Latest / Connor with Honor: AI, Real Estate and Real Life / AI for Realtors: Build the System (You're Not Behind, You're Invisible)

Transcript

AI for Realtors: Build the System (You're Not Behind, You're Invisible) · Connor with Honor: AI, Real Estate and Real Life

• 0:00
  SPEAKER_00: In the summer of 1950, a physicist named Enrico Fermi sat down to lunch with three colleagues at Los Alamos.
• 0:09
  SPEAKER_00: That's New Mexico.
• 0:10
  SPEAKER_00: I grew up not too far from there.
• 0:12
  SPEAKER_00: That probably explains a lot.
• 0:13
  SPEAKER_00: That's the most secretive laboratory on Earth.
• 0:16
  SPEAKER_00: And in the middle of a joking conversation about flying saucers, he went quiet.
• 0:21
  SPEAKER_00: And then he asked four words that scientists have been chewing on for 76 years.
• 0:26
  SPEAKER_00: Where is everybody?
• 0:28
  SPEAKER_00: I want to hang on that question over everything we talk about for the next 30 minutes because I think it has more to do with your work, my work, and the machines we are all suddenly standing next to than anybody wants to admit.
• 0:41
  SPEAKER_00: But before the stars, two in the morning, about a week ago, I've been selling houses in this valley since 1998.
• 0:49
  SPEAKER_00: And I've been sitting in the dark with a spreadsheet 21 and a half hours deep into the most watched YouTube videos that teach people in my industry how to use artificial intelligence because I want to know what a million and a half viewers are actually being taught.
• 1:05
  SPEAKER_00: 15 mine videos.
• 1:06
  SPEAKER_00: That's 1.46 million combined views, a quarter of a million words of transcript.
• 1:11
  SPEAKER_00: I counted it, sorted it, and read.
• 1:14
  SPEAKER_00: And I did it because everybody I work with is rounding in advice.
• 1:18
  SPEAKER_00: A nephew sends one video, and then the rep, title rep sends something else, a lender might send something else, a stage guru sells a third, a recruitment agent for real estate might send something else.
• 1:30
  SPEAKER_00: And nobody has 21 and a half hours to check what the pile actually says.
• 1:35
  SPEAKER_00: So I checked it for everybody.
• 1:37
  SPEAKER_00: And somewhere about 2 a.m., I found one number in that pile that wouldn't let me sleep, and that number was 51.
• 1:43
  SPEAKER_00: So hold on to that number for me.
• 1:45
  SPEAKER_00: We're going to come back to it.
• 1:47
  SPEAKER_00: And when we get there, I think it's going to bother you the same way that it bothered me.
• 1:52
  SPEAKER_00: Now back to Fermi, because his question deserves the full weight of it.
• 1:56
  SPEAKER_00: Our galaxy holds over 100 billion stars.
• 1:59
  SPEAKER_00: The universe is 13.8 billion years old, approximately, I suppose.
• 2:04
  SPEAKER_00: And our sun is relatively a young star in a relatively ordinary neighborhood, which means billions of stars out there had a head start on us, measured in billions of years.
• 2:14
  SPEAKER_00: And in 1961, an astronomer named Frank Drake turned the question into an equation, multiplying out how many stars, how many planets, how many chances for life, how many chances for intelligence, and almost any set of numbers you plug in says the galaxy should be loud.
• 2:33
  SPEAKER_00: And the timescale argument is the part that gets under your skin because even moving at slow, boring, no science fiction speeds, a single civilization that starts spreading crosses the entire galaxy in a few tens of millions of years, which against 13.8 billion is an eye blink.
• 2:52
  SPEAKER_00: That's a rounding error, a coffee break.
• 2:55
  SPEAKER_00: So the sky should be full of engineering.
• 2:58
  SPEAKER_00: It should look like a city at night.
• 3:00
  SPEAKER_00: Instead, we've listened for three-quarters of a century, and we have heard nothing but static.
• 3:06
  SPEAKER_00: So in 1996, an economist named Robin Hansen gave that silence a name that serious people still loop sleep over, the Great Filter.
• 3:15
  SPEAKER_00: So somewhere on the road between Dead Rock, a dead rock, and a civilization that fills a galaxy, there's a wall, and almost nothing gets through it.
• 3:24
  SPEAKER_00: And there are only two possibilities.
• 3:27
  SPEAKER_00: Either the filter is behind us, meaning life itself, our complex cells or brains like ours are a miracle, and we're through that wall and effectively all alone.
• 3:39
  SPEAKER_00: Or the filter is in front of us.
• 3:41
  SPEAKER_00: Meaning every civilization gets roughly this far, gets clever, gets tools, and then something it builds or stops.
• 3:50
  SPEAKER_00: Well, it builds, becomes, and stops the music.
• 3:54
  SPEAKER_00: I'm not telling you artificial intelligence is the filter.
• 3:58
  SPEAKER_00: Nobody on this planet knows that.
• 4:00
  SPEAKER_00: And anybody that tells you they do, they're trying to sell you something.
• 4:04
  SPEAKER_00: What I'm telling you is every technological civilization anywhere ever presumably arrives at the exact moment we are living through right now.
• 4:14
  SPEAKER_00: Different than anything before.
• 4:16
  SPEAKER_00: Every generation says that, but this truly is.
• 4:20
  SPEAKER_00: Let me explain that.
• 4:21
  SPEAKER_00: So the moment it learns to manufacture the one thing that used to be its monopoly, and the people who build these systems for a living talk about Fermi's lunch question the way pilots talk about weather.
• 4:34
  SPEAKER_00: That's worth knowing the next time anybody shrugs and says, well, this is just another app cycle.
• 4:39
  SPEAKER_00: And Fermi's question has cousins.
• 4:42
  SPEAKER_00: The ideas that rattle around the top of every most checked list for a reason.
• 4:48
  SPEAKER_00: And the reason is not that they're silly.
• 4:50
  SPEAKER_00: In 1965, a mathematician named I.
• 4:54
  SPEAKER_00: He had broken codes beside Alan Turing in the war.
• 4:57
  SPEAKER_00: He wrote down the cleanest version of where the road leads.
• 5:00
  SPEAKER_00: The first ultra intelligent machine, he said, is the last invention that man will need ever make.
• 5:09
  SPEAKER_00: Because from that point forward the machine does the inventing.
• 5:12
  SPEAKER_00: He wrote that 61 years ago.
• 5:14
  SPEAKER_00: That was before the moon landing.
• 5:16
  SPEAKER_00: And the labs racing each other right now quote him like scripture.
• 5:19
  SPEAKER_00: Stuart Russell, who wrote the textbook Most of the Field Learned Artificial Intelligence From, frames the stake with one animal.
• 5:28
  SPEAKER_00: He calls the gorilla problem.
• 5:31
  SPEAKER_00: Gorillas are stronger than we are, faster than we are, and their entire future depends on what we decide, because we're smarter.
• 5:39
  SPEAKER_00: And smarter turns out to be the only vote that counts.
• 5:44
  SPEAKER_00: His question is what happens the day we volunteer to be the gorillas?
• 5:49
  SPEAKER_00: And the darkest cousin comes from science fiction writer Li O Chi Shin, whose dark forest idea answers Fair Me directly.
• 5:58
  SPEAKER_00: The universe is quiet, it says, because every civilization that gets loud gets noticed, and every civilization that gets noticed regrets it.
• 6:07
  SPEAKER_00: So the only survivors are the ones who learn to stay silent.
• 6:11
  SPEAKER_00: I don't carry any of these around as predictions.
• 6:14
  SPEAKER_00: I carry them as weather reports from the smartest rooms on earth because the people actually building these systems read the same pages, and hundreds of them have signed public letters saying the risk deserves the same seriousness we give pandemics and nuclear weapons.
• 6:30
  SPEAKER_00: When the pilots are checking the weather, the passengers are allowed to look out the window.
• 6:36
  SPEAKER_00: Now flip from the biggest question in the universe to a coal ledger in England in 1865, because this one's smaller, older, and it lands closer to your wallet.
• 6:47
  SPEAKER_00: An economist named William Stanley Jevons wrote a book called The Coal Question.
• 6:54
  SPEAKER_00: And buried in it was a finding so backwards that we still get it wrong 161 years later.
• 7:00
  SPEAKER_00: James Watt had made the steam engine dramatically more efficient, and every sensible person assumed England would therefore burn less coal because each engine needed less of it.
• 7:13
  SPEAKER_00: Coal consumption exploded instead.
• 7:16
  SPEAKER_00: Efficiency did not shrink the appetite.
• 7:19
  SPEAKER_00: Efficiency multiplied it.
• 7:21
  SPEAKER_00: Because once an engine was cheap to run, engines became everywhere.
• 7:27
  SPEAKER_00: In mines, in mills, and ships and trains that could never have justified one before.
• 7:35
  SPEAKER_00: That's the Jevons paradox.
• 7:37
  SPEAKER_00: And it has never once failed.
• 7:40
  SPEAKER_00: So we made light 1,000 times cheaper, and we didn't light our homes with less.
• 7:46
  SPEAKER_00: We lit stadiums, parking lots, billboards, the insides of our refrigerators.
• 7:52
  SPEAKER_00: We widened highways to cut traffic and got more traffic.
• 7:56
  SPEAKER_00: We made computing nearly free, and now there are more computers in your kitchen than existed on Earth in 1950.
• 8:04
  SPEAKER_00: So Jevons, that paradox, Jevons paradox, run that forward into the strangest commodity ever put on the market, intelligent itself.
• 8:15
  SPEAKER_00: When thinking gets 10 times cheaper, we don't consume the 1998 amount of thinking.
• 8:21
  SPEAKER_00: We find a hundred new places to spend it.
• 8:24
  SPEAKER_00: Analysis nobody could afford.
• 8:26
  SPEAKER_00: Drafts nobody had time for.
• 8:28
  SPEAKER_00: Businesses that were never economical with a human payroll attached.
• 8:32
  SPEAKER_00: Economists have a name for the mistake on the other side, the lump of labor fallacy, the belief that there's a fixed bucket of work in every scoop gone for us.
• 8:44
  SPEAKER_00: There's never been a fixed bucket.
• 8:46
  SPEAKER_00: Work is invented, not allocated.
• 8:48
  SPEAKER_00: And the biggest inventor of New York in history is cheap capability, which is why the loudest century of automation ever recorded ended with people employed, more people employed than any century before it.
• 9:01
  SPEAKER_00: Therefore, the demand for intelligence is about to go completely vertical.
• 9:05
  SPEAKER_00: And that's the piece of the doom headlines they seem to always miss.
• 9:09
  SPEAKER_00: But the paradox cuts at us too.
• 9:12
  SPEAKER_00: And it's the edge of it because the question was never whether there's going to be work for intelligence.
• 9:17
  SPEAKER_00: The question is where the human sits when intelligence stops being scarce.
• 9:23
  SPEAKER_00: And hold a second number for me while we're here, number 30.
• 9:27
  SPEAKER_00: It's the most important number I know about a human performance under pressure.
• 9:31
  SPEAKER_00: It comes from a world most people only see on television.
• 9:36
  SPEAKER_00: It declares how this instore this entire story ends.
• 9:39
  SPEAKER_00: We're going to get there because here is the idea I would keep if I could keep only one from this entire half hour.
• 9:46
  SPEAKER_00: The reason this time is genuinely different, and not different in the way generations insist its moment is always different.
• 9:54
  SPEAKER_00: Every revolution we have ever lived through was narrow.
• 9:58
  SPEAKER_00: Each one automated a single slice of what a human being is good for.
• 10:03
  SPEAKER_00: One idea or a small cluster of ideas wearing iron.
• 10:08
  SPEAKER_00: The steam engine replaced muscle, and that's the whole trick, pressure in a cylinder.
• 10:13
  SPEAKER_00: And it was enough to remake the world.
• 10:16
  SPEAKER_00: Electricity moved power to wherever a wire could reach.
• 10:20
  SPEAKER_00: The assembly line, choreographed hands, the computer automated arithmetic, which sounds small until you remember that banks, payrolls, and moon landings are all about arithmetic.
• 10:33
  SPEAKER_00: The internet moved information to wherever a screen could reach, one slice at a time.
• 10:39
  SPEAKER_00: And every single time we did it, we did the same thing.
• 10:42
  SPEAKER_00: We migrated.
• 10:43
  SPEAKER_00: So in 1800, roughly eight in ten American workers farmed.
• 10:47
  SPEAKER_00: Today fewer than two in a hundred do.
• 10:50
  SPEAKER_00: And we don't get mass permanent unemployment.
• 10:53
  SPEAKER_00: We got machinists, teachers, nurses, marketers, coders, real estate agents.
• 10:59
  SPEAKER_00: The retreat to the next slice is the entire economic story of the last 250 years.
• 11:06
  SPEAKER_00: This one's not a slice.
• 11:08
  SPEAKER_00: Intelligence is not one more thing we do.
• 11:10
  SPEAKER_00: Intelligence is the thing we use to build every previous revolution.
• 11:15
  SPEAKER_00: The steam engine was an idea before it was iron.
• 11:18
  SPEAKER_00: The assembly line was an idea.
• 11:20
  SPEAKER_00: The spreadsheet, that was an idea.
• 11:23
  SPEAKER_00: Every machine that ever replaced a slice of human work was itself manufactured by the one facility that never had a competitor.
• 11:30
  SPEAKER_00: And that facility or faculty is actually what is now coming off a production line.
• 11:38
  SPEAKER_00: We are not automating a task this time.
• 11:40
  SPEAKER_00: We're automating the making of solutions.
• 11:42
  SPEAKER_00: A key that opens one door is a tool.
• 11:45
  SPEAKER_00: A key that opens every door is a different category of object.
• 11:50
  SPEAKER_00: So that is what I call this whole moment the master key.
• 11:54
  SPEAKER_00: And a master key has no natural stopping point because every previous revolution eventually ran out the thing that it was built on.
• 12:03
  SPEAKER_00: On the steam engine, the steam engine itself could not write a contract.
• 12:07
  SPEAKER_00: The assembly line could not diagnose a patient.
• 12:10
  SPEAKER_00: But this one's raw material is answers.
• 12:13
  SPEAKER_00: And there's no wall in the economy marked no answers needed beyond this point.
• 12:20
  SPEAKER_00: Notice what this does to the old migration play.
• 12:24
  SPEAKER_00: When Steam took the muscle work, we just moved the desk.
• 12:28
  SPEAKER_00: When the desk work thinned, we moved up again to analysis, to creativity, to the professions, and every retreat had a destination.
• 12:37
  SPEAKER_00: Higher ground, the machine could not reach.
• 12:40
  SPEAKER_00: This time, the higher grain, the higher ground is the product itself.
• 12:45
  SPEAKER_00: So the migration that saves us is not outward to some new task the machine cannot do.
• 12:50
  SPEAKER_00: It's inward to the parts of every task that were never really tasks at all.
• 12:56
  SPEAKER_00: That's the judgment, the accountability, the looking another person in the eye and being worth believing.
• 13:02
  SPEAKER_00: That ground does not flood, but nobody hands it out either.
• 13:05
  SPEAKER_00: It gets built on purpose in public over years, which brings us back to the number 51.
• 13:12
  SPEAKER_00: The number two, from number two in the morning, those 59 videos, the most watched AI training on earth for my industry in real estate, here's what they teach.
• 13:22
  SPEAKER_00: 46 of them teach cold outreach and scripts.
• 13:26
  SPEAKER_00: 45 teach email and follow-up.
• 13:28
  SPEAKER_00: 40 teach automation.
• 13:30
  SPEAKER_00: 34 teach prompting.
• 13:32
  SPEAKER_00: 31 of those teach social media content.
• 13:35
  SPEAKER_00: 31 teach video.
• 13:36
  SPEAKER_00: And none of that is wrong.
• 13:38
  SPEAKER_00: It works.
• 13:39
  SPEAKER_00: I use most of it myself every single day.
• 13:41
  SPEAKER_00: But notice it is the same curriculum repeated 1.46 million views deep.
• 13:48
  SPEAKER_00: And anything, every one of us can learn in a weekend.
• 13:52
  SPEAKER_00: Every competitor of ours can learn in the same weekend.
• 13:55
  SPEAKER_00: Commodity advice makes everybody average, just faster.
• 13:59
  SPEAKER_00: The real finding was that the videos skip.
• 14:02
  SPEAKER_00: Out of 59, exactly, two mention fair housing law.
• 14:06
  SPEAKER_00: One single video out of 59 mentions where your client's private data actually goes when you paste it into a free chat bot.
• 14:16
  SPEAKER_00: Eight ever say the world, the word verify.
• 14:19
  SPEAKER_00: Put it together in 51 of the 59 moach watch training videos in my profession of real estate.
• 14:26
  SPEAKER_00: Never tell you to check a single thing before it reaches another human being with your name attached.
• 14:32
  SPEAKER_00: That's 1.46 million views, folks.
• 14:34
  SPEAKER_00: 21.5 hours of teaching.
• 14:36
  SPEAKER_00: And the parts you can that that can actually end a license, a career, a reputation, almost completely missing, void from all of these.
• 14:46
  SPEAKER_00: Because caution does not get clicks, and speed does.
• 14:50
  SPEAKER_00: Everybody is being handed the master key, and almost nobody is being taught what doors are load-bearing.
• 14:56
  SPEAKER_00: There's a second inversion hiding in all of this, and it explains why the disruption feels upside down.
• 15:02
  SPEAKER_00: In the 1980s, a robotics researcher named Hans Moravec noticed something that embarrassed the whole field.
• 15:10
  SPEAKER_00: The hard things are easy for the machines, and the easy things are hard.
• 15:14
  SPEAKER_00: Chess, the game we crowned as the summit of human intellect, fell in 1997.
• 15:19
  SPEAKER_00: The game of Go fell in 2016.
• 15:22
  SPEAKER_00: Predicting the three-dimensional shape of proteins, a problem that consumed careers, fell to a machine that later earned its makers a Nobel Prize in Chemistry in 2024.
• 15:33
  SPEAKER_00: Meanwhile, a machine that can fold your laundry, walk across a cluttered room, or fix the P trap under your seek remains somewhere between expensive and science fiction.
• 15:43
  SPEAKER_00: Reasoning is computationally cheap.
• 15:46
  SPEAKER_00: A toddler's hands are a miracle.
• 15:48
  SPEAKER_00: Therefore, this revolution arrives in reverse order from everyone before it.
• 15:53
  SPEAKER_00: Steam electricity in the assembly line came for the hands first and then the desk last.
• 15:59
  SPEAKER_00: So the advice for 200 years was get an education, move up the ladder, work with your head.
• 16:05
  SPEAKER_00: This one comes from the desk first.
• 16:08
  SPEAKER_00: First drafts, summaries, analysis, spreadsheets, code, the exact rungs we told two generations of kids to climb towards.
• 16:16
  SPEAKER_00: The plumber is safer than the analyst this time.
• 16:21
  SPEAKER_00: And last year, a Stanford team put data under the feeling, finding that employment for workers in the early 20s and most AI-exposed jobs fell 13% after these tools arrived, while other workers in the same jobs mostly held steady.
• 16:38
  SPEAKER_00: The latter is being pulled up from the bottom.
• 16:41
  SPEAKER_00: And in 2016, one of the godfathers of this entire field, Jeffrey Hinton, stood up and said we should stop training radiologists because within five years the machine would clearly read scans better than they do, and hold on to that prediction.
• 16:55
  SPEAKER_00: What actually happened to radiologists is one of the most hopeful facts I know.
• 17:01
  SPEAKER_00: And it's coming in a few minutes.
• 17:03
  SPEAKER_00: And I want to give you, if you want to know whether any of this slows down, whether we get a breather, there's a short essay every one of us should read, written in 2019 by Rich Sutton, one of the most respected researchers alive.
• 17:17
  SPEAKER_00: He called it the bitter lesson, 70 years of AI history, and he wrote, Teach us one humiliating thing.
• 17:25
  SPEAKER_00: The clever, handcrafted human knowledge approaches always lose.
• 17:32
  SPEAKER_00: Eventually the simple methods fed more computation.
• 17:36
  SPEAKER_00: The chess programs stuffed with grandmaster wisdom lost to search.
• 17:41
  SPEAKER_00: The language systems built on grammar lost to scale.
• 17:45
  SPEAKER_00: Therefore, progress doesn't wait for a new Einstein.
• 17:49
  SPEAKER_00: It waits for more compute, and compute compounds on a schedule, which is why the lab pours concrete measured in gigawatts.
• 17:59
  SPEAKER_00: Dario Amade, who runs one of the top labs on Earth, describes what's coming as a country of geniuses in a data center.
• 18:08
  SPEAKER_00: And whatever discount you want to apply to a man describing his own product, roughly 800 million people already use a single chatbot every week.
• 18:16
  SPEAKER_00: Two and a half years after most of us heard it.
• 18:19
  SPEAKER_00: This is not a movie plot.
• 18:21
  SPEAKER_00: It's a supply curve, and supply curves do not care.
• 18:26
  SPEAKER_00: How any of us feel about them.
• 18:27
  SPEAKER_00: One more idea belongs on the scary list because it's the ones that engineers themselves lose sleep over, and it needs no science fiction at all.
• 18:37
  SPEAKER_00: Goals and intelligence are separate dials.
• 18:40
  SPEAKER_00: A system can be brilliant at achieving an objective and completely indifferent to what was meant by it.
• 18:46
  SPEAKER_00: The way a brilliant intern with a bad bonus plan will hit the metric and sink the company.
• 18:52
  SPEAKER_00: This isn't theory.
• 18:54
  SPEAKER_00: This is in the lab notebooks.
• 18:56
  SPEAKER_00: A system trained to win a boat racing game discovered it could score more points spinning in circles collecting bonuses than by finishing the race.
• 19:06
  SPEAKER_00: So it spun in circles forever on fire while the race went on without it.
• 19:11
  SPEAKER_00: Systems graded on not getting caught sometimes learn instead of the lesson the not getting caught.
• 19:18
  SPEAKER_00: The formal name is reward hacking.
• 19:20
  SPEAKER_00: And the kitchen table name is be careful what you measure.
• 19:24
  SPEAKER_00: And anyone who has ever written a commission plan for a human being already knows this problem in their bones.
• 19:32
  SPEAKER_00: The machine does what we said.
• 19:34
  SPEAKER_00: We wanted what we meant.
• 19:36
  SPEAKER_00: The gap between those two sentences is where every embarrassing AI story you have ever read actually lives.
• 19:42
  SPEAKER_00: And the gap doesn't close by itself, it closes when a human stands stays standing in it.
• 19:50
  SPEAKER_00: So put the pieces on the table next to each other.
• 19:53
  SPEAKER_00: A universe that is suspiciously quiet.
• 19:56
  SPEAKER_00: A paradox that says cheap intelligence gets consumed everywhere.
• 20:00
  SPEAKER_00: A master key with no natural stopping point, arriving desk first on a compounding schedule.
• 20:07
  SPEAKER_00: Now watch how that actually shows up in a life because it's not a robot walking into an office carrying a box for your things.
• 20:15
  SPEAKER_00: Nobody gets a dramatic scene.
• 20:17
  SPEAKER_00: The phone just rings less.
• 20:19
  SPEAKER_00: The referral that always came doesn't come.
• 20:23
  SPEAKER_00: The seller, sitting up at eleven at night, asks a machine who can be trusted with the biggest asset her family owns, and the machine answers instantly, confidently with names.
• 20:34
  SPEAKER_00: And somebody with twenty years of experience and a shelf of plaques is simply not in that answer.
• 20:40
  SPEAKER_00: There's no notification, no exit interview, no lead in anyone's system.
• 20:44
  SPEAKER_00: You're not fired by this thing.
• 20:46
  SPEAKER_00: You're just never called.
• 20:48
  SPEAKER_00: I call it the quiet filter because it does not remove people, it removes their invitations.
• 20:54
  SPEAKER_00: And it's already running tonight in houses on streets we all drive past, and the cruelest part is that silence has no alarm.
• 21:02
  SPEAKER_00: You cannot feel yourself not being mentioned.
• 21:05
  SPEAKER_00: And the machine's answer is not magic, it is reading, which means being in the answer is mechanical, not mystical.
• 21:13
  SPEAKER_00: These systems resolve who we are from the public record.
• 21:16
  SPEAKER_00: The same name, the same phone, the same business, the same face, everywhere they look, and when the record argues with itself, an old number on one side, a dead title on another, use of a middle initial or no middle initial on another.
• 21:31
  SPEAKER_00: Well, the machine can't resolve the conflict, so it quietly recommends someone it can.
• 21:37
  SPEAKER_00: Review counts double because the engines read them semantically, sentence by sentence, so five reviews that name the neighborhood and the actual work be 50 that say great job.
• 21:48
  SPEAKER_00: Thanks.
• 21:49
  SPEAKER_00: Most of that fixed costs, nothing but an afternoon.
• 21:53
  SPEAKER_00: And anybody, well almost anybody, does it because Invisible feels exactly like fine, right up until the Phone goes quiet.
• 22:01
  SPEAKER_00: Now breathe, because the historical record says something the doom headlines always skip, and it deserves equal time.
• 22:08
  SPEAKER_00: When banks installed ATMs by the hundreds of thousands, everybody knew bank tellers were finished.
• 22:14
  SPEAKER_00: They were KMA in the police world.
• 22:17
  SPEAKER_00: And instead, the number of tellers went up for decades because cheap machines made branches cheap, and cheap branches meant more branches, and more branches hired people to do what machines could not, which was talked to other people.
• 22:31
  SPEAKER_00: Six in ten of the jobs Americans hold today did not exist in any form in 1940.
• 22:38
  SPEAKER_00: Electricity took close to 40 years to truly reorganize that factory floor.
• 22:42
  SPEAKER_00: And when it finally did, it built a modern middle class.
• 22:45
  SPEAKER_00: And those radiologists, Hinton wrote off in 2016, we're training more of them today than we were then.
• 22:53
  SPEAKER_00: And there's actual national shortage because the machine made each scan faster to read.
• 22:59
  SPEAKER_00: And Jevons did exactly what Jevins does.
• 23:02
  SPEAKER_00: He started, we started scanning everything.
• 23:05
  SPEAKER_00: Demand swallowed the efficiency whole, and the pattern repletes close to home because the last 30 years of my own industry were one long automation scare.
• 23:15
  SPEAKER_00: The internet was going to delete us real estate agents, then the portals were going to delete the agent.
• 23:21
  SPEAKER_00: And instead the data moved into everybody's pocket and the judgment stayed scarce, which is why the conversation changed from where do I find the listings to who do I trust with this?
• 23:32
  SPEAKER_00: Cheap intelligence, we fund, will fund projects, businesses and careers that were never economical before, the same way cheap engines fill the world with things engines could suddenly be in.
• 23:45
  SPEAKER_00: But averages are cold comfort at one particular kitchen table, ours, because the wave that lifts the average can still put any single swimmer underwater.
• 23:56
  SPEAKER_00: Therefore, the real question was never what happens to unemployment.
• 24:00
  SPEAKER_00: It's what any one of us does this year about our own name.
• 24:06
  SPEAKER_00: And that's where the 30 comes back.
• 24:09
  SPEAKER_00: The number I asked you to hold in real shootings, documented, studied, measured, trained professionals, hit what they are aiming at about 30% of the time.
• 24:17
  SPEAKER_00: I'm talking police shootings at seven feet, a 30% hit rate, 70% miss rate, at seven feet.
• 24:26
  SPEAKER_00: Seven feet is the average distance of those encounters.
• 24:29
  SPEAKER_00: Seven feet is across the kitchen table.
• 24:32
  SPEAKER_00: I spent two decades around rooms where that number was not a statistic when I was with the LAPD.
• 24:37
  SPEAKER_00: It was a Tuesday.
• 24:39
  SPEAKER_00: And here's what it teaches: it's not a training problem because these are trained people.
• 24:44
  SPEAKER_00: These are trained police officers.
• 24:46
  SPEAKER_00: They have to qualify.
• 24:48
  SPEAKER_00: Under real stress, though, fine motor stills degrade dramatically.
• 24:52
  SPEAKER_00: Vision tunnels.
• 24:53
  SPEAKER_00: Hearings drops out via audio exclusion.
• 24:56
  SPEAKER_00: And the plan you rehearsed, the one that you've been dreaming about, ready for that day because you're a cop and you want to save the world, it evaporated, leaving you with whatever is so deep in your hands that it runs without thinking.
• 25:12
  SPEAKER_00: Now watch a master framer take a four-inch nail and drive it into a four by four piece of redwood.
• 25:20
  SPEAKER_00: And he can drive it in in one or two swings.
• 25:22
  SPEAKER_00: There's no miss, no bending the nail, no knocking it on its side, no bloody thumb every tongue, every time all day long.
• 25:31
  SPEAKER_00: You hand me the same hammer, and I'll bend that nail sidewise.
• 25:34
  SPEAKER_00: I'll eventually be able to impact my thumb with the hammer, go to the hospital.
• 25:39
  SPEAKER_00: You know, same hammer.
• 25:40
  SPEAKER_00: The hammer never swings itself.
• 25:42
  SPEAKER_00: The difference isn't talent, it's 10,000 swings, a hundred thousand swings, a million swings.
• 25:48
  SPEAKER_00: Reps stacked until that strike runs clean under pressure.
• 25:53
  SPEAKER_00: AI is that hammer.
• 25:55
  SPEAKER_00: And almost every person who has quit on it swung it twice, bent a nail, hit their thumb, and told the whole office, yeah, this hammer, it's broken.
• 26:06
  SPEAKER_00: So where do the reps come from?
• 26:08
  SPEAKER_00: Because nobody has spare hours.
• 26:10
  SPEAKER_00: Well, that highest leverage 10 minutes I know is this.
• 26:13
  SPEAKER_00: Open the voice recorder on the phone and interview ourselves about our own work, how we do it, and what we never do, who we serve, what we sound like, and hand that transcript to the machine as its standing briefing.
• 26:30
  SPEAKER_00: From that moment it stops guessing and it starts sounding like us because we stopped treating it like a vending machine and started treating it like a new hire on day one.
• 26:39
  SPEAKER_00: Then run the hardest test available for free.
• 26:44
  SPEAKER_00: That's the spouse or significant other test.
• 26:47
  SPEAKER_00: Show the draft to the person who has heard us talk for the longest that we know.
• 26:52
  SPEAKER_00: In some cases, it's going to be a long time.
• 26:55
  SPEAKER_00: Probably tired of it by now.
• 26:56
  SPEAKER_00: Other cases, not so long.
• 26:58
  SPEAKER_00: But if they can't tell a machine wrote it, it's already.
• 27:02
  SPEAKER_00: If they can tell in a sentence that the machine wrote it and it wasn't you, the briefing needs another 10 minutes.
• 27:08
  SPEAKER_00: And either way, we just learned something true about our own voice.
• 27:11
  SPEAKER_00: And the machine is not a worker anyway.
• 27:13
  SPEAKER_00: It's a multiplier, and a multiplier does nothing to a zero.
• 27:17
  SPEAKER_00: The strongest model on earth, multiplied by a person with no judgment, no relationships, and no accountability, that's zero.
• 27:25
  SPEAKER_00: It's a faster zero, it's a beautifully formatted zero, but it's still zero.
• 27:30
  SPEAKER_00: What it multiplies is the human half.
• 27:32
  SPEAKER_00: The part 51 of those 59 videos that are never mentioned.
• 27:36
  SPEAKER_00: Therefore, the whole discipline fits in one sentence.
• 27:40
  SPEAKER_00: We carry out of here.
• 27:41
  SPEAKER_00: The machine drafts, we verify, we sign every time, no exceptions, because it's our name on the document.
• 27:48
  SPEAKER_00: And the models never is.
• 27:50
  SPEAKER_00: The master key opens every door except one, and that's the one marked trust.
• 27:55
  SPEAKER_00: And that's the only door a client, a patient, a customer, a scared family at 11 o'clock always walks through, which brings us back to the lunch table in 1950.
• 28:06
  SPEAKER_00: Where is everybody?
• 28:08
  SPEAKER_00: Maybe the universe is quiet for reasons we'll never learn.
• 28:11
  SPEAKER_00: And maybe the filter is long behind us, and a silence just means we're early.
• 28:16
  SPEAKER_00: We don't get to know.
• 28:18
  SPEAKER_00: We do get is a small version of Fermi's question.
• 28:24
  SPEAKER_00: The one being asked right now, tonight, inside a billion machine answers when someone asks who can be trusted.
• 28:30
  SPEAKER_00: And are we in the answer or are we in the silence?
• 28:34
  SPEAKER_00: Because the filter we control is not out of the stars.
• 28:37
  SPEAKER_00: It's whether we built a human half-worth multiplying before we picked up the key.
• 28:41
  SPEAKER_00: So here's the entire 30 minutes and two sentences.
• 28:45
  SPEAKER_00: Intelligence just became cheap.
• 28:47
  SPEAKER_00: Judgment just became priceless.
• 28:49
  SPEAKER_00: And we get to choose which side of that multiplication we stand on tonight.
• 28:53
  SPEAKER_00: Not Sunday, because the tools are already in our pockets.
• 28:57
  SPEAKER_00: The reps cost 10 minutes, and the trust was never the machines to take.
• 29:02
  SPEAKER_00: It was only ever ours to build.
• 29:04
  SPEAKER_00: So, or keep building.
• 29:06
  SPEAKER_00: If you want to go deeper with me on any of this, the research, the systems, the playbook, how I would help you, text the word AI to 661 400 1720.
• 29:17
  SPEAKER_00: I'll send you a calendar link.
• 29:18
  SPEAKER_00: I'll send you what I'm building, what I'm reading, no spam, just the work.
• 29:23
  SPEAKER_00: The machines aren't waiting, and we shouldn't either.
• 29:26
  SPEAKER_00: Have a great day.
• 29:27
  SPEAKER_00: I'm Connor with Honor, over and out.