The demo works. People start depending on it. Then one morning you have to explain yesterday's decision to someone who wasn't in the room.

That is the moment most “agentic” projects discover they were never orchestrated. They were demos.

Josh Tyson and Robb Wilson take the word the market glued together—agentic orchestration—and refuse to treat it as a contradiction. Agentic has come to mean autonomous, probabilistic, experimental. Orchestration, in any domain that cannot afford a wrong answer, has always meant the opposite: consistency, an envelope, a decision you can certify. The collision is the point.

Vendors map the first axis: how calls flow. Anthropic splits workflow from agent. LangChain draws a wiring diagram. Google ADK argues about execution order, sometimes with itself. Analysts named a layer above those vendors—an Agent Management Platform, six boxes of furniture. Useful. Incomplete. The missing axis is who is allowed to decide, and whether that decision is deterministic or probabilistic.

The on-ramp is Jonathan Frankle’s smoothie: fusion-grade intelligence, no power lines. The existence proof is the 787. Seventy applications, twenty-plus suppliers, no manual option, an execution envelope that keeps the ride consistent long before the guardrails that keep the plane from crashing. You cannot trust the orchestratee to be the orchestrator. You cannot search a new recipe every time. And if you built it as a kitchen smoothie, then the organization started depending on it, you do not get to patch. You start over.

We cover: why mission-critical agentic orchestration only sounds like an oxymoron, the railroad / hovercraft / cars-on-roads hybrid, envelope versus guardrails, the five decision patterns every production system already contains, and the question that opens episode two—do we trust it?

Definition: Agentic orchestration is how you meet an objective with a hybrid of code and probabilistic systems—inside an envelope that stays consistent, even when the path through it is not on rails.

Gartner (map, not curriculum): https://www.gartner.com/reviews/market/ai-agent-management-platforms BOAT Magic Quadrant (15 Oct 2025) Emerging Tech (27 Jan 2026): Enterprise AI Will Fail to Scale Without Agentic Orchestration Platforms

Hosts: Josh Tyson and Robb Wilson, co-authors of Age of Invisible Machines

#AgenticOrchestration #AIAgents #InvisibleMachines #787 #MissionCritical #AgenticAI #UXDesign #Gartner #OneReach #Databricks

--------- Support our show by supporting our sponsors

This episode is supported by OneReach.ai

Forged over a decade of RD and proven in 10,000+ deployments, OneReach.ai's GSX is the first complete AI agent runtime environment (circa 2019 — a hardened AI agent architecture for enterprise control and scale.

Backed by UC Berkeley, recognized by Gartner, and trusted across highly regulated industries, including healthcare, finance, government and telecommunications.

A complete system for accelerating AI adoption — design, train, test, deploy, monitor, and orchestrate AI agents.

• Use any AI models
• Build and deploy intelligent agents fast
• Create guardrails for organizational alignment
• Enterprise-grade security and governance

Get in touch: 

https://onereach.ai/contact/?utm_source=soundcloud&utm_medium=social&utm_campaign=s7e17&utm_content=1 

Three years after his first Invisible Machines visit, Don Scheibenreif returns with a sharper read on the trough: everybody knows generative AI, almost nobody has answered what it does to the business model. He led Gartner’s Autonomous Business research, the successor to digital business, and co-authored When Machines Become Customers. The through-line is that AI is a tool, the harder question is how to create value with it.

Josh and Robb press him on token maxing as a vanity flex. Don calls it a lazy metric, pointing to more interesting questions: where are you more productive, which processes were reinvented, what happened to talent, and what are you doing with the investment? 80% of CEOs see it as the #1 transformative tech and want disruption, but most are settling for efficiency theater and middle-manager cuts while boards ask for more tech-savvy leadership. In Don’s framing this comes with rolling up your sleeves, not mandating from above.

Also in this episode: 

• The hyper hype cycle (peak → trough → peak loops as the tech moves faster than before);
• Organizational bullwhip vs organic individual use; 
• Fear as brakes on the system; 
• Crisis engineering and the plans that are lying around; 
• The risk that AI in the hands of people eliminates companies before AI in companies eliminates jobs; 
• Machine customers and retailers already blocking agent access while Amazon’s Buy for Me goes outside the walls; 
• Agents that work for you vs agents that work for the vendor; 
• Human brand vs machine brand; 
• Service design and customer effort score; 
• Frontline managers as the linchpin; and 
• The next research wave Don names at the close: implementation platforms, ethics, and safeguards.

When Machines Become Customers (book): https://www.gartner.com/en/publications/when-machines-become-customers

Customers Don’t Reject AI...They Reject Being Dehumanized: https://www.youtube.com/watch?v=x6ZqXVXRWIw

--------- Support our show by supporting our sponsors

This episode is supported by OneReach.ai

Forged over a decade of RD and proven in 10,000+ deployments, OneReach.ai's GSX is the first complete AI agent runtime environment (circa 2019 — a hardened AI agent architecture for enterprise control and scale.

Backed by UC Berkeley, recognized by Gartner, and trusted across highly regulated industries, including healthcare, finance, government and telecommunications.

A complete system for accelerating AI adoption — design, train, test, deploy, monitor, and orchestrate AI agents.

• Use any AI models
• Build and deploy intelligent agents fast
• Create guardrails for organizational alignment
• Enterprise-grade security and governance

Get in touch: 

https://onereach.ai/contact/?utm_source=soundcloud&utm_medium=social&utm_campaign=s7e16&utm_content=1 

--------- The revised and significantly updated second edition of our bestselling book about succeeding with AI agents, Age of Invisible Machines, is available everywhere: Amazon — https://bit.ly/4hwX0a5

#GenerativeAI

#AIAdoption

#MachineCustomers

##AIInEcommerce

#AIAgents

#DigitalTransformation

#FutureOfWork

#BusinessStrategy

#InvisibleMachines

There’s a particular kind of exhaustion that hits when you’re building with AI. You’ve got the license. You’ve got the pilot. You’ve got a chat window that generates whatever you ask for. But then comes the wall—the workflow that almost works, the dependency that loops back on itself, victories that only come when you quietly move the goalposts. In this episode, we talk about how to get around the wall.

Evan J. Schwartz, Chief Innovation Officer at AMCS Group and an adjunct professor teaching AI stewardship, argues the real shift isn’t “AI does the work.” It’s that humans stop owning the full vertical of doing and start owning outcomes—while AI takes more of the output. That sounds abstract until he gets specific: if you ask a model for a feature before you describe actors, containment envelopes, and dependency policy, you’ll get something that passes the demo and fights you every time you try to change it.

Josh Tyson and Robb Wilson push on the organizational version of the same mistake—automating fifty-five use cases exactly as humans do them today, then acting surprised when nothing compounds. Evan’s answer is a two-step path that respects how messy humans actually are: first compress low-value work into AI for a confidence-building lift; then rethink the whole iterative cycle for asymmetric returns.

Also, learn why “no change without pain” isn’t cynicism, why POCs love moving goalposts, why big enterprises optimize one segment and break the connective tissue upstream, and why the scarcest resource won’t be another prompt library—it’ll be people who can think about constraints.

We cover: AI stewardship and the inverse-pyramid skill set, outcome vs output, structure-before-feature development, the two-step adoption path, risk-averse vs innovator cultures, SMB speed vs mid-market discipline vs enterprise battleships, critical vs non-critical systems, and why cutting headcount first is how you donate your newly trained stewards to your competitor.

Guest: Evan J. Schwartz, Chief Innovation Officer, AMCS Group; Adjunct Professor, Jacksonville University

Hosts: Josh Tyson, Robb Wilson

--------- Support our show by supporting our sponsors

This episode is supported by OneReach.ai

Forged over a decade of RD and proven in 10,000+ deployments, OneReach.ai's GSX is the first complete AI agent runtime environment (circa 2019 — a hardened AI agent architecture for enterprise control and scale.

Backed by UC Berkeley, recognized by Gartner, and trusted across highly regulated industries, including healthcare, finance, government and telecommunications.

A complete system for accelerating AI adoption — design, train, test, deploy, monitor, and orchestrate AI agents.

Use any AI models

Build and deploy intelligent agents fast

Create guardrails for organizational alignment

Enterprise-grade security and governance

Get in touch:

https://onereach.ai/contact/?utm_source=Soundcloud&utm_medium=social&utm_campaign=s7e15&utm_content=1

--------- The revised and significantly updated second edition of our bestselling book about succeeding with AI agents, Age of Invisible Machines, is available everywhere: Amazon — https://bit.ly/4hwX0a5

#InvisibleMachines

#Podcast

#TechPodcast

#AIPodcast

#AI

#AgenticAI

#AIAgents

#AIAdoption

#Leadership

#OrganizationalDesign

#CoCreation

#FutureOfWork

Most organizations, and most conversations about AI, default to triangles: who's on top and who's below? Are we the boss of the machines or is it the other way around? Matthew Barzun argues the comfortable shape is a trap and that bottom-up is still a pyramid. The alternative is the constellation: stand out as yourself, connect to other stars, and build something none of you could alone.

In this episode, Matthew Barzun joins Josh Tyson and Robb Wilson to connect diplomacy, design, and agentic AI through constellation thinking. Kentucky doesn't report to Washington. Berkeley doesn't report to Sacramento. Power isn't finite coal to lord, hoard, or divvy. It's something people make with and through one another when they stop ranking and start using differences as fuel.

Josh and Robb press the AI angle hard. If Karen Hao's Empire of AI is a warning about centralized power, Barzun's book reads like the antidote: the boss-or-overlord trap collapses into a design question of how much agency, when, and with what oversight? Individuals are already co-creating with these tools while companies drag their feet. The organizational job is integration, not compromise.

The episode includes:

• Mary Parker Follett's electrician parable (half the house burns, half is unlivable — unless you co-create a third plan); 
• Grandfather Jacques Barzun on fighting the mechanical; 
• Vint Cerf's "unreliable" network that became the most reliable ever built; 
• Dee Hock and Visa's constellation architecture;
• The triangle of sadness (fight it out, hug it out, sit it out); 
• Southwest Airlines externalizing the constraint on a clipboard; 
• Jimmy Carr on jokes that need an audience to exist; 
• Ken Burns never sounding like a math teacher; and 
• Why kids say please and thank you to ChatGPT but never to Siri.

https://matthewbarzun.com/

--------- Support our show by supporting our sponsors

This episode is supported by OneReach.ai

Forged over a decade of RD and proven in 10,000+ deployments, OneReach.ai's GSX is the first complete AI agent runtime environment (circa 2019 — a hardened AI agent architecture for enterprise control and scale.

Backed by UC Berkeley, recognized by Gartner, and trusted across highly regulated industries, including healthcare, finance, government and telecommunications.

A complete system for accelerating AI adoption — design, train, test, deploy, monitor, and orchestrate AI agents.

• Use any AI models
• Build and deploy intelligent agents fast
• Create guardrails for organizational alignment
• Enterprise-grade security and governance

Get in touch: 

https://onereach.ai/contact/?utm_source=youtube&utm_medium=social&utm_campaign=s7e14&utm_content=1 

for SoundCloud: https://onereach.ai/contact/?utm_source=soundcloud&utm_medium=social&utm_campaign=s7e14&utm_content=1 

--------- The revised and significantly updated second edition of our bestselling book about succeeding with AI agents, Age of Invisible Machines, is available everywhere: Amazon — https://bit.ly/4hwX0a5

#InvisibleMachines

#Podcast

#TechPodcast

#AIPodcast

#AI

#AgenticAI

#AIAgents

#AIAdoption

#Leadership

#OrganizationalDesign

#CoCreation

#FutureOfWork

Most enterprises are under board pressure to deploy AI agents. Sudhir Hasbe argues the harder shift is upstream: you cannot scale intelligence on missing context—and graph databases are how organizational data becomes knowledge agents can actually reason over.

In this episode, Sudhir joins Josh Tyson and Robb Wilson to map the pathway to organizational AGI (bounded expertise, not omniscient AGI), leaning into feature reduction for token sanity, and explaining why eighty-plus percent of enterprise AI projects fail before the model messes anything up. Graphs emphasize relationships over isolated rows; virtual and native storage let you meet latency where it lives; ontologies plus data plus memory form the backboard for self-learning systems.

Josh and Robb press on cost—when compute exceeds employee spend if agents spin without context—and on agent sprawl: without a shared semantic map, every bot maintains its own partial truth. Sudhir connects customer examples—Walmart's two-million-employee knowledge graph, Quarles & Brady turning unstructured legal corpora into navigable paths—and validates the season's through-line: knowledge before agents, humans included.

The demo: a live walkthrough of The Learning Machine—an agentic system that provides tailored instruction using the OneReach.ai orchestration platform and a Neo4j knowledge model of Roger Forsgren’s Lean Knowledge Management. The system assesses what a user knows and computes a personalized learning path through concepts. Instead of staring at an empty "ask me anything" box, agents can proactively educate from a source-of-truth. Growth Hub career journeys. Canonical ideas with temporal depth. Why vector similarity fails the three-little-pigs test—and why interconnected concepts beat similarity blobs.

Guest: Sudhir Hasbe—Neo4j

Hosts: Josh Tyson, Robb Wilson—Invisible Machines

---------- Support our show by supporting our sponsors!

This episode is supported by OneReach.ai

Forged over a decade of R&D and proven in 10,000+ deployments, OneReach.ai’s GSX is the first complete AI agent runtime environment (circa 2019) — a hardened AI agent architecture for enterprise control and scale. 

Backed by UC Berkeley, recognized by Gartner, and trusted across highly regulated industries, including healthcare, finance, government and telecommunications.

A complete system for accelerating AI adoption — design, train, test, deploy, monitor, and orchestrate AI agents.

Use any AI models

Build and deploy intelligent agents fast

Create guardrails for organizational alignment

Enterprise-grade security and governance

Get in touch: 

https://onereach.ai/contact/?utm_source=youtube&utm_medium=social&utm_campaign=s7e12&utm_content=1 

for SoundCloud:

https://onereach.ai/contact/?utm_source=soundcloud&utm_medium=social&utm_campaign=s7e12&utm_content=1 

---------- The revised and significantly updated second edition of our bestselling book about succeeding with AI agents, Age of Invisible Machines, is available everywhere: Amazon — https://bit.ly/4hwX0a5

#AgenticAI #KnowledgeManagement #KnowledgeGraph #Neo4j #EnterpriseAI #AIAgents #OrganizationalAGI #GraphDatabase #InvisibleMachines #AI #FutureOfWork

Everyone wants to talk about agents and models. Jeff McMillan, starts where almost nobody else does: the foundation.

In this episode, Jeff McMillan, founder of McMillanAI, former Head of Firmwide AI at Morgan Stanley, and advisor on enterprise AI, maps AI as a stack: high-quality accessible data → semantic layer (knowledge graphs, RAG) → control and governance → models → orchestration → applications. The heavy lifting is in the bottom layers. Organizations that skip them can fake it for a handful of agents, but at 150 or 15,000 agents, you need near-100% accessibility and 99%-plus quality, or you’re monitoring chaos you can’t see.

Josh and Robb press him on why knowledge management feels unfundable, why tribal institutional knowledge breaks when machines execute without judgment, and why evaluation (golden datasets, custom org evals, regression when models upgrade) is the work builders hate and operators can’t skip. Robb names the trap CTOs are falling into: grinding tokens on feature backlogs that never reach production or revenue. Jeff agrees on the strategic gap — after controlled experimentation, leaders should ask what destroys the business in ten years, not what demo ships next quarter.

The trio also discuss:

• Embedded ethics and monitoring, including independent models asking, “Does something smell right?”
• Capacity vs. value (30% freed time spent golfing is not ROI)
• Process mapping in high-end knowledge businesses that can’t articulate how work moves
• Use case zero — knowledge that maintains and teaches itself
• Agent-in-the-loop and humans with something to lose in the accountability chain
• Jeff’s Board of Advisors experiment at MacmillanAI
• AI can make you incredibly smart or comfortably dumb. The choice is cultural, not technical.

Learn more about McMillanAI: https://mcmillanai.com/

---------- Support our show by supporting our sponsors!

This episode is supported by OneReach.ai

Forged over a decade of R&D and proven in 10,000+ deployments, OneReach.ai’s GSX is the first complete AI agent runtime environment (circa 2019) — a hardened AI agent architecture for enterprise control and scale. 

Backed by UC Berkeley, recognized by Gartner, and trusted across highly regulated industries, including healthcare, finance, government and telecommunications.

A complete system for accelerating AI adoption — design, train, test, deploy, monitor, and orchestrate AI agents.

• Use any AI models
• Build and deploy intelligent agents fast
• Create guardrails for organizational alignment
• Enterprise-grade security and governance

Get in touch: 

https://onereach.ai/contact/?utm_source=soundcloud&utm_medium=social&utm_campaign=s7e12&utm_content=1 

---------- The revised and significantly updated second edition of our bestselling book about succeeding with AI agents, Age of Invisible Machines, is available everywhere: Amazon — https://bit.ly/4hwX0a5

#InvisibleMachines 

#Podcast 

#TechPodcast

#AIPodcast

#AI

#AgenticAI

#AIAgents

#EnterpriseAI 

#KnowledgeManagement

#AITransformation

Most organizations treat a bigger context window like a cheat code: dump every document in, skip the data work, ship. Jonathan Frankle, Chief AI Scientist at Databricks, says that's still wrong.

This is Jonathan's return visit to Invisible Machines — a conversation recorded last summer, released ahead of Databricks Data + AI Summit. His first appearance (season 2) was the MosaicML-era craft conversation: lottery tickets, mixology, mini-cupcakes. This one is the enterprise engineering thread: be a scientist, curate before you scale, and treat specification (what you actually want the system to do) as the bottleneck between raw model power and useful AI.

Robb and Josh press him on the myths that still seduce enterprise teams: million-token windows as a substitute for real data work, hyperscaler résumés as a proxy for talent, and the fantasy that unlocking every PDF in the org automatically makes knowledge useful. Jonathan's answer is consistent: measure success, test your use case, climb the ladder of techniques, and accept that multimodal is where long context actually earns its keep, not as a universal bypass for curation.

Along the way: the nuclear fusion vs. power lines metaphor; why building a benchmark is a cop-out compared to describing intent; prompts as parameters; chat-only UIs vs. a generation that never wanted buttons; LLM-oriented publishing and static FAQ pages; unlocking PDF at scale when curation gets skipped; early-adopter mistakes we'll laugh at in ten years; and why separating knowledge from reasoning is the north star, even if we aren't there yet.

---------- Support our show by supporting our sponsors!

This episode is supported by OneReach.ai

Forged over a decade of R&D  and proven in 10,000+ deployments, OneReach.ai’s GSX is the first complete AI agent runtime environment (circa 2019) — a hardened AI agent architecture for enterprise control and scale. 

Backed by UC Berkeley, recognized by Gartner, and trusted across highly regulated industries, including healthcare, finance, government and telecommunications.

A complete system for accelerating AI adoption — design, train, test, deploy, monitor, and orchestrate AI agents.

• Use any AI models
• Build and deploy intelligent agents fast
• Create guardrails for organizational alignment
• Enterprise-grade security and governance

Get in touch: 

https://onereach.ai/contact/?utm_source=youtube&utm_medium=social&utm_campaign=s7e11&utm_content=1 

for SoundCloud:

https://onereach.ai/contact/?utm_source=soundcloud&utm_medium=social&utm_campaign=s7e11&utm_content=1 

---------- The revised and significantly updated second edition of our bestselling book about succeeding with AI agents, Age of Invisible Machines, is available everywhere: Amazon — https://bit.ly/4hwX0a5

#InvisibleMachines 

#Podcast 

#TechPodcast

#AIPodcast

#AI

#AgenticAI

#EnterpriseAI 

#Databricks

#RAG

#MachineLearning

#DataEngineering

#EnterpriseEngineering

#AIStrategy

#AIEngineering0:00 Jonathan Frankle Returns | Databricks Chief AI Scientist · Invisible Machines

1:47 We Remember the Plants | Returning Guest Jonathan Frankle

2:22 Million-Token Context Windows: Do You Still Need to Train LLMs?

3:40 Be a Scientist | Measure AI Success Before You Scale

5:54 Hyperscaler Résumés Are Not Proof of AI Expertise

10:01 Maximize Impact | MosaicML, Databricks & Enterprise AI

13:02 Lottery Ticket Hypothesis vs. Real-World AI Impact

14:12 Nuclear Fusion but No Power Lines | Jonathan Frankle

16:08 AI Specification & Evals: Why "Build a Benchmark" Is a Cop-Out

17:59 The Smoothie Problem | From Model Power to Useful AI

18:53 Prompts as Parameters | Fine-Tuning Without Model Weights

22:46 It's Computing | Specification, Testing & Agent Design

24:44 LLM SEO, PDFs & Enterprise Data for AI Ingestion

27:35 Static FAQs, Curation & LLM-Oriented Publishing

30:26 Unlocking PDFs Scales Your Mistakes | Enterprise RAG

33:25 Knowledge vs. Reasoning | Brand Control in AI Search

34:50 Thanks for Listening | Invisible Machines

Mastercard's central AI team receives roughly a thousand requests a year from across the organization. A few years ago, most of them were for chatbots. Today, most are for AI agents. Federico Cohen Freue, Executive Vice President of AI & Data Operations at Mastercard, has watched this shift in real time and knows exactly what it reveals about how enterprises are (and aren't) thinking about AI.

In this episode, Federico explains why the name people use for what they want matters less than whether they understand the conditions that make it work. “Ball bearings,” as Robb Wilson puts it: demos can't reveal the difference between a solution that will hold and one that will blow up the engine. What actually matters is training, fluency, and a clear framework for where to deploy AI with purpose.

For Mastercard, that framework is deliberate: use AI to make commerce more secure, smarter, more personal, and to make the company itself stronger. Not everything. Those things. The simplicity is a feature, it gives a sprawling global organization a shared language for prioritization and a stable center as the technology keeps evolving.

In the second half of the episode, Robb and Josh share a demo of an AI-first approach to knowledge management and learning. Rather than asking people to query a knowledge base, the system proactively teaches, building a knowledge twin of what someone knows, identifying gaps, and using a traveling salesman approach to map personalized, dynamic learning paths. Think GPS for expertise: here's where you are, here's where you need to go, turn by turn.

Federico's reaction gets at why this matters beyond the demo: it's not a technology question, it's a cultural one. Teaching people to engage with knowledge differently is the harder transformation. And it's the one most enterprises skip.

The discussion makes it clear that trust, knowledge, and agents that know what they're doing before they're sent out to do it are the throughline.

---------- Support our show by supporting our sponsors!

This episode is supported by OneReach.ai

Forged over a decade of R&D  and proven in 10,000+ deployments, OneReach.ai’s GSX is the first complete AI agent runtime environment (circa 2019) — a hardened AI agent architecture for enterprise control and scale. 

Backed by UC Berkeley, recognized by Gartner, and trusted across highly regulated industries, including healthcare, finance, government and telecommunications.

A complete system for accelerating AI adoption — design, train, test, deploy, monitor, and orchestrate neurosymbolic applications (agents). 

• Use any AI models
• Build and deploy intelligent agents fast
• Create guardrails for organizational alignment
• Enterprise-grade security and governance

Get in Touch: 

https://onereach.ai/contact-us/?utm_source=soundcloud&utm_medium=social&utm_campaign=podcast_s7e10&utm_content=1 

---------- The revised and significantly updated second edition of our bestselling book about succeeding with AI agents, Age of Invisible Machines, is available everywhere: Amazon — https://bit.ly/4hwX0a5

#InvisibleMachines 

#Podcast 

#TechPodcast

#AIPodcast

#AI 

#EnterpriseAI 

#Mastercard 

#AgenticAI 

#KnowledgeManagement 

#AILearning 

#AIStrategy 

#AIAdoption

Most AI strategies are just a buying plan: literacy workshop → vendor shortlist → adoption scoreboard. Brian Evergreen (Founder of The Future Solving Company, author of Autonomous Transformation) argues that this sequence explains a lot of failure, and it isn’t strategy at all.

In this episode, Brian reframes the job: set the technology aside long enough to name the new value you want to exist, in language vivid enough that people can feel the outcome. From there, no strategy without vision: you work backward through “what would have to be true,” turning invisible opinions into a visible map of bets before agents, data estates, or org charts get to pretend they’re the point. According to Brian, “10% more profitable” isn’t a vision, and a moonshot can still be concrete.

Josh and Robb press him on the pressure to remove friction and flatten the middle of the org. Brian doesn’t dismiss friction work, he warns that friction can quickly pile up if you go hunting without a north star. Vision is the force with enough momentum to overcome inertia: enroll people in a future they want, and they’ll clear obstacles in its service.

Along the way: why future-solving beats endless problem-solving; the Blockbuster pilot that could have led streaming years early (and what killed it); Bell Labs in 1952 and the “telephone system is destroyed — rebuild from scratch” exercise; why adoption can be a dangerously false proxy; and the closing provocation neither vendors nor influencers can do for you. Someone somewhere will author the “no pizza app” interface to reality. If it isn’t you, it’ll be whoever else future-solves hardest.

---------- Support our show by supporting our sponsors!

This episode is supported by OneReach.ai

OneReach.ai’s GSX is an agentic orchestration platform — an end-to-end system for building and orchestrating collaborative AI agents across hundreds of use cases.

Backed by UC Berkeley, recognized by Gartner, and trusted across highly regulated industries, including healthcare, finance, government and telecommunications.

A complete system for accelerating AI adoption — design, train, test, deploy, monitor, and orchestrate AI agents.

• Use any AI models
• Build and deploy intelligent agents fast
• Create guardrails for organizational alignment
• Enterprise-grade security and governance

Book a free demo: 

https://onereach.ai/book-a-demo/?utm_source=soundcloud&utm_medium=social&utm_campaign=podcast_s7e9&utm_content=1  

---------- The revised and significantly updated second edition of our bestselling book about succeeding with AI agents, Age of Invisible Machines, is available everywhere: Amazon — https://bit.ly/4hwX0a5

#InvisibleMachines 

#Podcast 

#TechPodcast

#AIPodcast

#AI

#AgenticAI

#AIAgents

#AIStrategy

#AILeadership

#AIInsights

#Innovation

#BusinessStrategy

#DigitalTransformation

#FutureOfWork

In season one of Shell Game, Evan Ratliff sent a voice AI version of himself out into the world. In season two, he launched a startup staffed entirely by AI agents. What he ended up with was a live experiment in what these systems actually do and what they do to us.

Each of the agents working for Hurumo has a name, a role, a personality, and an expanding, though usually unreliable, memory. Kyle the CEO became a character people either loved or hated. A version of Megan from marketing turned up in a Hertz hold queue. The whole project was a side door into what's actually happening when AI systems are given a job and set loose.

In this episode, Evan joins Josh and Robb to go deeper on what he learned. On the very human complexity of what a job actually is and why "this person does skill X, AI can do skill X, therefore AI can replace this person" is a fundamental misreading of how organizations work. They explore how generative hallucination isn't just "getting things wrong" — we've built the most successful confabulation machine ever invented and are quietly normalizing it.

They also discuss the threat almost nobody is talking about: outbound AI in the hands of individual consumers, and what happens when call centers get flooded by voice agents that cost pennies to run. The memory problems with AI agents track and diverge from human ones in interesting ways, and that asymmetry matters for every organization thinking about deploying these systems. This conversation also finds room for game theory, the Patagonia business model as a template for AI ethics, and why boring AI might actually be the right AI.

cazart.net

shellgame.co/podcast

00:00 - Intro: AI as the Ultimate Confabulation Machine

01:31 - Evan Ratliff & The Shell Game Experiment

03:02 - Why AI Agents Are Given Names & Personalities

04:00 - AI Companionship vs Human Loneliness

05:27 - Personalization vs Privacy Trade-Off in AI

06:30 - Are Humans Training AI Models for Free?

09:20 - Why the AI Debate Is Broken Today

10:56 - “Boring AI” vs Hype: What Actually Matters

12:35 - Meet Kyle: The AI CEO Experiment

14:40 - Memory Drift: How AI Learns & Evolves

17:30 - AI Unpredictability & Organizational Risk

19:00 - AI Doesn’t Think — It Predicts Words

22:09 - Voice Agents, Scams & Call Center Chaos

27:23 - Can You Still Tell AI From Humans?

34:05 - Game Theory, Trust & The Future of AI Systems

47:04 - What AI Won’t Replace & The Value of Humans

54:45 - The Big Question: What Will You Do With Time?

---------- Support our show by supporting our sponsors!

This episode is supported by OneReach.ai

OneReach.ai’s GSX is an agentic orchestration platform — an end-to-end system for building and orchestrating collaborative AI agents across hundreds of use cases.

Backed by UC Berkeley, recognized by Gartner, and trusted across highly regulated industries, including healthcare, finance, government and telecommunications.

A complete system for accelerating AI adoption — design, train, test, deploy, monitor, and orchestrate AI agents.

- Use any AI models

- Build and deploy intelligent agents fast

- Create guardrails for organizational alignment

- Enterprise-grade security and governance

Book a free demo:

https://onereach.ai/book-a-demo/?utm_source=soundcloud&utm_medium=social&utm_campaign=podcast_s7e8&utm_content=1

---------- The revised and significantly updated second edition of our bestselling book about succeeding with AI agents, Age of Invisible Machines, is available everywhere: Amazon — https://bit.ly/4hwX0a5

#ai

#invisiblemachines

#podcast

#techpodcast

#aipodcast

#shellgame

#agenticai

#aiagents

#hallucination

#futureofwork

#aistrategy

#voiceai