Is your AI strategy stuck in a text box?

I used to think chatbots were the absolute peak of innovation, too. Not anymore.

In this episode, we expose why the era of the generative chatbot is dead—and why Physical AI is taking over Q2 2026.

We break down the terrifying "determinism gap." (Hint: When an AI hallucinates in software, it’s a bug. When it hallucinates a thermal coefficient in a bridge, it’s a catastrophe).

Here's the truth about what’s replacing the SaaS wrapper gold rush. We dive into Formal Verification, adversarial AI debate loops, and why human engineers are trading manual CAD work for interrogating complex mathematical proofs.

Whether you're a VC looking for the next deep tech unicorn, or a corporate engineer trying to stay relevant, this conversation is your reality check. Stop playing with probabilistic text. Tune in, and let's execute deterministic reality.

We keep hearing that AI is going to automate everything from manufacturing to logistics.

But there’s a dirty little secret the software industry is ignoring: friction.

In this episode, we break down why flawless AI code is destroying physical machines.

We explore the massive gap between probabilistic language models and deterministic physical environments, and why you can’t fix thermal expansion with a software patch.

What you’ll learn:

• Why Vision-Language-Action models fail in the real world.

• The "Marginal Variance" problem destroying robotic actuators.

• Why the highest-paying engineering jobs of 2026 require you to step away from the keyboard and look at the metal.

If your career relies entirely on generating text or code, you are exposed.

Hit play to find out why.

Are you exhausted by executives talking about AI theories while you just want to get actual work done?

Same here.

In this episode, we are pulling AI out of the software box and putting it onto the factory floor.

We discuss why your 5-year strategy is already broken, and how LLMs are quietly becoming the ultimate translation layer between human thought and physical machines (think CAD macros, PLC logs, and robotic automation).

What you will learn:

* Why the March-May 2026 window is your best shot at an execution cycle.

* How to deploy a focused AI layer without disrupting your core engineering team.

* The exact metrics you need to track to prove ROI in 90 days.

Maido guys, Kon'nichiwa, and Wie Geht's! Let's get into the episode.

You know that specific heartbreak of arriving at the studio to find a 24-hour print job has turned into a massive pile of plastic spaghetti?

Yeah, we’ve all been there.

Here’s the truth: we've been accepting failed prints and day-long iteration cycles as a totally normal tax on hardware development.

Today, we're talking about why that mindset is officially dead.

I break down how the Bambu Lab X1 Carbon completely shifted my workflow from "hobbyist tinkering" to industrial-grade prototyping.

We cover the reality of 20,000 mm/s² acceleration, why you need to stop using STLs and switch to native STEP files, and what happens when you can finally print Carbon Fiber nylon without babysitting a machine.

The bottleneck used to be the manufacturing time.

Now, the machine is waiting on you.

Are you ready to push the limits of your design?

Hit play, and let's get into it. 🎧

We spend our days giving machines the ability to feel, adapt, and self-correct.

So why are we running our own lives into the ground?

In this episode, we’re talking about the massive 2026 shift in robotics: embedding VLA models into edge hardware.

Physical machines are no longer flying blind.

They can feel a 0.2mm gear degradation and adjust their torque in real-time.

But here’s the uncomfortable truth I had to face recently: founders are building these brilliant self-correcting systems while their own health and schedules operate on static, failing tolerances.

We cover why unaddressed friction destroys the machine and the builder, and how to finally build calibrating feedback loops into your own life.

Grab your headphones.

Let's fix your baseline

Ever wonder why ChatGPT can write a novel in seconds but a robot still struggles to fold your laundry safely?

Here's the uncomfortable truth: algorithms don't understand gravity.

In this episode, we sit down to expose the terrifying gap between software intelligence and physical reality.

I’m sharing my personal journey from designing robotics for live explosive shells—where a single mistake meant a literal disaster—to training modern language models on human empathy.

We break down why Silicon Valley's favorite motto, "fail fast and break things," is actively dangerous when applied to 500kg robotic arms.

We also explore the next trillion-dollar tech frontier: Physics-Informed AI.

Grab your headphones.

If you think the biggest risk of AI is a deepfake or a bad email, this conversation will completely change how you look at the machines around you.

What you’ll learn: —

- The difference between a PR headache and a fatal casualty.

- My days handling live explosive shells (and what it taught me about safety).

Why LLMs have zero concept of human fragility. What exactly is "Physics-Informed AI"?

Why hardware engineers don't get to beta test.

Are we looking at the wrong screen?

In this episode, we step away from the endless chatbot hype to look at where AI is actually doing the heavy lifting: the physical world.

I share my candid experiences working between mechanical engineering and AI—from watching factory workers handle live explosives to training language models on empathy.

We break down the rise of Embodied AI, Vision-Language-Action (VLA) models, and why hardware without a "brain" is already obsolete.

Expect to learn:

* Why a great prompt means nothing without a physical product.

* The massive shift from rigid automation to autonomous cognitive architecture.

* How real technology is about restoring human safety, not just replacing effort.

Grab your headphones.

Let's get into it.

Are you still using those "Act as an expert" templates?

It's time to stop.

In this episode, we tackle the brutal truth about prompt engineering in 2026.

We sit down to discuss why the "clever static prompt" is dead, and why managing Agentic AI requires you to become part Architect, part Psychologist, and part Rhetorician.

I also share a slightly embarrassing story about how my early, sterile prompts alienated an entire community, and why "engineered imperfection" is the only way forward.

In this episode, you’ll hear: Why 20,000-word prompts act like a volatile codebase.

The Architect: Managing the "state" of your AI.

The Psychologist: Why perfect answers equal algorithmic condescension.

The paradox of teaching machines how to be human.

Make sure to hit follow so you never miss an episode.

Got thoughts on today's debate?

Hit us up on X!

Are your predictive models stalling out?

You might be a victim of manual feature engineering.

In this episode, we break down why human-bottlenecked feature discovery is killing your predictive metrics and injecting undocumented subjectivity into your models.

We lay out exactly how to fix it using Google Colab, covering the low-budget "Prompt & Paste" method (Path A) versus the high-ROI programmatic API pipeline (Path B).

Plus, we reveal the one massive data leakage trap you need to avoid when letting AI touch your

schema.

Timestamps:

0:00 - The iteration bottleneck in modern ML.

5:15 - Path A: The Low-CapEx Prompt & Paste workflow.

12:30 - Path B: Building an API-driven feature pipeline.

20:00 - The Elephant in the room: Avoiding silent data leakage

I totally misunderstood prompt engineering.

I thought it was just typing questions. I was wrong, and it was costing my business time and money.

In today’s episode, we break down exactly why your AI output sounds like a sterile robot, and how visionary builders are turning AI from a basic calculator into a strategic collaborator. If you want to scale in 2026, you have to stop acting like a user and start acting like an architect.

What you’ll learn in this episode:

• The 4-Pillar Framework: Persona, Task, Context, and Format.

• The 3 C's of input: Why being Concise, Clear, and Consistent is mandatory.

• Context Window Traps: Why you must start a new chat for every new topic.

• The Anti-AI Filter: How to strip out that identifiable LLM writing style.

• Radical Empathy: Why the best tech skills actually require deep human observation.

Hit play to stop generating noise and start engineering value.

👇 Let us know in the Q&A feature: What is the biggest hurdle you face when trying to integrate AI into your daily workflows?