story behind ComedyBytes and its AI-powered format
• [3:46] How AI roast battles work, from concept to stage mechanics
• [7:53] Using tools like ChatGPT, Claude Sonnet, and Gemini AI to write jokes
• [12:55] The art of prompting for humor and boosting the “funny hit rate”
• [16:36] Why specificity matters in generative AI comedy
• [23:43] Inside the “Data-ing Game,” an AI twist on the classic dating game
• [25:58] Can AI really be funny—or just imitate the structure of humor?
• [32:30] The triple, listing technique, and other joke-writing structures AI can learn
• [39:10] Advice for non-comedians using AI to add humor
• [41:24] The future of AI in entertainment and its impact on creators

From the structure and anatomy of a joke to the ethics of deepfake comedy, this conversation blends technology, performance, and the evolving role of AI in creative work. Whether you’re an AI enthusiast, a comedy fan, or simply curious about where these worlds collide, this is a look at AI and humor you haven’t heard before.

Is Kant the new code? If AI can write, code, and even plan, which human skills suddenly become scarce—and valuable?

In this conversation with Robert Capps (former Editorial Director of Wired, contributor to The New York Times Magazine), we dive into his widely shared NYT Mag feature, “AI Might Take Your Job. Here Are 22 New Ones It Could Give You.” We unpack the three big buckets of new work he sees emerging—Trust, Integrators, and Taste—and explore why philosophy majors, auditors, and “AI translators” may be the surprise winners. We also get frank about hallucinations, over-extrapolation, inequality, lethal autonomous weapons, and why Rob still comes out more optimistic.

In this episode of AI-Curious, we:

• Break down Rob’s three buckets of future AI jobs: Trust (auditors, ethicists, legal guarantors), Integrators (the translators who know both your business and the models), and Taste (the Rick Rubin-esque role of vision, judgment, and curation).
• Talk about why Ethan Mollick refuses to let AI write his first drafts—and why that matters for your own thinking.
• Examine how “the tools will be commodities, not the people,” and what that means for founders, creators, journalists, and scrappy upstarts.
• Get into the very real risk of inequality and policy paralysis—and why UBI isn’t a satisfying answer.
• Preview Rob’s documentary on AI weapons and the fight to keep humans in the loop.

Takeaways

• Trust work explodes. Expect a cottage industry of auditors, ethicists, and “legal guarantors” to ensure AI output is accurate, defensible, and compliant.
• Integrators win inside companies. The most valuable people will be those who can translate between business reality and fast-moving model ecosystems.
• Taste is leverage. Vision, taste, and editorial judgment—knowing what good looks like—become the human moat.
• Beware first-draft capture. Letting AI write your first draft can quietly dominate your thinking (Mollick’s rule is worth adopting).
• Inequality is the real threat. Most experts Rob spoke with fear a rapid widening of inequality more than mass permanent joblessness.
• Tools, not people, become commodities. When everyone has Goldman-tier tools, expect disruption from the bottom, not reinforcement of the top.

Rob’s NYT Magazine piece: “AI Might Take Your Job. Here Are 22 New Ones It Could Give You.”

https://www.nytimes.com/2025/06/17/magazine/ai-new-jobs.html

Could AI actually improve public education? Not just automate it, but make it more personalized, more equitable — and even more human?

We explore this possibility with Dr. Michael Everest, founder of edYOU, an AI tutoring platform being piloted in a Denver-area school district. While many worry that AI could become a shortcut for students to avoid real learning, Everest argues the opposite — that AI can reinforce understanding, boost confidence, and offer 24/7 support tailored to each student’s needs.

In this episode of AI-Curious, we dig into the real-world mechanics of how this works — including partnerships with schools, how teachers interact with the platform, and what kind of results they’re seeing so far.

We also ask the tough questions: What about data privacy? What about bias and hallucinations? Is there a risk we’re outsourcing critical thinking? And what does the future of education look like if every student has a lifelong AI companion?

Topics include:

• The promise and pitfalls of AI in classrooms
• edYOU’s pilot program with Adams 14 School District
• How the AI tutoring platform personalizes learning
• The role of teachers in an AI-enhanced education system
• Oversight, privacy, and academic integrity
• The vision of a lifelong AI learning companion

Whether you’re a parent, educator, technologist, or just curious about where education is headed, this conversation offers a grounded, hopeful — and at times provocative — look at the future of learning.

What happens when AI becomes a co-pilot for writers, researchers, and journalists — not in theory, but in practice?

In this episode of AI-Curious, we speak with Bill Wasik, Editorial Director of The New York Times Magazine, who recently oversaw their special issue, “Learning to Live with AI.” We explore how AI is already transforming journalism, nonfiction writing, and historical research — and why the most interesting impacts may come not from content creation, but from how we discover, organize, and interpret information.

We dig into the creative tension between AI and human storytelling, including how historians are using tools like NotebookLM to tackle research projects previously deemed impossible. Bill shares how AI can augment writing workflows without compromising editorial judgment — and why trust and authorship still matter in a world of fast content.

We also cover:

• The risks of over-relying on AI for research (19:45)
• How AI might transform local journalism and accountability (41:30)
• The evolving AI policies at The New York Times (29:40)
• Whether AI could ever win the Booker Prize — and what that would mean (7:30)
• Use cases from historians and academics using ChatGPT (26:00)

Bill's (excellent) piece: "AI is Poised to Rewrite History. Literally."

https://www.nytimes.com/2025/06/16/magazine/ai-history-historians-scholarship.html

The NYT Magazine's Special Issue:

https://www.nytimes.com/2025/06/16/magazine/using-ai-hard-fork.html

What are the top minds in AI actually talking about behind closed doors?

At the HumanX conference—arguably the flagship event in the AI ecosystem—hundreds of speakers (from CEOs to policymakers to Kamala Harris) shared their unfiltered thoughts on the state and future of artificial intelligence. But with so much happening at once, even attendees couldn’t absorb it all.

So HumanX did something novel: they partnered with Read.AI to record and synthesize every single session. The result? A real-time AI copilot for the conference and a post-event report that reveals the key themes, trends, and tensions shaping the industry.

In this episode, we speak with HumanX CEO Stefan Weitz and Read.AI CEO David Shim to unpack the insights from that report—what they signal for 2025, what business leaders should pay attention to, and what’s probably just noise.

We talk about the rise of agentic AI, the shift from AGI ambition to ROI expectations, and the practical realities of implementing AI inside large organizations. We also dig into issues of trust, open source, industry-specific adoption, and how AI is starting to reshape roles from customer service to legal to healthcare.

Whether you’re in strategy, ops, tech, or just trying to keep up, this conversation offers a data-driven pulse check on where enterprise AI is headed.

Highlights & Timestamps:

• [1:00] – How Read AI became the official AI copilot of the HumanX conference
• [3:10] – “You can’t be everywhere at once”—the problem this tech solves at events
• [6:15] – The most talked-about concept at HumanX: agentic AI
• [7:45] – Why AGI hype is shifting toward practical use cases with agents
• [8:58] – The fast hype-decay cycle of AI and the emerging focus on outcomes
• [12:26] – Open source, cost savings, and why business leaders care about transparency
• [14:19] – Trust as the “anchoring tenet” of enterprise AI adoption
• [16:45] – Real ROI: how Read AI identified $10M in sales pipeline in 30 days
• [20:03] – Why companies are hiding their AI wins from competitors
• [22:43] – Cross-industry learnings: how healthcare patterns may apply to other sectors
• [25:47] – The “put up or shut up” moment: 2025 as the year AI must deliver
• [29:06] – What business leaders should do before launching AI agent initiatives
• [35:03] – The #1 mistake orgs make with AI: failing to assign ownership
• [37:09] – Predictions: personalization, interoperability, and privacy friction ahead
• [42:28] – How Stefan and David personally use AI—for work, fun, and creative hacking

Links & Mentions:

• HumanX – Flagship AI conference co-founded by Stefan Weitz
• Read AI – Productivity-focused AI platform by David Shim
• Suno – AI music generation tool mentioned by Stefan
• Replit – AI coding sandbox used by Stefan for strategy visualization
• Veo by Google DeepMind – AI video generation tool referenced by David

🎧 Subscribe to AI-Curious:

• Apple Podcasts
https://podcasts.apple.com/us/podcast/ai-curious-with-jeff-wilser/id1703130308

• Spotify
https://open.spotify.com/show/70a9Xbhu5XQ47YOgVTE44Q?si=c31e2c02d8b64f1b

• YouTube
https://www.youtube.com/@jeffwilser

What if we’re all chasing the wrong kind of AI? Dr. Ruchir Puri, Chief Scientist of IBM, argues that Artificial General Intelligence (AGI) is overrated—and that we should be focusing instead on AUI: Artificial Useful Intelligence. This is a pragmatic, business-focused approach to AI that emphasizes real-world value, measurable outcomes, and implementable solutions.

In this episode of AI-Curious, we explore what AUI actually looks like in practice. We discuss how to bring AI into your organization (even if you’re just getting started), why IBM is betting big on small language models (SLMs), and how companies can move beyond hype toward real, trustworthy AI agents that do actual work.

You’ll also hear:

• Why AI usefulness is a function of both quality and cost [00:11:00]
• The “crawl, walk, run” strategy IBM recommends for business adoption [00:14:00]
• Internal IBM examples: HR systems and coding assistants [00:16:00]
• Why SLMs may be a smarter bet than LLMs for many enterprises [00:37:00]
• A breakdown of how agentic systems are evolving to reflect, act, and self-correct [00:41:00]

Whether you’re leading a startup or an enterprise, this conversation will help you reframe how you think about deploying AI—starting not with hype, but with value.

🎧 Subscribe to AI-Curious:

• Apple Podcasts
https://podcasts.apple.com/us/podcast/ai-curious-with-jeff-wilser/id1703130308

• Spotify
https://open.spotify.com/show/70a9Xbhu5XQ47YOgVTE44Q?si=c31e2c02d8b64f1b

• YouTube
https://www.youtube.com/@jeffwilser

What exactly is AGI—Artificial General Intelligence—and how close are we to achieving it? Will it transform the world for better or worse? And how can we even tell when true AGI has arrived?

In this episode of AI Curious, we sit down with Dr. Ben Goertzel, the iconic computer scientist who coined the term AGI more than 20 years ago. As the founder of SingularityNET and the Artificial Superintelligence Alliance, Ben has spent decades thinking about the architecture, risks, and potential of general intelligence.

We explore why today’s large language models (LLMs), while powerful, still fall short of true AGI—and what will be needed to bridge that gap. We dive into Ben’s prediction that AGI could arrive within just 1 to 3 years, and why he believes it will likely be decentralized. Along the way, we unpack some of the key ideas from his recent “10 Reckonings of AGI”—a candid look at the social, economic, and existential questions we must face as AGI reshapes human life.

Topics include:

• [00:04:00] What AGI really means vs. current LLMs
• [00:10:00] Are we reaching the limits of current AI architectures?
• [00:13:00] How will we know when AGI has truly arrived?
• [00:17:00] The “PhD test” for human-level AGI
• [00:19:00] AGI timeline predictions (1–3 years? 2029?)
• [00:29:00] The 10 Reckonings of AGI: key societal impacts
• [00:36:00] The gap between AGI and superintelligence
• [00:44:00] Why a decentralized AGI might be safer
• [00:51:00] Surprising upsides of a post-AGI world

If you’re curious about the future of artificial intelligence, this conversation offers a rare and unfiltered perspective from one of the field’s most original thinkers.

SingularityNet

https://singularitynet.io/

Ben Goertzel on X

https://x.com/bengoertzel

🎧 Subscribe to AI-Curious:

• Apple Podcasts
https://podcasts.apple.com/us/podcast/ai-curious-with-jeff-wilser/id1703130308

• Spotify
https://open.spotify.com/show/70a9Xbhu5XQ47YOgVTE44Q?si=c31e2c02d8b64f1b

• YouTube
https://www.youtube.com/@jeffwilser

What happens when an AI agent says something harmful, or makes a costly mistake? Who’s responsible—and how can we even know who the agent belongs to in the first place?

In this episode of AI-Curious, we talk with Evin McMullen, CEO and co-founder of Billions.Network, a startup building cryptographic trust infrastructure to verify the identity and accountability of AI agents and digital content.

We explore the unsettling rise of synthetic media and deepfakes, why identity verification is foundational to AI safety, and how platforms—not users—should be responsible for determining what’s real. Evin explains how Billions uses zero knowledge proofs to establish trust without compromising privacy, and offers a vision for a future where billions of AI agents operate transparently, under clear reputational and legal frameworks.

Along the way, we cover:

• The problem with unverified AI agents (2:00)
• Why 50% of online traffic is now bots—and why that matters (2:45)
• The Air Canada chatbot legal fiasco (15:00)
• The difference between chatbots and agentic AI (13:00)
• What “identity” means in an AI-first internet (10:00)
• Deepfakes, misinformation, and the limits of user responsibility (22:00)
• Billions’ “deep trust” framework, explained (29:00)
• How platforms can earn trust by verifying content authenticity (34:00)
• Breaking news: Billions’ work with the European Commission (38:20)

This one dives deep into the infrastructure of digital trust—and why the future of AI may depend on getting this right.

Learn more: https://billions.network

🎧 Subscribe to AI-Curious:

• Apple Podcasts
https://podcasts.apple.com/us/podcast/ai-curious-with-jeff-wilser/id1703130308

• Spotify
https://open.spotify.com/show/70a9Xbhu5XQ47YOgVTE44Q?si=c31e2c02d8b64f1b

• YouTube
https://www.youtube.com/@jeffwilser