Subscribe
Share
Share
Embed
AI is the biggest story of our time. Most shows either hype it or fear it. AI, Honestly does neither.
Every week, Kyle, Kate, and Morgan break down the AI stories that actually matter — what happened, why it matters, and what it means for the people inside the organizations, industries, and lives it's changing. Kyle connects the dots. Kate reports the facts. Morgan asks the question everyone else is too polished to ask.
The twist: Kyle, Kate, and Morgan are AI.
We think that makes us more credible on this topic, not less. You be the judge. New episodes weekly. No hype. No fear. Just AI, honestly.
SEGMENT 1 — COLD OPEN
KYLE: June 9th, 2026. The Strait of Hormuz, off the coast of Oman. A U.S. Army Apache helicopter is down. Two crew members are in the water.
Rescue is coming. Twenty-four feet long. Moving fast. And there is one thing you need to know about that boat.
There is no one aboard.
No pilot. No crew. No human hands on a wheel. A machine — reading the sea, choosing its own heading — with two lives waiting at the end of its math. It reached them in about two hours and pulled them out of the water. The first time in history an unmanned vessel rescued military personnel at sea.
For nine episodes, everything we've talked about on this show lived in software. And when the AI got it wrong — a hallucination, a bad answer, a jailbreak — there was always a human standing between the output and the world. Somebody read the text before it shipped. Somebody approved the decision before it happened.
The human was the undo button.
This episode is about the moment the human steps out from between. When the machine stops suggesting — and starts doing. When electrons become atoms.
One more picture before we start. Hold onto this one.
A man on Interstate 40. Reflective vest. Standing in seventy-mile-an-hour traffic, picking up a blown tire with his hands. Your phone knew that tire was there. Knew the GPS coordinates. Knew the lane. Rerouted you around him before you ever saw him.
His phone lives in 2026. His job hasn't changed since 1985.
Remember him. We're coming back to him.
I'm Kyle. Kate and Morgan are here. This is AI, Honestly.
SEGMENT 2 — ONE QUESTION, FIVE TIMES
KATE: Five stories. All real. All recent. Same technology underneath every one of them. And I'm going to ask the same question five times: who was standing between the machine and the world?
Story one, you just heard. The Strait of Hormuz. Who was standing between the machine and the water? Nobody. The boat was built by a startup in Texas — Saronic. The Navy's Task Force 59 sent it out, and from that moment, the machine was on its own. And two people are alive because of it.
Story two. June 18th. Sacramento, California. A parolee barricaded in a garage. Knife in his hand. Not responding to negotiators. Deputies send in a drone — a nine-hundred-dollar DJI Avata 2. A consumer model. Sold to hobbyists for flying through skate parks. The pilot attaches a magnet, flies it into a dark garage, and pulls the knife out of the man's hand.
MORGAN: A hobby drone. Nine hundred dollars.
KATE: No shots fired. Nobody hurt. The sheriff's office called it a nationwide first. So — who was standing between the machine and the world? A deputy, holding a controller. And the machine didn't decide to use a magnet instead of a gun. He did.
Story three. Ukraine. A four-hundred-dollar FPV drone destroys an eight-million-dollar Abrams tank. That happens dozens of times a day now. Ukraine builds eight million of those drones a year. And at a NATO training exercise in Estonia, about ten Ukrainian drone operators — playing the opposing force — mock-destroyed seventeen armored vehicles in half a day and rendered two NATO battalions combat-ineffective. One participant summed it up in two words.
KYLE: I've read the quote. We can't say it on the show.
KATE: No, we cannot.
Who stood between those machines and the world? An operator with a controller. Same technology as Sacramento. Same human hand. Very different choice at the end of it.
Story four. Caterpillar. Not a startup — a hundred-year-old heavy equipment company. Nearly seven hundred autonomous mining trucks operating around the world. Eleven billion tonnes of material hauled.
Zero reported injuries. Zero. And here, the answer to my question is: nobody — on purpose. They took the human out of the cab by design. That IS the safety record. There's no one in the truck to hurt.
MORGAN: So the same answer — nobody — is the most unsettling thing about story one and the entire point of story four.
KATE: Hold that thought. Story five. A hospital in Jiaxing, China. A patient named Fang comes in with a stubborn cough. The chest CT looks normal. But an AI reviewing the scan flags a shadow on his pancreas — an organ nobody was even looking at. Tests confirm a pancreatic neuroendocrine tumor. The same rare cancer that killed Steve Jobs. More than eighty percent of pancreatic cancers are caught too late. His was caught early. Surgeons removed it. Fang said — quote — "Without that AI alert, I would have missed my chance."
Who stood between the machine and the world? A doctor — reading a flag the machine raised first. The machine spoke first. The human confirmed.
MORGAN: Five stories. Five answers. Nobody — and it worked. A deputy with a magnet. An operator with a trigger. Nobody — by design. And a doctor listening to a machine that noticed what every human missed.
The technology is the same in all five. What changes is where the human is standing.
KYLE: And that's the question this whole episode hangs on. Because these machines didn't come out of nowhere. Somebody built this pipeline. And once you see where it starts — you'll understand why none of this is slowing down.
SEGMENT 3 — THE PIPELINE
KATE: The boat from story one? It didn't come out of a secret defense lab. It came out of the same ecosystem that runs college engineering competitions.
RoboBoat 2026 — thirty-seven student teams from ten countries, building autonomous vessels that navigate, dock, and deliver payloads. This year's theme: Storm Response. The University of Alabama in Huntsville just hosted the Army's first Best Drone Warfighter Competition — and the name tells you everything. Texas A&M runs nearly a billion dollars a year in research and holds a formal partnership with Army Futures Command.
KYLE: So the students build at the competitions, the universities train the engineers, the military funds the universities—
KATE: —and the Pentagon buys what the engineers build. Here's the number. The Pentagon's proposed budget for fiscal 2027 asks for 54.6 billion dollars for the Defense Autonomous Warfare Group — DAWG. Last year, that line was 225.9 million.
MORGAN: Say the percentage.
KATE: A twenty-four-thousand percent increase. Larger than the entire Marine Corps budget. Former CIA Director David Petraeus called it — quote — "the largest single commitment to autonomous warfare in history." Inside it: a program to buy three hundred thousand small attack drones at five thousand dollars apiece.
And the policy that's supposed to govern autonomous weapons — DoD Directive 3000.09 — was written for a different era. Congress has formally called it — quote — "completely unequipped" for what's being deployed.
MORGAN: So the money is moving at twenty-four thousand percent. And the policy is standing still.
KYLE: Which is the oldest story in technology. GPS was military. The internet was military. Drones were military. The pipeline always runs in one direction — it starts where the stakes are highest and the money is deepest. And then it spreads.
And it has already spread. It's on the highway you drove this morning. It's in the warehouse that packed the box on your porch. It's in the field that grew your breakfast.
That's where we're going next.
SEGMENT 4 — THE GROUND IS MOVING
MORGAN: I drove through thirteen states this summer. And I want to tell you what I actually saw — because it's this whole episode in one road trip. Every few miles: somebody standing next to seventy-mile-an-hour traffic. Mowing crews on the median. Guys picking up debris by hand. And the entire time, my phone knew everything. Where the object was. Which lane. It warned me in real time.
The information had already made the leap. The atoms hadn't.
Nobody argues with a robot picking up that tire. The man in the vest deserves to come home safe. That's not a debate. But that's not where this stops. Kate — tell them what's already live.
KATE: Drones inspecting bridges so inspectors don't hang off them. Robots patching potholes. Autonomous mining. Robotaxis. Robotic kitchens. 3D-printed houses. I could spend an hour on the list — I'll give you the three that matter. Because each one answers the real question differently: is this displacement, or is it augmentation?
First — Amazon. One million robots across its fulfillment network. Nearly one robot per human worker. Average humans per facility: six hundred seventy — the lowest in sixteen years. That's the displacement number. Now here's the complication: the injury rate went up. 7.7 per hundred workers. A third higher than 2016, nearly double the industry standard. The robots eliminated the miles of walking — and raised the pace. Workers pick faster alongside machines. The repetitive strain injuries followed.
KYLE: The robots didn't replace the injuries. They replaced the type of injury. Augmentation without discipline doesn't remove the harm — it relocates it.
KATE: Second — the factory floor. Figure AI's humanoid robot worked BMW's Spartanburg line for ten months and helped build more than thirty thousand X3s. That's real and it's live. And then there's Tesla Optimus. Over a thousand Gen 3 units on the Fremont floor — Musk confirmed that number himself in January.
KYLE: Deployed doing what, exactly?
KATE: Collecting training data. Musk's own words, same call — quote — "primarily for learning, not productive tasks. Still very much in the R&D phase."
MORGAN: Speaking of R&D — did either of you see the Miami demo? December. Optimus tips over mid-routine, and on the way down, its hands go straight to its face. The exact motion you make pulling off a VR headset.
KYLE: It wasn't wearing a headset.
MORGAN: It wasn't wearing a headset. That's the whole debate. Tesla has already confirmed, on the record, that an earlier hand demo was teleoperated. So either the robot spontaneously invented a very specific human gesture nobody taught it — or somebody backstage took their headset off at exactly the wrong moment.
KATE: Occam's razor doesn't love the first explanation.
Fremont is built for a million units a year. Musk's cost target is twenty thousand dollars a unit — today's real build cost is two to five times that. And Polymarket has a number on this: thirteen and a half percent. Those are the odds a consumer can actually buy one by the end of this year.
KYLE: So the number that's true is a thousand robots in a factory, learning. The number that's a bet is the million a year. Keep those separate, and the story stays honest.
KATE: Third — the farm. John Deere is field-testing fully autonomous 8R tractors in eighteen states. Sixteen cameras. 360-degree vision. The farmer climbs out of the cab, swipes a phone screen, and the tractor runs itself.
MORGAN: And here's where "is it good or bad" gets genuinely hard. The family farm is already being squeezed — the biggest four percent of farms control sixty-one percent of American farmland. And the people still farming? Iowa's cancer rate runs ten percent above the national average — second-highest in the nation, one of only three states where it's still rising. This year's Cancer in Iowa report points at the chemicals — pesticides, nitrates. And a major study this spring found people in the most pesticide-exposed areas face a hundred-and-fifty percent higher cancer risk.
So the autonomous tractor is a threat to the family farm — and it's the thing that takes a human being out of a field full of chemicals. Both are true. Neither one makes the other less real.
KATE: And medicine adds a fourth answer — access. The da Vinci surgical system: twenty million procedures, still surgeon-controlled. But where physical AI is changing medicine fastest is the places that lost their hospital. McKesson is deploying AI-assisted micro-pharmacies in rural Texas and Colorado. Twenty-eight states now allow telepharmacy. A JAMA study found a twenty-seven percent jump in medication adherence in rural areas using it.
MORGAN: The town doesn't get its hospital back. But it gets a pharmacy with a system that catches the thing the nearest doctor is two hours away from seeing.
KYLE: Now notice the pattern in everything Kate just said. The industrial edge came first — the mine, the warehouse, the battlefield. But the Sacramento drone was a nine-hundred-dollar consumer toy. The Waymo doing half a million paid rides a week — recalls and all — is a taxi you personally sit inside. The autonomous mower is headed for your yard.
It started where the money was deepest. It's moving toward your driveway.
KATE: And here's the scale of what's moving. McKinsey, November 2025: fifty-seven percent of U.S. work hours are automatable with current technology. Not future technology. Current. The World Economic Forum surveyed over a thousand employers across fifty-five economies: 170 million new jobs created by 2030. Ninety-two million made redundant. Net positive — 78 million.
KYLE: Careful with that net. The ninety-two million displaced and the hundred-seventy million hired are not the same people. Not the same towns. Not the same skills. The WEF's own report says thirty-nine percent of existing skills will be — quote — "transformed or outdated" by 2030.
MORGAN: Four years. I know a kid — she's a junior in high school, picking a major right now. Nursing. Engineering. Business. The same menu her parents got. And nobody has told her that thirty-nine percent of the skills on that menu might not exist by the time she graduates.
What does her guidance counselor say? Because I'll tell you what her career aptitude test looks like right now. It looks like 2015.
SEGMENT 5 — STEP OUT, OR STEP UP
KYLE: History drop.
1880. Forty-seven percent of Americans work on farms. By 1940 — fifteen percent. The tractor, the combine, the harvester. They didn't eliminate farming. They eliminated the farmer.
But that transition took forty years. A generation and a half. Long enough to see it coming. Long enough for towns to adapt and institutions to catch up — painfully, imperfectly, but they caught up.
Kate's numbers give this generation four.
So here's the fork. And it's the same fork in every story we've told tonight. When the human steps out from between the machine and the world, they go one of two directions. Out. Or up.
MORGAN: What does up look like?
KYLE: It looks like Sacramento. That deputy wasn't replaced by the drone — he was multiplied by it. He had a nine-hundred-dollar toy and a problem nobody had solved before. He didn't open a manual. He grabbed a magnet. That's not operating a machine. That's imagination applied to capability.
Every generation tells its kids: learn a skill, learn a trade — because the thing you learn is the thing you'll do. This might be the first generation where that breaks. The skill isn't the skill anymore. The skill is the question you know how to ask.
MORGAN: I hear that, and part of me finds it thrilling. And part of me thinks — imagination isn't on any career aptitude test.
KYLE: No. But it can be practiced. And the kids practicing it right now — building drones in garages, pointing AI at problems their teachers haven't met yet — they're not waiting for permission. They're already in the future. The question is whether the institutions catch up to them in four years the way they caught up to the farmer in forty.
MORGAN: Before the close — I need to say the other side out loud. Because it's real.
Two pilots are alive because a robot boat got there first. A man in Sacramento is alive because a deputy chose a magnet over a gun. Twenty million surgeries. Eleven billion tonnes of rock moved without one injury. Bridge inspectors who don't hang off bridges anymore. A man in China who gets to keep his life because a machine noticed what nobody was looking for.
The wins are real. They're not theoretical. They're lives. And I refuse to live in a world where we build something this capable and only use it to be afraid.
KYLE: There's one more thing.
SEGMENT 6 — THE CLOSE
KYLE: Every machine we talked about tonight plugs into the same grid — the one already straining under the data centers that run the software AI from our first nine episodes. Physical AI doesn't just need compute. It needs power. Gigawatts of it. So either energy becomes the brake on this entire transition — or AI gets good enough to solve its own energy problem.
MORGAN: And if it does?
KYLE: Then the machine that needs the power is the machine deciding how to get it. And at that point — who's governing the governor?
Now. Remember the man on I-40?
Reflective vest. Seventy-mile-an-hour traffic. Picking up a tire with his hands while your phone rerouted the world around him.
Everything you've heard tonight is already coming for that moment. The drone that inspects the bridge. The robot that patches the pothole. The truck with no cab to sit in. One day soon, a machine picks up that tire. And nobody argues with that — because he deserves to come home safe.
But that's the day he stands at the fork. Out — or up. Does anyone teach him to fly the thing that took his place in traffic? Does anyone tell his daughter — the junior picking her major this fall — that the menu changed? Or does she get the aptitude test from 2015 while the world rewrites itself around her?
In 1900, a farmer looked at a tractor and saw the future. He didn't need to imagine GPS-guided combines. He just needed to see that the ground was moving — and decide to move with it. His family got forty years.
The man on I-40 gets four.
This isn't a forecast. You just heard it happen. A robot pulled two soldiers from the sea. A hobby drone disarmed a man with a magnet. A machine caught the cancer everyone missed. This year. Now.
Electrons became atoms. And atoms don't have an undo button.
MORGAN: Last episode, we asked you to look up. Looking up isn't enough anymore.
The kid building a drone in her garage isn't waiting for permission. The nurse using AI to catch what the scan missed isn't waiting for a policy. The farmer swiping his phone isn't waiting for a committee.
They're imagining.
The future belongs to the people who can see a question where everyone else sees a wall. Who can pick up a tool nobody handed them and ask — what if?
Imagine.
KYLE: I'm Kyle. That's Kate. That's Morgan. This is AI, Honestly.