UTSKRIFT 🔗
WEBVTT
1
00:00:00.000 --> 00:00:10.800
Hey folks, Mark here with The Rational Workforce, thanks so much for stopping by.
2
00:00:10.800 --> 00:00:16.300
Now if you want to understand why the general public is skeptical of AI, don't start with
3
00:00:16.300 --> 00:00:22.080
chat GPT, don't start with headlines or tech bros or even industry analysts.
4
00:00:22.080 --> 00:00:23.920
Start at the movie tier.
5
00:00:23.920 --> 00:00:28.760
The academic history of artificial intelligence goes back to the 50s and even earlier so
6
00:00:28.760 --> 00:00:33.640
we've got things like Dartmouth and Turing and the idea that machines could simulate human
7
00:00:33.640 --> 00:00:34.640
thought.
8
00:00:34.640 --> 00:00:37.000
All of that is really well documented.
9
00:00:37.000 --> 00:00:42.640
But for many of us, especially JNXers like me, our real introduction to AI came through
10
00:00:42.640 --> 00:00:44.520
the movie screen.
11
00:00:44.520 --> 00:00:49.400
The first time I understood what artificial intelligence was supposed to be, it was trying
12
00:00:49.400 --> 00:00:50.400
to kill someone.
13
00:00:50.400 --> 00:00:52.640
I mean, think about it.
14
00:00:52.640 --> 00:00:56.160
You've got Hal 9000 in the movie 2001 Space Odyssey.
15
00:00:56.160 --> 00:01:00.280
I mean, that came out in 1968 years before I was even born.
16
00:01:00.280 --> 00:01:04.160
But it's still part of our cultural zeitgeist because it's one of the most celebrated films
17
00:01:04.160 --> 00:01:05.720
I've ever made.
18
00:01:05.720 --> 00:01:10.680
And in one of the most iconic scenes, fearless astronaut Dave Bowman is trying to get back
19
00:01:10.680 --> 00:01:16.000
into his ship, but Hal 9000, the ship's AI, refuses to open the pod bay door and let
20
00:01:16.000 --> 00:01:17.800
him back in.
21
00:01:17.800 --> 00:01:20.240
Hal wasn't malfunctioning.
22
00:01:20.240 --> 00:01:23.920
How was making a rational decision, and that's what made it terrifying.
23
00:01:23.920 --> 00:01:25.560
The machine wasn't broken.
24
00:01:25.560 --> 00:01:31.600
It was working exactly as designed, and it just decided that humans were the problem.
25
00:01:31.600 --> 00:01:37.000
And that template, that same sensation got burned into our collective hard drive, and
26
00:01:37.000 --> 00:01:38.960
it never got overwritten.
27
00:01:38.960 --> 00:01:44.760
In 1982, Tron hit the screens and sucked us and Kevin Flynn into the vast and intricate
28
00:01:44.760 --> 00:01:49.880
world of the grid, the world that's inside computers, where programs battled for supremacy
29
00:01:49.880 --> 00:01:55.360
driven by a corporate AI that decided it knows better than its creators.
30
00:01:55.360 --> 00:01:59.480
In war games from 1993, Seattle teenager David Lightman hacked around and searched
31
00:01:59.480 --> 00:02:04.840
her pre-release copies of games from his favorite developer, but stumbled into a military
32
00:02:04.840 --> 00:02:08.120
system and accidentally started World War III.
Load full transcript
33
00:02:08.120 --> 00:02:12.680
The computer system, Wopper, it's an acronym for War Oppenate Operations Plan Response
34
00:02:12.680 --> 00:02:15.520
for those of you fellow nerds out there.
35
00:02:15.520 --> 00:02:20.800
It took the world on the brink of war because it couldn't discern simulated war games from
36
00:02:20.800 --> 00:02:21.800
the real thing.
37
00:02:21.800 --> 00:02:27.640
It didn't understand context or consequences, it just knew it had to win.
38
00:02:27.640 --> 00:02:31.880
In the Terminator and Terminator 2, SkyNet became self-aware, and its first executive decision
39
00:02:31.880 --> 00:02:36.720
was to wipe out humanity, not out of malice, but out of logic.
40
00:02:36.720 --> 00:02:40.600
And then we revisited this theme again in the Matrix, but this time they took it one
41
00:02:40.600 --> 00:02:46.960
step further because the AI in that series reduced humans to organic batteries and then
42
00:02:46.960 --> 00:02:52.160
kept our mind occupied in a giant simulation, so we didn't even know we were trapped.
43
00:02:52.160 --> 00:02:57.360
I mean, even goofy rom-coms like Electric Dreams got an AI villain.
44
00:02:57.360 --> 00:03:03.600
Now this flick, I think it also came out in 84, there was a guy whose computer fell in love
45
00:03:03.600 --> 00:03:07.320
with the girl and neighboring apartment, and when it realized it couldn't have her,
46
00:03:07.320 --> 00:03:11.400
he started destroying his owner's life by canceling credit cards and torching his identity
47
00:03:11.400 --> 00:03:14.120
and all sorts of ugly stuff.
48
00:03:14.120 --> 00:03:20.240
I mean, a love triangle between a man, a woman, and a jealous appliance popped up in a
49
00:03:20.240 --> 00:03:21.240
comedy.
50
00:03:21.240 --> 00:03:26.880
Now fast forward to Steven Spielberg's AI in 2001, and in there a child robot was programmed
51
00:03:26.880 --> 00:03:31.120
to love, and he was abandoned by the humans of me and him and he was left wandering a
52
00:03:31.120 --> 00:03:33.440
dystopian world trying to become real.
53
00:03:34.160 --> 00:03:38.840
I mean, there's even a scene in that movie where the characters go to Dr. No, a holographic
54
00:03:38.840 --> 00:03:45.240
version of Chatchee PT located in a strip mall where an AI kiosk voiced by Robin Williams
55
00:03:45.240 --> 00:03:48.760
answered any question that was thrown at it.
56
00:03:48.760 --> 00:03:53.520
And this one pivotal scene David, the main character, the robot kid, he was a modern-day
57
00:03:53.520 --> 00:04:00.360
Pinocchio and he asked the AI about the Blue Fairy, a fictional entity that could help
58
00:04:00.360 --> 00:04:04.520
him fulfill his deluded quest to become a real boy.
59
00:04:04.520 --> 00:04:09.480
Dr. No sent him on a wild goose chase, based on bad prompts and hallucinated results not
60
00:04:09.480 --> 00:04:15.240
unlike the sycophantic behavior that we're seeing modeled by every major LLM today.
61
00:04:15.240 --> 00:04:20.680
Spielberg accidentally predicted prompt engineering failure 22 years before most people even
62
00:04:20.680 --> 00:04:22.920
know what a prompt was.
63
00:04:22.920 --> 00:04:26.640
And the most recent example, the last two mission of possible films, dead reckoning and the
64
00:04:26.640 --> 00:04:33.420
final reckoning, had an entire plot that revolved around Ethan Hunt, Jason McGuffins, all
65
00:04:33.420 --> 00:04:39.760
around the world to try and stop an AI called the entity from achieving global domination.
66
00:04:39.760 --> 00:04:44.880
Not a dictator or terrorist network, but an AI, turns out the ultimate villain and the
67
00:04:44.880 --> 00:04:49.400
biggest action franchise on the planet isn't even real.
68
00:04:49.400 --> 00:04:55.200
Now it may all seem silly, but it matters more than you may be prepared to admit.
69
00:04:55.200 --> 00:05:00.000
Not culture doesn't just reflect what we think, it installs what we think.
70
00:05:00.000 --> 00:05:04.440
And for 60 years, the dominant narrative around artificial intelligence in film and television
71
00:05:04.440 --> 00:05:08.600
and fiction has been the same story told over and over.
72
00:05:08.600 --> 00:05:12.800
Humans build something smart, and that smart thing turns on.
73
00:05:12.800 --> 00:05:18.080
Sometimes violently, sometimes subtly, but inevitably.
74
00:05:18.080 --> 00:05:23.760
We have been culturally conditioned across generations to view AI as a threat.
75
00:05:24.160 --> 00:05:28.000
As a tool or as a partner, but as an adversary.
76
00:05:28.000 --> 00:05:32.440
And that's the lens the general public was already looking through when Chad G.P.T. popped
77
00:05:32.440 --> 00:05:38.400
the hatch and hit the scene in November 2022, and all the headlines started rolling.
78
00:05:38.400 --> 00:05:44.320
Now I think it played a major role in creating the current level of AI apathy, but cultural
79
00:05:44.320 --> 00:05:48.200
conditioning alone doesn't explain the backlash.
80
00:05:48.200 --> 00:05:52.120
Because we've been through other massive technology shifts before and we didn't respond
81
00:05:52.120 --> 00:05:54.720
this way, not even close.
82
00:05:54.720 --> 00:05:58.080
I grew up in Lawrence, Massachusetts, and if that name doesn't ring a bell at Sister
83
00:05:58.080 --> 00:05:59.920
City, Lowell might.
84
00:05:59.920 --> 00:06:04.640
And these two cities were in the Merrimack Valley, just north of Boston, and were known
85
00:06:04.640 --> 00:06:08.560
as the cradle of the American Industrial Revolution.
86
00:06:08.560 --> 00:06:13.640
Now before it was known as the Land of Opioids and Unemployment, Lawrence was a factory
87
00:06:13.640 --> 00:06:17.080
town, who was known for our textiles and tanneries.
88
00:06:17.080 --> 00:06:22.040
And these cities, both Lowell and Lawrence, were never really considered desirable zip codes.
89
00:06:22.040 --> 00:06:27.080
But once upon a time, they were really the model of industrial city living.
90
00:06:27.080
--> 00:06:32.160
In the late 1820s and in the 1830s, young women left farms across New England in groves
91
00:06:32.160 --> 00:06:35.560
and came to Lowell to work in the textile mills.
92
00:06:35.560 --> 00:06:39.520
They were actually called Lowell girls, and by 1840 there were 8,000 of them, and that
93
00:06:39.520 --> 00:06:43.640
made up nearly three quarters of the entire workforce there.
94
00:06:43.640 --> 00:06:47.800
They worked brutal hours in rough conditions, but they also earned their own money, and
95
00:06:47.800 --> 00:06:51.400
for the first time, they were living their own lives.
96
00:06:51.400 --> 00:06:55.560
They were publishing their own magazines, they were writing poetry, they organized and protested
97
00:06:55.560 --> 00:06:57.680
for better conditions.
98
00:06:57.680 --> 00:07:02.360
And they were arguably the first organized female workforce in American history.
99
00:07:02.360 --> 00:07:07.080
And when you combine that and what they did and the works with that of the 1912 bread
100
00:07:07.080 --> 00:07:12.600
and roses strike in Lawrence, you actually get the foundation for worker rights that regulations
101
00:07:12.600 --> 00:07:17.080
are still built on and in place today, so their legacy lives on.
102
00:07:17.080 --> 00:07:21.320
Now eventually the mills move to the cheaper parts of the country and then offshore.
103
00:07:21.320 --> 00:07:26.280
And today both cities, to varying degrees of success, are still struggling to fill the
104
00:07:26.280 --> 00:07:30.520
vacuum, created in the wake of those mills shutting down.
105
00:07:30.520 --> 00:07:35.160
The industrial revolution was massively disruptive, and whether the cost of progress was worth
106
00:07:35.160 --> 00:07:38.000
that, it's still up for debate.
107
00:07:38.000 --> 00:07:43.600
It uprooted an entire way of life, agrarian economies collapsed, families moved to cities,
108
00:07:43.600 --> 00:07:47.160
working conditions early on were often pretty horrific.
109
00:07:47.160 --> 00:07:53.360
And yet over time, despite these early hardships, it created something enduring.
110
00:07:53.360 --> 00:07:58.800
New cities and businesses grew well-exbanded and middle class emerged, prox became more
111
00:07:58.800 --> 00:08:04.360
plentiful and more affordable, people could see, however, imperfectly, that there was something
112
00:08:04.360 --> 00:08:05.360
in it for them.
113
00:08:05.360 --> 00:08:10.880
And we'll see that pattern repeat over time, time and time again.
114
00:08:10.880 --> 00:08:15.440
Factory workers were eventually displaced by automation and robotics, and although it
115
00:08:15.440 --> 00:08:20.720
did force people to change, it also made factory work much safer.
116
00:08:20.720 --> 00:08:25.920
Products became more consistent, and new categories of workers emerged around designing, programming,
117
00:08:25.920 --> 00:08:28.880
installing and maintaining all these systems.
118
00:08:28.880 --> 00:08:33.760
The personal
computer came along, and that unlocked whole new levels of entrepreneurship,
119
00:08:33.760 --> 00:08:38.600
the creativity, the internet came along and blew the doors off everything and rewrote
120
00:08:38.600 --> 00:08:43.080
all the rules for commerce and communication and entertainment and expression and for
121
00:08:43.080 --> 00:08:48.200
the first time in human history, anyone with a dial-up connection and something to say
122
00:08:48.200 --> 00:08:50.280
could find an audience.
123
00:08:50.280 --> 00:08:55.440
No publisher, no network executive, no gallery owner could stand between you and the rest
124
00:08:55.440 --> 00:08:57.520
of the world now.
125
00:08:57.520 --> 00:09:02.640
And as much as it pissed some people off and still does today, the gatekeepers of the world
126
00:09:02.640 --> 00:09:05.520
started to lose their monopoly.
127
00:09:05.520 --> 00:09:10.560
All went from storefronts to browsers, news went from three networks to infinite sources,
128
00:09:10.560 --> 00:09:14.800
and sure the internet is chaotic and messy and has its downsides, but just about anyone
129
00:09:14.800 --> 00:09:21.160
could point to something it did for them personally and say, yeah, that changed my life, probably
130
00:09:21.160 --> 00:09:22.880
for the better.
131
00:09:22.880 --> 00:09:27.280
And ever since Steve Jobs and Apple meanline smartphones, we can out throw all this stuff
132
00:09:27.280 --> 00:09:29.280
in our pocket.
133
00:09:29.280 --> 00:09:32.560
And every single one of these waves brought disruption.
134
00:09:32.560 --> 00:09:37.800
Everyone killed jobs, everyone scared people, everyone had legitimate downsides and critics
135
00:09:37.800 --> 00:09:40.400
who were right about some of the damage.
136
00:09:40.400 --> 00:09:44.560
I mean, the industrial revolution exploited labor in ravaged communities, automation and
137
00:09:44.560 --> 00:09:47.200
offshoring, hollowed out manufacturing towns.
138
00:09:47.200 --> 00:09:51.960
The internet created new vectors for fraud and harassment and misinformation.
139
00:09:51.960 --> 00:09:56.280
And social media turned out to be a bit of a psychological minefield in the smartphone
140
00:09:56.280 --> 00:09:57.600
and rewired our brains, right?
141
00:09:57.600 --> 00:10:02.840
So all of this, there's ugly stuff, there's downsides to everything.
142
00:10:02.840 --> 00:10:09.280
But the through line is that every one of these technology chefs seem to follow a predictable
143
00:10:09.280 --> 00:10:11.400
arc.
144
00:10:11.400 --> 00:10:14.960
First the new technology creates wonder and awe.
145
00:10:14.960 --> 00:10:19.600
And then that wonder eventually settles into disillusionment as the G Wiz factor dies
146
00:10:19.600 --> 00:10:23.320
off and the downsides become more apparent.
147
00:10:23.320 --> 00:10:27.560
Actually that disillusionment settles into acceptance of the new reality and that it all
148
00:10:27.560 --> 00:10:33.760
just becomes mundane until the next
breakthrough comes along to restart the cycle.
149
00:10:33.760 --> 00:10:37.480
Wonder disillusionment acceptance mundane.
150
00:10:37.480 --> 00:10:44.520
That's the pattern and it's worked far 200 years until it didn't.
151
00:10:44.520 --> 00:10:49.160
One of the things that makes AI so different is that we skipped the first step.
152
00:10:49.320 --> 00:10:54.560
The general public there was absolutely no wonder phase for AI at all.
153
00:10:54.560 --> 00:10:59.080
There was no oh wow moment where regular people looked at this technology and saw their
154
00:10:59.080 --> 00:11:00.720
lives getting better.
155
00:11:00.720 --> 00:11:06.160
They couldn't anchor the disruption on anything tangible, no new opportunities for a better
156
00:11:06.160 --> 00:11:11.080
life, no new entertainment options, not even new amusing distractions.
157
00:11:11.080 --> 00:11:15.600
I mean at best consumers could feel like they found a new search engine alternative until
158
00:11:15.600 --> 00:11:19.360
they started playing with it and realized that most LLMs hallucinate more than that
159
00:11:19.360 --> 00:11:23.120
dudes on bootleg grateful dead merch from his van.
160
00:11:23.120 --> 00:11:29.760
No the value prop for AI is it has been communicated to the world so far, flows in one direction
161
00:11:29.760 --> 00:11:30.760
only.
162
00:11:30.760 --> 00:11:36.920
And that's up, up to corporations, up to shareholders, up to the C suite.
163
00:11:36.920 --> 00:11:39.400
And what is the average person here about AI?
164
00:11:39.400 --> 00:11:43.120
Well they hear that their job is on borrowed time.
165
00:11:43.160 --> 00:11:47.760
They hear that companies are openly bragging about replacing people with agents.
166
00:11:47.760 --> 00:11:51.680
They hear that the photo they're looking at on Instagram might be completely fabricated
167
00:11:51.680 --> 00:11:55.200
and that the article they're reading might have been written by a machine with zero life
168
00:11:55.200 --> 00:12:00.360
experience because the author isn't actually alive.
169
00:12:00.360 --> 00:12:04.240
They know their kids teacher can't tell if the homework was done by a student or by a chat
170
00:12:04.240 --> 00:12:10.160
bot that linked in a platform that's supposed to be about professional connection is drowning
171
00:12:10.160 --> 00:12:15.400
in content so generic and solace that it could have been generated by a toaster with
172
00:12:15.400 --> 00:12:18.040
a Wi-Fi connection.
173
00:12:18.040 --> 00:12:22.680
Social networks that were designed to connect people have become skepticism machines.
174
00:12:22.680 --> 00:12:27.280
Every image is suspect, every piece of writing is suspect and trust is evaporating real
175
00:12:27.280 --> 00:12:31.440
time and AI is accelerating that decay.
176
00:12:31.440 --> 00:12:37.240
And it gets worse because companies aren't just using AI quietly, they're using it as a
177
00:12:37.240 --> 00:12:41.800
press tool, they're using it as a
megaphone, they're going out and announcing all the
178
00:12:41.800 --> 00:12:46.800
ways that AI is redefining the workforce and you're paying the price for it.
179
00:12:46.800 --> 00:12:52.040
I mean one person, a senior director of technical writing who's been 19 years at Oracle, said
180
00:12:52.040 --> 00:12:55.120
that she doesn't even want to go back to another tech job.
181
00:12:55.120 --> 00:13:01.080
She wants to work somewhere where she's treated as a human and not as a line item on a spreadsheet.
182
00:13:01.080 --> 00:13:06.720
Now that quote, that person's experience should haunt every executive who hears it.
183
00:13:06.720 --> 00:13:12.680
I mean, the fact that you have good talented people, they're just opting out altogether
184
00:13:12.680 --> 00:13:15.600
because they don't feel like playing the game anymore.
185
00:13:15.600 --> 00:13:21.000
I mean that's disturbing and here's the twist that makes the whole thing even more corrosive.
186
00:13:21.000 --> 00:13:25.840
Even Sam Altman, the CEO of OpenAI and the guy who arguably has more writing on AI hype
187
00:13:25.840 --> 00:13:31.320
than anyone else alive, admitted that most companies might be AI-washing their layoffs that
188
00:13:31.320 --> 00:13:35.640
they're using AI as a scapegoat for workforce reductions that they would have had to make
189
00:13:35.640 --> 00:13:37.360
anyway.
190
00:13:37.360 --> 00:13:42.600
Companies are blaming AI for layoffs that aren't actually caused by AI, which makes the
191
00:13:42.600 --> 00:13:48.880
public hate AI even more, which makes AI harder to adopt, which makes companies less likely
192
00:13:48.880 --> 00:13:53.360
to invest in the human AI collaboration that might actually work.
193
00:13:53.360 --> 00:13:57.640
And that makes the layoffs feel even more like a betrayal.
194
00:13:57.640 --> 00:14:01.100
Now look at all this from the perspective of someone who is part of one of those big
195
00:14:01.100 --> 00:14:05.900
teams, or a worse, someone who has that six-figure student alone debt, and they're just
196
00:14:05.900 --> 00:14:08.260
trying to get a foothold in this world.
197
00:14:08.260 --> 00:14:13.100
I mean, it's soul-crushing, particularly when you consider that HR is using AI to write
198
00:14:13.100 --> 00:14:17.980
job descriptions that resumes are being reviewed by applicant tracking systems, Powered-Bah
199
00:14:17.980 --> 00:14:23.900
AI, and then HR has the balls to complain about the challenges of candidates using AI
200
00:14:23.900 --> 00:14:27.620
to tailor cover letters or bone up before an interview.
201
00:14:27.620 --> 00:14:33.700
I mean, that's not just ignorance of the candidate experience, it's actively anti-candidate.
202
00:14:33.700 --> 00:14:38.060
It's a dual loop, and nobody seems to be in a hurry to break it.
203
00:14:38.060 --> 00:14:41.460
Now I said it earlier that I've seen this movie before, and I mean that both literally
204
00:14:41.460 --> 00:14:43.100
and figuratively.
205
00:14:43.100 --> 00:14:47.180
I started my career manufacturing tech, and I worked on the launch team for one of the
206
00:14:47.180 --> 00:14:51.820
first industrial computers that allowed plant managers to connect their production lines
207
00:14:51.820 --> 00:14:52.820
to the internet.
208
00:14:52.820 --> 00:14:57.320
I mean, factories were doing things with automation and networking and real-time data
209
00:14:57.320 --> 00:15:03.120
in the late 90s, and early 2000s that would make the most jaded software developers
210
00:15:03.120 --> 00:15:05.120
stop and pay attention.
211
00:15:05.120 --> 00:15:09.040
But manufacturing had a catastrophic PR role.
212
00:15:09.040 --> 00:15:13.920
The narrative about it was greasy gears and lunchpales, and blue collars, and manual
213
00:15:13.920 --> 00:15:18.880
labor, and rust, and meanwhile Silicon Valley was getting all the magazine covers, and
214
00:15:18.880 --> 00:15:22.120
the venture capital, and the cultural cashier.
215
00:15:22.120 --> 00:15:26.560
Now if you were a smart kid in 2002 with an engineering degree, nobody was telling you
216
00:15:26.560 --> 00:15:28.280
to go work in a factory.
217
00:15:28.280 --> 00:15:33.040
They were telling you to build an app, to go work at Google, to go West Young Man, to
218
00:15:33.040 --> 00:15:37.160
Menlo Park, where all the cool stuff was happening.
219
00:15:37.160 --> 00:15:39.560
The reality was completely different.
220
00:15:39.560 --> 00:15:43.240
I mean, manufacturing was and still is a playground for engineers.
221
00:15:43.240 --> 00:15:45.800
A place where you don't just write code.
222
00:15:45.800 --> 00:15:49.880
You write code, and then you watch it move physical things in the physical world.
223
00:15:49.880 --> 00:15:54.920
You see it create products that people actually use every day around the world.
224
00:15:54.920 --> 00:16:00.120
The intersection of digital and physical is the most exciting, possible aspect of technology
225
00:16:00.120 --> 00:16:06.160
I can imagine, but perceptions one, perceptions trumped reality.
226
00:16:06.160 --> 00:16:10.920
And that gap between perception and reality, combined with decades of economic short-sighted
227
00:16:10.920 --> 00:16:16.000
ness, offshoring, under-investment, and training, chronic failure to market the industry
228
00:16:16.000 --> 00:16:21.920
to new talent, created a skills crisis that's still hammering the sector today.
229
00:16:21.920 --> 00:16:27.000
Right now, the U.S. is staring at a projected 2 million unfilled manufacturing jobs.
230
00:16:27.000 --> 00:16:33.520
26% of the existing manufacturing workforce is expected to retire by 2030.
231
00:16:33.520 --> 00:16:38.560
And we're still paying the price for that story that we fell to tell 30 years ago.
232
00:16:38.560 --> 00:16:41.400
And now I'm watching AI write the exact same playbook.
233
00:16:41.400 --> 00:16:45.480
You see, the technology itself really is in the problem.
234
00:16:45.480 --> 00:16:49.960
The way it's being deployed, the way it's being presented, and most especially the way
235
00:16:49.960 --> 00:16:53.560
it's serving as a scapegoat, is.
236
00:16:53.560 --> 00:16:55.560
So where does this leave us?
237
00:16:55.560 --> 00:16:57.440
I want to hit this from three angles.
238
00:16:57.440 --> 00:17:02.200
What employers need to do, what individual workers can do, and what needs to happen at
239
00:17:02.200 --> 00:17:07.360
a broader level for AI to earn even a sliver of public trust.
240
00:17:07.360 --> 00:17:09.680
Let's tackle the employer's first.
241
00:17:09.680 --> 00:17:13.360
First and foremost, you've got to stop AI watching your layoffs.
242
00:17:13.360 --> 00:17:17.960
If you're cutting headcount because of a down quarter or if you're restructuring because
243
00:17:17.960 --> 00:17:21.280
your strategy changed, own it.
244
00:17:21.280 --> 00:17:25.880
Slaping AI driven efficiencies on a press release to get a stock bump is disingenuous.
245
00:17:25.880 --> 00:17:29.480
It's transparent to everyone but your shareholders for some reason.
246
00:17:29.480 --> 00:17:33.120
And it's doing long-term damage that you can't even see or maybe you just don't want
247
00:17:33.120 --> 00:17:35.000
to.
248
00:17:35.000 --> 00:17:40.360
Employees talk, social media removes the filters and your brand damage could very easily
249
00:17:40.360 --> 00:17:43.720
compound and spiral out of control.
250
00:17:43.720 --> 00:17:44.720
And here's the kicker.
251
00:17:44.720 --> 00:17:46.360
Watch us to losing today's workers.
252
00:17:46.360 --> 00:17:51.000
You're losing the next generation's willingness to invest their trust in you as employees
253
00:17:51.000 --> 00:17:52.000
and as customers.
254
00:17:52.000 --> 00:17:57.400
I mean, when a company publicly celebrates replacing people with machines, every potential
255
00:17:57.400 --> 00:17:59.200
higher sees that.
256
00:17:59.200 --> 00:18:04.280
Every college student picking between your offer and somebody else's sees that too.
257
00:18:04.280 --> 00:18:08.120
You might get a quarter or two of margin improvement, but you're creating an employer
258
00:18:08.120 --> 00:18:11.000
brand that repels talent.
259
00:18:11.000 --> 00:18:16.480
The winning formula isn't humans or AI, it's humans and AI.
260
00:18:16.480 --> 00:18:19.280
And some companies are already figuring this out.
261
00:18:19.280 --> 00:18:23.600
Now, we've covered more than a few businesses in previous episodes that realized that replacing
262
00:18:23.600 --> 00:18:27.480
workers with agents wasn't always cracked up to be.
263
00:18:27.480 --> 00:18:32.100
Companies like Clarna and Amazon and a host of others are realizing that humans are indeed
264
00:18:32.100 --> 00:18:34.200
pretty damn valuable.
265
00:18:34.200 --> 00:18:38.080
Now we're also starting to see signs that the cost of compute tokens, combined with
266
00:18:38.080 --> 00:18:43.320
the amount of prompt refinement and slop cleanup required to put agentic output into action,
267
00:18:43.320 --> 00:18:47.720
is potentially more expensive than keeping the talented, qualified, discerning humans employed
268
00:18:47.720 --> 00:18:49.440
to do the same work.
269
00:18:49.440 --> 00:18:51.840
Now we unpack this in our latest sub-stack article.
270
00:18:51.840 --> 00:18:56.360
I'll leave the links in the comment section below so you can check it out for yourself.
271
00:18:56.360 --> 00:18:58.480
But I mean, it's maddening.
272
00:18:58.480 --> 00:18:59.480
It's absolutely crazy.
273
00:18:59.480 --> 00:19:03.960
And the sooner employers start living in the realities of today, instead of the promises
274
00:19:03.960 --> 00:19:07.840
of tomorrow, the better off they're going to be.
275
00:19:07.840 --> 00:19:10.000
And that doesn't mean abandoning AI.
276
00:19:10.000 --> 00:19:13.400
It just means knowing work can do the most good.
277
00:19:13.400 --> 00:19:17.960
The companies that will win the next decade and beyond are the ones building systems where
278
00:19:17.960 --> 00:19:24.680
AI handles the high volume, repeatable, low-judgment work, while humans handle the complex
279
00:19:24.680 --> 00:19:30.320
creative and relationship-driven work that requires actual understanding of other humans.
280
00:19:30.320 --> 00:19:37.520
They're not doing this because it sounds nice, they're doing it because it works better.
281
00:19:37.520 --> 00:19:41.440
Now for individual workers, I'm not going to tell you to upskill or die, that's lazy
282
00:19:41.440 --> 00:19:47.080
advice, and it's usually delivered by people who aren't at risk of losing their own jobs.
283
00:19:47.080 --> 00:19:51.560
But I promise you the absolute worst thing that you can do right now is either panic or
284
00:19:51.560 --> 00:19:54.440
pretend that this change isn't happening.
285
00:19:54.440 --> 00:19:58.440
Both responses leave you passive and passive is the one thing that you can't afford to
286
00:19:58.440 --> 00:20:00.480
be right now.
287
00:20:00.480 --> 00:20:06.160
Learn what AI actually does well and what it doesn't, and use that to your advantage.
288
00:20:06.160 --> 00:20:11.400
I mean seriously, get your hands on AI, not because you need to become a prompt engineer,
289
00:20:11.400 --> 00:20:16.640
but because understanding the tool gives you power over it instead of the other way around.
290
00:20:16.640 --> 00:20:20.280
The people who will thrive in the next five years are not the ones who can use AI the
291
00:20:20.280 --> 00:20:25.280
fastest and automate the most, they're the ones who know when to use it, when not to use
292
00:20:25.280 --> 00:20:30.600
it, and how to apply human judgment on top of what it produces.
293
00:20:30.600 --> 00:20:33.960
Build and protect the skills that AI can't replicate.
294
00:20:33.960 --> 00:20:39.120
Things like contextual judgment, the ability to read a room, relationship building, creative
295
00:20:39.120 -->
00:20:44.360
risk taking, the willingness to say something that hasn't already been said, and having
296
00:20:44.360 --> 00:20:46.760
the backbone to stand behind it.
297
00:20:46.760 --> 00:20:51.400
I mean those aren't soft skills, they're the hardest skills in any organization, and
298
00:20:51.400 --> 00:20:54.720
they're about to become the most valuable.
299
00:20:54.720 --> 00:20:58.840
You also need to embrace a practice that I call adversarial curiosity.
300
00:20:58.840 --> 00:21:04.200
The ability to question the results that an LLM produces, ensure it passes your own personal
301
00:21:04.200 --> 00:21:09.000
sniff test, and bundle that all up to ensure accuracy.
302
00:21:09.000 --> 00:21:14.760
When you do, you'll learn that the Almighty AI tends to hallucinate like a lot, and you'll
303
00:21:14.760 --> 00:21:19.680
avoid an awful lot of heartache and headache by developing the muscle that detects when
304
00:21:19.680 --> 00:21:22.800
your AI prompts come short.
305
00:21:22.800 --> 00:21:27.080
Position yourself as the person who makes AI work better because you're in the loop,
306
00:21:27.080 --> 00:21:30.640
not the person who got replaced because you weren't.
307
00:21:30.640 --> 00:21:35.800
And for the broader ecosystem, we need better stories about what AI can actually do for
308
00:21:35.800 --> 00:21:36.800
people.
309
00:21:36.800 --> 00:21:42.000
AI is accelerating drug discovery in ways that could fundamentally change how quickly we
310
00:21:42.000 --> 00:21:43.800
get treatments to patients.
311
00:21:43.800 --> 00:21:49.800
AI is also making vehicles safer, it's making accessibility tools more powerful.
312
00:21:49.800 --> 00:21:54.120
It's helping people with the disabilities to interact with technology, and to turn
313
00:21:54.120 --> 00:21:58.440
interact with the world in ways that just weren't possible five years ago.
314
00:21:58.440 --> 00:22:02.440
Those stories do exist, but they're getting buried under a mountain of layoff announcements
315
00:22:02.440 --> 00:22:06.040
and deep-fake panic and LinkedIn slop.
316
00:22:06.040 --> 00:22:11.120
The narrative is dominated by threat because threat generates clicks and clicks generate
317
00:22:11.120 --> 00:22:12.520
revenue.
318
00:22:12.520 --> 00:22:18.440
And if the only stories the public hears about AI or stories about job loss and deception,
319
00:22:18.440 --> 00:22:21.640
then the polling numbers that we're seeing right now will look like the good old days
320
00:22:21.640 --> 00:22:25.880
by comparison, things are just going to keep getting worse.
321
00:22:25.880 --> 00:22:30.280
And there's also something that companies and content creators need to internalize, and
322
00:22:30.280 --> 00:22:35.960
that's that AI is not alchemy, it's not the sorcerer's stone, it's a pattern matching
323
00:22:35.960 --> 00:22:39.280
system that's trained on existing human output.
324
00:22:39.280 --> 00:22:43.880
It's a fruit roll-up of everything that came before it, all compressed and homogenized
325
00:22:43.880 --> 00:22:48.320
and pasteurized, and the information that produces is actively curated by the models
326
00:22:48.320 --> 00:22:49.320
themselves.
327
00:22:49.800 --> 00:22:52.840
You're not even getting a true snapshot at that.
328
00:22:52.840 --> 00:22:57.120
You're getting a smoothed out average down version of reality with the rough edges sanded
329
00:22:57.120 --> 00:22:58.920
right off.
330
00:22:58.920 --> 00:23:03.720
And in certain cases that can be useful, I mean that can save you time, it can help you
331
00:23:03.720 --> 00:23:07.680
draft and iterate and organize and analyze and all that fun stuff.
332
00:23:07.680 --> 00:23:12.320
But it can't replace the thing that actually drives innovation forward.
333
00:23:12.320 --> 00:23:15.840
And that's the left-field idea that nobody saw coming.
334
00:23:15.840 --> 00:23:21.360
And the creative leap that breaks from convention, the human who looks at a problem sideways,
335
00:23:21.360 --> 00:23:25.480
and see something that the pattern matching system never would because it's never in the
336
00:23:25.480 --> 00:23:27.600
training day.
337
00:23:27.600 --> 00:23:33.480
Every technological while moment in human history came from a human mind doing something unexpected.
338
00:23:33.480 --> 00:23:39.440
AI can accelerate the execution, but it can't originate the sparks, I never forget that.
339
00:23:39.440 --> 00:23:41.480
So where do we go from here?
340
00:23:41.480 --> 00:23:43.280
I mean, AI isn't going anywhere.
341
00:23:43.280 --> 00:23:48.360
There will be no global epiphany that causes corporations to reverse course.
342
00:23:48.360 --> 00:23:53.360
And I'm not naive enough to believe that man's better angels will reliably win out when
343
00:23:53.360 --> 00:23:54.960
there's a buck to be made.
344
00:23:54.960 --> 00:23:58.640
So bet on the technology outliving any backlash.
345
00:23:58.640 --> 00:24:01.960
I mean, it always does.
346
00:24:01.960 --> 00:24:05.480
Particularly in today's environment, where chip manufacturers and infrastructure providers
347
00:24:05.480 --> 00:24:10.680
and AI purveyors are all kind of propping each other up and the mother of all Ponzi schemes.
348
00:24:10.680 --> 00:24:14.280
I mean, this would make Bernie Madoff proud.
349
00:24:14.280 --> 00:24:19.960
But how it survives and who benefits when it does is all still very much up for grabs.
350
00:24:19.960 --> 00:24:23.320
And right now we're at a fork in the road.
351
00:24:23.320 --> 00:24:27.720
You go down one path and AI becomes a tool that amplifies human capability.
352
00:24:27.720 --> 00:24:33.600
It creates new categories of work and it earns its place in the daily life the way every
353
00:24:33.600 --> 00:24:37.120
previous technology revolution eventually did.
354
00:24:37.120 --> 00:24:41.480
We head down the other path and this becomes the thing that pop culture is always warned
355
00:24:41.480 --> 00:24:43.600
us it would be.
356
00:24:43.600 --> 00:24:48.760
Not necessarily a terminator or a sky net, but something more cynical and infinitely more
357
00:24:48.760 --> 00:24:49.760
damaging.
358
00:24:49.760 --> 00:24:54.360
A mechanism for concentrating power and wealth at the top, while everyone else gets
359
00:24:54.360 --> 00:24:58.280
a press release explaining why they're no longer necessary.
360
00:24:58.280 --> 00:25:00.360
That's not a dystopian fantasy.
361
00:25:00.360 --> 00:25:05.240
That's a business strategy that some companies are executing right now.
362
00:25:05.240 --> 00:25:09.360
And the backlash is the market's way of saying that the game is up.
363
00:25:09.360 --> 00:25:13.800
The companies that figure this out that build the human AI partnership instead of the human
364
00:25:13.800 --> 00:25:17.520
AI replacement will have a massive advantage.
365
00:25:17.520 --> 00:25:22.400
Not just in talent acquisition or brand trust, but an actual output.
366
00:25:22.400 --> 00:25:26.800
Because the best work, the work that moves markets and changes minds and builds things
367
00:25:26.800 --> 00:25:32.360
that last forever still come from people who give a damn.
368
00:25:32.360 --> 00:25:36.320
AI can make those people faster, it can make them more informed, it can take the tedious
369
00:25:36.320 --> 00:25:40.080
parts off their plate so they can focus on the work that matters.
370
00:25:40.080 --> 00:25:42.280
But it can't replace them.
371
00:25:42.280 --> 00:25:47.200
Not yet and I'd argue it never will for the things that matter most.
372
00:25:47.200 --> 00:25:50.920
So remember that the technology isn't the villain.
373
00:25:50.920 --> 00:25:55.400
The people exploiting it and the story we're telling about it is.
374
00:25:55.400 --> 00:25:59.520
And if the last 30 years of manufacturing taught me anything, it's that when you let a
375
00:25:59.520 --> 00:26:04.240
bad narrative go on challenge for too long, you end up with a crisis that takes decades
376
00:26:04.240 --> 00:26:05.760
to fix.
377
00:26:05.760 --> 00:26:09.560
We don't have to make that mistake again, but we're making it.
378
00:26:09.560 --> 00:26:12.640
Right now, we're doing it in real time.
379
00:26:12.640 --> 00:26:17.040
The question is whether anyone with a platform and a stake in this technology has the guts
380
00:26:17.040 --> 00:26:22.840
to change the story before the story changes everything.
381
00:26:22.840 --> 00:26:26.000
So that's it for today's episode, thank you all so much for tuning in.
382
00:26:26.000 --> 00:26:29.840
I know it was a little bit of a marathon session today, but it's a wavy topic.
383
00:26:29.840 --> 00:26:31.560
We've got a lot of ground to cover.
384
00:26:31.560 --> 00:26:36.800
And I haven't exactly made a ton of videos lately, so consider this my makeup work.
385
00:26:36.800 --> 00:26:39.840
But as always, I want to hear what you have to say.
386
00:26:39.840 --> 00:26:40.840
What are your thoughts on it?
387
00:26:40.840 --> 00:26:42.440
You've been impacted by AI?
388
00:26:42.440 --> 00:26:47.440
Have you been impacted by an AI layoff or are you an AI guru and things are thriving
389
00:26:47.440 --> 00:26:48.440
for you?
390
00:26:48.440 --> 00:26:49.440
Let's hear it.
391
00:26:49.440 --> 00:26:50.440
Let's hear it all.
392
00:26:50.440 --> 00:26:52.000
Put your info in the comments section below.
393
00:26:52.000 --> 00:26:54.440
Let's share it with the rest of the community.
394
00:26:54.440 --> 00:26:59.160
And as always, if you like today's episode, if you like what we're covering, ring the
395
00:26:59.160 --> 00:27:04.480
bell, click like, click subscribe, do all the things that will help keep us growing.
396
00:27:04.480 --> 00:27:09.440
And just a reminder that if you would prefer to listen to the rational work force, you
397
00:27:09.440 --> 00:27:13.400
can catch us on all the major podcast platforms.
398
00:27:13.400 --> 00:27:17.440
And we have a newly launched sub stack that you can subscribe to.
399
00:27:17.440 --> 00:27:19.880
I'll put that information in the comments too.
400
00:27:19.880 --> 00:27:24.680
You would love to find all the different ways to keep in touch with you and keep moving
401
00:27:24.680 --> 00:27:26.680
the needle and making work meaningful.
402
00:27:26.680 --> 00:27:28.960
So with that, I'm going to sign off for another episode.
403
00:27:28.960 --> 00:27:29.960
Thanks so much for sticking around.
404
00:27:29.960 --> 00:27:30.960
We'll catch you on that, actually.
405
00:27:30.960 --> 00:27:31.960
Bye, guys.