WEBVTT

00:00:00.000 --> 00:00:02.419
Imagine for a second that you're holding like

00:00:02.419 --> 00:00:05.059
just a standard wooden handled hammer. Okay,

00:00:05.099 --> 00:00:07.679
standard hammer. Got it. Right. And you close

00:00:07.679 --> 00:00:10.699
your eyes and you tap the steel head of that

00:00:10.699 --> 00:00:13.859
hammer against a brick wall. Yeah. Your brain

00:00:13.859 --> 00:00:16.059
doesn't just, you know, register a vibration

00:00:16.059 --> 00:00:18.940
in the palm of your hand. It does something far

00:00:18.940 --> 00:00:22.140
more wild. It definitely does. It actually maps

00:00:22.140 --> 00:00:25.859
the very head of the hammer as a temporary part

00:00:25.859 --> 00:00:28.620
of your own physical body. Like you literally

00:00:28.620 --> 00:00:31.000
feel the texture of a wall at the end of the

00:00:31.000 --> 00:00:33.340
tool, not your fingertips. Right, which is just

00:00:33.340 --> 00:00:35.719
incredible. It's this well -documented psychological

00:00:35.719 --> 00:00:38.719
phenomenon known as proprioceptive extension.

00:00:39.740 --> 00:00:42.219
Proprioceptive extension. Exactly. We don't just

00:00:42.219 --> 00:00:45.619
use tools as separate external objects. We fluidly,

00:00:45.619 --> 00:00:47.859
instantly extend our physical boundaries through

00:00:47.859 --> 00:00:50.320
them. Yeah. Our neurological map literally grows

00:00:50.320 --> 00:00:52.859
to encompass the wood and the steel. Which is

00:00:52.859 --> 00:00:54.859
perfectly fine, right? I mean, it's even miraculous

00:00:54.859 --> 00:00:57.460
for a simple hand tool. Yeah. But consider what

00:00:57.460 --> 00:00:59.539
happens when the tool you are extending yourself

00:00:59.539 --> 00:01:02.420
through isn't made of wood and steel. Right.

00:01:02.539 --> 00:01:05.680
What if it's a sophisticated, opaque, algorithmic

00:01:05.680 --> 00:01:08.219
intelligence? What happens to your mind and your

00:01:08.219 --> 00:01:11.219
agency when you are outsourcing not just your

00:01:11.219 --> 00:01:14.260
physical reach, but your perception of reality?

00:01:14.459 --> 00:01:16.480
And your daily decision making. Exactly. And

00:01:16.480 --> 00:01:19.159
even your moral judgments. Well, we undergo this

00:01:19.159 --> 00:01:23.040
profound and... honestly largely unexamined shift

00:01:23.040 --> 00:01:26.340
in what it means to be human. We're no longer

00:01:26.340 --> 00:01:28.659
just extending our physical bodies to hit a nail.

00:01:28.989 --> 00:01:31.750
We're extending our cognitive architecture and

00:01:31.750 --> 00:01:34.450
our ethical agency into a black box. A total

00:01:34.450 --> 00:01:37.469
black box. Yeah. The stakes of the tools we use

00:01:37.469 --> 00:01:39.969
change completely when the tool has the capacity

00:01:39.969 --> 00:01:43.409
to talk back, to filter information, and to subtly

00:01:43.409 --> 00:01:45.409
alter the very parameters of how we view the

00:01:45.409 --> 00:01:47.049
world. And that is exactly what we are getting

00:01:47.049 --> 00:01:49.810
into today. Welcome to another deep dive. Today's

00:01:49.810 --> 00:01:52.609
mission might sound a little paradoxical at first

00:01:52.609 --> 00:01:54.489
glance. A bit of a stretch. But stick with us.

00:01:54.549 --> 00:01:57.129
We are going to decode how a 2 ,500 -year -old

00:01:57.129 --> 00:01:58.799
nearly forgotten Chinese... Chinese philosophical

00:01:58.799 --> 00:02:01.980
concept might actually be the ultimate foundational

00:02:01.980 --> 00:02:04.819
key to solving the modern crisis of artificial

00:02:04.819 --> 00:02:08.139
intelligence. It is a massive claim. But the

00:02:08.139 --> 00:02:10.219
source material really backs it up. It really

00:02:10.219 --> 00:02:13.379
does. We're grounding this exploration in a brilliant

00:02:13.379 --> 00:02:16.400
piece of newly published research from March

00:02:16.400 --> 00:02:19.520
2026, featured in the journal Synthese. Such

00:02:19.520 --> 00:02:21.659
a great journal. Yeah, always good stuff. The

00:02:21.659 --> 00:02:24.620
work is by Wing -Yin Pang from Tsinghua University,

00:02:24.879 --> 00:02:28.340
and it's titled Shinshin and the Ethics of Belief

00:02:28.340 --> 00:02:31.219
in Human -AI Relations. Which is a mouthful.

00:02:31.530 --> 00:02:33.349
But we'll break it down. OK, let's unpack this.

00:02:33.729 --> 00:02:36.389
Usually when we hear the phrase AI ethics, we

00:02:36.389 --> 00:02:39.270
immediately jump to, you know, highly technical,

00:02:39.389 --> 00:02:41.750
modern jargon. Right. We think about data filters.

00:02:42.169 --> 00:02:45.729
Exactly. Or bias mitigation software, algorithmic

00:02:45.729 --> 00:02:49.490
weight adjustments or those endless congressional

00:02:49.490 --> 00:02:51.909
hearings about data privacy. The glowing server

00:02:51.909 --> 00:02:54.210
racks and the panicked headlines. Yeah. But this

00:02:54.210 --> 00:02:56.289
analysis forces us to step away from all that.

00:02:56.370 --> 00:02:58.430
It asks us to look at the moral, philosophical

00:02:58.430 --> 00:03:01.240
and almost biological bedrock of what. it actually

00:03:01.240 --> 00:03:03.900
means to rely on an external agent. It demands

00:03:03.900 --> 00:03:06.360
that we look at the deep architecture of trust

00:03:06.360 --> 00:03:09.379
itself. The analysis takes us back to the Mohists.

00:03:09.460 --> 00:03:12.000
The Mohists, right. Yeah, an ancient Chinese

00:03:12.000 --> 00:03:14.419
philosophical movement. We're looking at how

00:03:14.419 --> 00:03:17.500
their sage kings managed the problem of delegation.

00:03:18.180 --> 00:03:21.150
Because the question isn't just whether... an

00:03:21.150 --> 00:03:23.430
algorithm is mathematically accurate, the real

00:03:23.430 --> 00:03:25.610
question is whether the tools you interact with

00:03:25.610 --> 00:03:28.090
every single day, your search engine, your social

00:03:28.090 --> 00:03:30.990
media feed, your digital assistant, are acting

00:03:30.990 --> 00:03:34.129
as faithful, virtuous extensions of your own

00:03:34.129 --> 00:03:36.900
intentions. Or... And this is the scary part.

00:03:37.060 --> 00:03:39.439
If they are structurally designed to undermine

00:03:39.439 --> 00:03:42.539
them for someone else's benefit. Exactly. But

00:03:42.539 --> 00:03:44.780
before we get too deep into the algorithm side,

00:03:45.039 --> 00:03:47.379
we should probably set a ground rule. Yes, definitely.

00:03:47.580 --> 00:03:50.159
I want to say a quick ground rule for our discussion

00:03:50.159 --> 00:03:52.659
today, especially as we get into the real world

00:03:52.659 --> 00:03:55.139
implications of this philosophy. It gets a bit

00:03:55.139 --> 00:03:57.400
heavy. It does. We are going to be looking at

00:03:57.400 --> 00:04:00.120
the structural engineering of social media algorithms.

00:04:00.240 --> 00:04:02.680
And that inevitably means looking at the real

00:04:02.680 --> 00:04:04.699
world events those algorithms have influenced.

00:04:05.099 --> 00:04:06.979
Right, you can't separate the two. The research

00:04:06.979 --> 00:04:09.819
we are analyzing specifically examines the data

00:04:09.819 --> 00:04:11.960
and corporate systems surrounding polarizing

00:04:11.960 --> 00:04:14.400
elements, including the U .S. Capitol riot on

00:04:14.400 --> 00:04:17.279
January 6th. Which is obviously a highly charged

00:04:17.279 --> 00:04:20.699
topic. Very much so. But looking strictly at

00:04:20.699 --> 00:04:23.139
the data from a systems analysis perspective,

00:04:23.399 --> 00:04:26.259
regardless of where anyone falls on the political

00:04:26.259 --> 00:04:29.500
spectrum, our goal today is to impartially report

00:04:29.500 --> 00:04:32.000
on the mechanics of these algorithms. The corporate

00:04:32.000 --> 00:04:34.800
incentives driving them. Exactly. We are here

00:04:34.800 --> 00:04:37.250
to convey the framework. presented in the research

00:04:37.250 --> 00:04:40.189
not to litigate the politics of the events themselves.

00:04:40.529 --> 00:04:42.629
And looking at it purely as a systems failure

00:04:42.629 --> 00:04:45.290
is exactly how the Mohists would have approached

00:04:45.290 --> 00:04:47.790
it anyway. Really? Oh yeah, they were obsessed

00:04:47.790 --> 00:04:50.350
with structural order and the practical mechanics

00:04:50.350 --> 00:04:53.610
of societal benefit. To understand their framework,

00:04:53.829 --> 00:04:56.129
we really have to start with the linguistic mystery

00:04:56.129 --> 00:04:58.970
at the very center of this research, the concept

00:04:58.970 --> 00:05:01.459
of sinchen. This is where the detective work

00:05:01.459 --> 00:05:03.399
of the research is just so fascinating to me.

00:05:03.480 --> 00:05:06.540
The term sentient is a complete anomaly in ancient

00:05:06.540 --> 00:05:09.379
literature. It really is. In the entire pre -Kin

00:05:09.379 --> 00:05:12.000
corpus of ancient Chinese texts, which by the

00:05:12.000 --> 00:05:14.579
way is a massive volume of historical and philosophical

00:05:14.579 --> 00:05:18.019
writings, this specific two -character term appears

00:05:18.019 --> 00:05:21.629
exactly one time. Just once. Yeah. It's located

00:05:21.629 --> 00:05:24.209
in the Mozi, which is the foundational text of

00:05:24.209 --> 00:05:27.589
Mohism, specifically in a chapter titled Exalting

00:05:27.589 --> 00:05:30.990
Unity 3. And when a word is a hapax logomenon.

00:05:31.069 --> 00:05:34.379
Wait, a what? A hapax logomenon. It's just a

00:05:34.379 --> 00:05:37.019
fancy academic term for a word that occurs only

00:05:37.019 --> 00:05:40.319
once within a context. When you have one of those,

00:05:40.519 --> 00:05:43.339
it becomes an absolute nightmare for translators

00:05:43.339 --> 00:05:45.879
and scholars. You don't have a luxury of cross

00:05:45.879 --> 00:05:48.220
-referencing it against ten other chapters to

00:05:48.220 --> 00:05:51.199
deduce its exact nuance through context clues.

00:05:51.439 --> 00:05:53.860
You're just flying blind. Exactly. You are flying

00:05:53.860 --> 00:05:57.079
blind. Because of this, centuries of incredibly

00:05:57.079 --> 00:06:00.040
smart scholars have struggled to pin it down,

00:06:00.120 --> 00:06:02.459
and they often end up projecting their own cultural

00:06:02.459 --> 00:06:04.560
or philosophical assumptions onto it. Like what?

00:06:04.660 --> 00:06:06.579
What were they guessing? Well, traditionally,

00:06:06.860 --> 00:06:10.100
many translated sentient as integrity. They interpreted

00:06:10.100 --> 00:06:12.300
it as a call for a leader to make their internal

00:06:12.300 --> 00:06:14.959
words match their external actions. Okay, that

00:06:14.959 --> 00:06:17.240
sounds nice, but what else? Others read it as

00:06:17.240 --> 00:06:19.670
self -confidence. And if you look at influential

00:06:19.670 --> 00:06:22.189
English translators like Ian Johnston or Mei

00:06:22.189 --> 00:06:24.170
-Yi Pao, they looked at the broader political

00:06:24.170 --> 00:06:26.930
context of the chapter and translated it loosely

00:06:26.930 --> 00:06:29.589
as trust in the people or trusting their staff.

00:06:30.009 --> 00:06:32.290
I mean, I'm really struggling to see how self

00:06:32.290 --> 00:06:34.529
-confidence or trusting your staff could both

00:06:34.529 --> 00:06:36.930
be derived from the exact same characters. It

00:06:36.930 --> 00:06:38.730
feels like scholars were kind of just throwing

00:06:38.730 --> 00:06:41.670
darts at a board. They absolutely were. But the

00:06:41.670 --> 00:06:44.310
research points out a massive, undeniable logical

00:06:44.310 --> 00:06:47.889
flaw in translating it as self -confidence. Okay.

00:06:47.930 --> 00:06:50.410
What's the flaw? If you actually read the surrounding

00:06:50.410 --> 00:06:53.509
text in the modzi, the passage is explicitly

00:06:53.509 --> 00:06:55.970
talking about the fundamental limitations of

00:06:55.970 --> 00:06:58.370
the individual human body. Like physical limits.

00:06:58.649 --> 00:07:01.310
Yeah. It says essentially that seeing with one

00:07:01.310 --> 00:07:03.629
eye is not nearly as effective as seeing with

00:07:03.629 --> 00:07:06.129
two, and hearing with one ear is not as effective

00:07:06.129 --> 00:07:07.949
as hearing with two. Right. It's establishing

00:07:07.949 --> 00:07:11.329
a baseline of human biological limits. Precisely.

00:07:11.550 --> 00:07:14.069
The ancient sage kings couldn't personally see

00:07:14.069 --> 00:07:16.490
or hear things happening a thousand lies away,

00:07:16.709 --> 00:07:20.250
a lie being an ancient unit of distance. Because

00:07:20.250 --> 00:07:22.509
of this physical limitation, they were forced

00:07:22.509 --> 00:07:25.259
to rely on others. So if the entire premise of

00:07:25.259 --> 00:07:27.800
the chapter is about the inescapable need to

00:07:27.800 --> 00:07:30.019
outsource your vision and hearing because your

00:07:30.019 --> 00:07:32.839
own body is insufficient? Then translating the

00:07:32.839 --> 00:07:35.439
core concept as self -confidence contradicts

00:07:35.439 --> 00:07:37.620
the entire argument. Exactly. That makes zero

00:07:37.620 --> 00:07:40.139
sense. What's fascinating here is how the analysis

00:07:40.139 --> 00:07:43.439
solves this linguistic puzzle by breaking down

00:07:43.439 --> 00:07:45.759
the etymology and usage of the two individual

00:07:45.759 --> 00:07:48.939
characters during that specific historical period.

00:07:49.160 --> 00:07:51.600
Let's take the second character first, Zhenzhen.

00:07:51.740 --> 00:07:55.060
Okay, Shen. In many contexts, shen can mean self

00:07:55.060 --> 00:07:58.240
or person in a vague, abstract way. But if you

00:07:58.240 --> 00:08:00.579
analyze contemporary texts from the same era,

00:08:00.680 --> 00:08:03.620
like the Zhuangzi, the usage of shen very clearly

00:08:03.620 --> 00:08:06.100
and consistently refers to the physical, biological,

00:08:06.220 --> 00:08:09.180
flesh and blood body. It is tangible. Yes. It

00:08:09.180 --> 00:08:11.740
is the limbs, the eyes, the ears. So shen isn't

00:08:11.740 --> 00:08:14.639
some metaphor for the soul or the ego. It's the

00:08:14.639 --> 00:08:17.740
actual meat and bones of a human being. Precisely

00:08:17.740 --> 00:08:20.139
the opposite of a metaphor. And then we look

00:08:20.139 --> 00:08:23.459
at the first character, sin. In modern usage,

00:08:23.699 --> 00:08:26.639
and even in many classical contexts, sinsin means

00:08:26.639 --> 00:08:30.699
belief or trust. But the research digs into ancient

00:08:30.699 --> 00:08:33.740
divinatory texts, specifically the Book of Changes,

00:08:33.899 --> 00:08:36.600
where sinsin operates with a very different mechanical

00:08:36.600 --> 00:08:39.500
meaning. And what's that? In those older contexts,

00:08:39.899 --> 00:08:42.559
sins means stretching or extending. It implies

00:08:42.559 --> 00:08:44.940
a physical elongation or projection outward.

00:08:45.200 --> 00:08:47.870
Okay, wow. When you snap those two revised definitions

00:08:47.870 --> 00:08:50.590
together, the entire landscape of the passage

00:08:50.590 --> 00:08:53.649
shifts. Completely. Sinton doesn't mean a vague

00:08:53.649 --> 00:08:56.330
internal sense of having integrity, and it doesn't

00:08:56.330 --> 00:08:58.269
just mean believing your employees are doing

00:08:58.269 --> 00:09:01.169
a good job. It literally translates to the extension

00:09:01.169 --> 00:09:03.429
of the physical body. Exactly. It fundamentally

00:09:03.429 --> 00:09:05.850
changes the relationship from a managerial one

00:09:05.850 --> 00:09:08.730
to an almost biological one. The Sage Kings weren't

00:09:08.730 --> 00:09:10.870
just delegating tasks to independent contractors.

00:09:11.190 --> 00:09:12.950
They were extending their sensory apparatus.

00:09:13.230 --> 00:09:15.639
Yes, their very capacity to perceive. reality

00:09:15.639 --> 00:09:18.779
across the vastness of the kingdom through these

00:09:18.779 --> 00:09:20.919
trusted human agents. I want to make sure I'm

00:09:20.919 --> 00:09:22.580
fully grasping the weight of this. It's like

00:09:22.580 --> 00:09:25.399
a CEO of a massive multinational corporation

00:09:25.399 --> 00:09:27.659
who obviously can't be in two places at once.

00:09:27.740 --> 00:09:31.259
Right. So they send a deeply trusted vice president

00:09:31.259 --> 00:09:34.279
to a vital make or break negotiation on the other

00:09:34.279 --> 00:09:37.379
side of the world. But the Sinjin framework implies

00:09:37.379 --> 00:09:39.840
that this VP isn't just a messenger carrying

00:09:39.840 --> 00:09:42.899
a list of demands. No, not at all. The VP is

00:09:42.899 --> 00:09:46.179
literally operating as the CEO's own eyes, ears

00:09:46.179 --> 00:09:50.120
and hands in that boardroom. If the VP misinterprets

00:09:50.120 --> 00:09:53.500
the tone of the room, the CEO effectively misinterprets

00:09:53.500 --> 00:09:56.539
it. If the VP acts unethically to close the deal,

00:09:56.659 --> 00:09:59.460
the CEO is the one acting unethically. There

00:09:59.460 --> 00:10:02.330
is no separation of agency. That is the exact

00:10:02.330 --> 00:10:05.370
dynamic, and it highlights why the standard concept

00:10:05.370 --> 00:10:08.330
of trust is just too weak a word for this. Cention

00:10:08.330 --> 00:10:10.870
requires a level of reliance that goes far beyond

00:10:10.870 --> 00:10:13.690
a standard employment contract or some blind

00:10:13.690 --> 00:10:16.509
leap of faith. It's deeper. It requires an absolute,

00:10:16.950 --> 00:10:20.710
ethically conditioned symbiosis. The agent must

00:10:20.710 --> 00:10:23.070
transmit the reality of the distant situation

00:10:23.070 --> 00:10:26.149
perfectly back to the king, without distortion,

00:10:26.590 --> 00:10:30.080
bias, or self -serving filters. Because if the

00:10:30.080 --> 00:10:32.419
transmission is flawed, the king's subsequent

00:10:32.419 --> 00:10:35.500
actions based on that flawed reality will inevitably

00:10:35.500 --> 00:10:38.740
be disastrous. So think about this in the context

00:10:38.740 --> 00:10:41.220
of your own daily life and your own modern extensions.

00:10:41.700 --> 00:10:44.220
When you pull your smartphone out of your pocket

00:10:44.220 --> 00:10:46.980
and type a question into a generative AI, you

00:10:46.980 --> 00:10:49.559
are engaging in sentient. You really are. You

00:10:49.559 --> 00:10:51.899
are throwing your cognitive reach out across

00:10:51.899 --> 00:10:55.299
the vast expanse of the Internet to gather, synthesize

00:10:55.299 --> 00:10:57.700
and perceive information that you couldn't possibly

00:10:57.700 --> 00:11:00.960
process on your own. you are extending your mind.

00:11:01.159 --> 00:11:03.299
But the deeply uncomfortable question this ancient

00:11:03.299 --> 00:11:06.559
framework forces us to ask is, are you extending

00:11:06.559 --> 00:11:08.840
your reach through a genuinely trustworthy agent?

00:11:09.139 --> 00:11:11.820
Right. Does the tool transmit reality perfectly,

00:11:12.000 --> 00:11:14.159
or does it distort it before it reaches your

00:11:14.159 --> 00:11:16.379
eyes? Which brings us perfectly to the second,

00:11:16.399 --> 00:11:18.820
much more demanding layer of this MOHIS blueprint.

00:11:19.139 --> 00:11:21.419
Because sentient isn't just about the mechanical

00:11:21.419 --> 00:11:23.659
accuracy of getting raw data from point A to

00:11:23.659 --> 00:11:26.240
point B. It's a dual structure. Yes. The first

00:11:26.240 --> 00:11:28.960
part, as we've established, is delegating perception

00:11:28.960 --> 00:11:32.039
and action. But the second part is what truly

00:11:32.039 --> 00:11:35.659
sets Mohism apart. It is the mandatory extension

00:11:35.659 --> 00:11:39.139
of one's moral virtues through those exact same

00:11:39.139 --> 00:11:41.840
trusted agents. Let's dig into the mechanics

00:11:41.840 --> 00:11:45.000
of that moral extension. The Mosey doesn't just

00:11:45.000 --> 00:11:47.860
say, find someone you like and trust them. It

00:11:47.860 --> 00:11:50.919
outlines a highly specific, rigorous methodology

00:11:50.919 --> 00:11:53.759
for how these agents were chosen. A concept known

00:11:53.759 --> 00:11:56.379
as exalting worthiness. Right. Because in the

00:11:56.379 --> 00:11:58.820
ancient world, power was almost entirely hereditary.

00:11:58.960 --> 00:12:01.279
You picked your cousins, your wealthy allies,

00:12:01.580 --> 00:12:03.480
or the aristocrats who just looked the part.

00:12:03.639 --> 00:12:06.080
And Mosey completely rejected that. He demanded

00:12:06.080 --> 00:12:08.960
that aides be chosen based entirely on demonstrated

00:12:08.960 --> 00:12:12.340
virtue and an unwavering commitment to the equitable

00:12:12.340 --> 00:12:14.899
distribution of resources. So Mosey was, what,

00:12:14.940 --> 00:12:17.000
a radical egalitarian for his time? Oh, absolutely.

00:12:17.519 --> 00:12:19.679
He lived during the Warring States period, an

00:12:19.679 --> 00:12:22.279
era of horrific violence, chaos, and aristocratic

00:12:22.279 --> 00:12:25.460
corruption. He saw firsthand that nepotism and

00:12:25.460 --> 00:12:27.860
hereditary power were literally destroying society.

00:12:28.139 --> 00:12:30.240
Yeah, that makes sense. The text is incredibly

00:12:30.240 --> 00:12:33.059
explicit on this point. It states that heaven

00:12:33.059 --> 00:12:35.139
does not discriminate between the rich and the

00:12:35.139 --> 00:12:38.200
poor, the noble and the base. The only metric

00:12:38.200 --> 00:12:41.250
for promotion is worthiness. And the Mosey doesn't

00:12:41.250 --> 00:12:44.029
leave worthiness as some vague spiritual ideal.

00:12:44.289 --> 00:12:47.309
Right. Not at all. It defines the way of worthiness

00:12:47.309 --> 00:12:51.049
as a relentless, active practice. It's a person

00:12:51.049 --> 00:12:53.330
who uses their physical strength to assist others,

00:12:53.490 --> 00:12:56.090
who uses their accumulated wealth to distribute

00:12:56.090 --> 00:12:58.789
to the needy, and who uses their knowledge to

00:12:58.789 --> 00:13:01.440
actively teach and elevate the community. It

00:13:01.440 --> 00:13:03.840
is a strictly applied commitment to collective

00:13:03.840 --> 00:13:06.480
welfare. Exactly. Now, this brings up a massive

00:13:06.480 --> 00:13:09.120
century -old philosophical debate within the

00:13:09.120 --> 00:13:11.059
academic literature that the research addresses.

00:13:11.179 --> 00:13:13.860
Because when modern ears hear phrases like collective

00:13:13.860 --> 00:13:15.899
welfare, practical mechanics, and distributing

00:13:15.899 --> 00:13:18.720
resources, it sounds incredibly utilitarian.

00:13:18.960 --> 00:13:21.919
Very Bentham. Exactly. Was Mosey just an ancient

00:13:21.919 --> 00:13:24.820
Chinese version of Jeremy Bentham? The research

00:13:24.820 --> 00:13:27.799
notes that back in the 1800s and early 1900s,

00:13:27.799 --> 00:13:30.399
early Western scholars and translators like Joseph

00:13:30.399 --> 00:13:33.399
A. Edkins and Daisetsu Suzuki completely framed

00:13:33.399 --> 00:13:36.039
Mosey through this Western lens. They labeled

00:13:36.039 --> 00:13:39.299
him a thorough utilitarian. Right. They pointed

00:13:39.299 --> 00:13:41.919
to the fact that Mosey vehemently hated what

00:13:41.919 --> 00:13:45.120
he called sentimental extravagances. For example,

00:13:45.159 --> 00:13:47.320
he wrote extensively against the practice of

00:13:47.320 --> 00:13:51.019
lavish, ruinously expensive funerals, which were

00:13:51.019 --> 00:13:53.700
culturally very important in ancient China. Because

00:13:53.700 --> 00:13:55.600
they drained the national wealth and provided

00:13:55.600 --> 00:13:58.259
zero practical benefit to the living. Right.

00:13:58.320 --> 00:14:00.379
So it's a classic case of historical projection.

00:14:00.759 --> 00:14:03.139
Early Western translators were operating during

00:14:03.139 --> 00:14:05.899
the height of the Industrial Revolution and Victorian

00:14:05.899 --> 00:14:09.299
utilitarian thought. They read Mosey's texts,

00:14:09.480 --> 00:14:12.860
saw his intense focus on concrete societal benefits,

00:14:13.100 --> 00:14:16.299
what Mosey called Lero, and immediately assumed

00:14:16.299 --> 00:14:19.039
he was performing a cold mathematical calculation

00:14:19.039 --> 00:14:22.000
of the greatest good for the greatest number.

00:14:22.460 --> 00:14:25.360
Suzuki even went so far as to contrast Mosey

00:14:25.360 --> 00:14:28.240
with Western Christianity, arguing that Mosey's

00:14:28.240 --> 00:14:30.500
famous doctrine of universal love wasn't about

00:14:30.500 --> 00:14:33.080
genuine internal empathy or human connection.

00:14:33.440 --> 00:14:36.259
He claimed it was merely a dry, practical mechanism

00:14:36.259 --> 00:14:39.159
engineered for political utility. They basically

00:14:39.159 --> 00:14:41.120
stripped the soul out of the philosophy to make

00:14:41.120 --> 00:14:43.639
it fit a Benthamite spreadsheet. But the research

00:14:43.639 --> 00:14:45.899
highlights that modern scholarship has completely

00:14:45.899 --> 00:14:48.970
dismantled that interpretation. Scholars like

00:14:48.970 --> 00:14:51.429
Dennis Ahern, Brian Van Norden, and Chris Frazier

00:14:51.429 --> 00:14:53.909
have reexamined the texts and concluded that

00:14:53.909 --> 00:14:56.950
Mosey was definitively not a standard utilitarian.

00:14:57.269 --> 00:14:58.970
Let's articulate the difference because it seems

00:14:58.970 --> 00:15:02.250
crucial. Classical Western utilitarianism, like

00:15:02.250 --> 00:15:06.190
Bentham's, is deeply tied to hedonism, the maximization

00:15:06.190 --> 00:15:08.830
of subjective pleasure and the minimization of

00:15:08.830 --> 00:15:11.370
pain. It's about whatever makes the most people

00:15:11.370 --> 00:15:14.139
feel the happiest. Right. But Mosey didn't care

00:15:14.139 --> 00:15:16.539
about subjective emotional states or maximizing

00:15:16.539 --> 00:15:19.259
pleasure. Exactly. Mosey's metrics were fiercely

00:15:19.259 --> 00:15:21.980
objective and observable. He cared about three

00:15:21.980 --> 00:15:25.080
primary goods. The increase of societal wealth,

00:15:25.259 --> 00:15:27.159
the growth of the population, and the maintenance

00:15:27.159 --> 00:15:28.960
of public order. He was trying to keep people

00:15:28.960 --> 00:15:30.620
from starving and being slaughtered in endless

00:15:30.620 --> 00:15:33.840
wars. But, and this is the absolute crux of Mohist

00:15:33.840 --> 00:15:36.399
ethics, that separates it entirely from cold

00:15:36.399 --> 00:15:39.730
calculation. Modern scholars like Fraser emphasize

00:15:39.730 --> 00:15:42.690
that for Mosey, these tangible benefits cannot,

00:15:42.850 --> 00:15:45.610
under any circumstances, exist in a moral vacuum.

00:15:46.090 --> 00:15:49.789
Benefit, or li, is fundamentally and inextricably

00:15:49.789 --> 00:15:52.769
bound to righteousness, or yi. Here's where it

00:15:52.769 --> 00:15:54.730
gets really interesting. I want to read a specific

00:15:54.730 --> 00:15:56.830
phrasing from the Mosey canons because it just

00:15:56.830 --> 00:16:01.070
blew my mind. The text literally states, righteousness

00:16:01.070 --> 00:16:04.970
is being of benefit. Wow. They aren't just related

00:16:04.970 --> 00:16:07.529
concepts. Right. They are mathematically fundamentally

00:16:07.529 --> 00:16:11.190
linked. If a ruler is pursuing wealth or even

00:16:11.190 --> 00:16:13.730
public order, but they are achieving it without

00:16:13.730 --> 00:16:16.570
moral righteousness, without an impartial universal

00:16:16.570 --> 00:16:19.710
love for all people, then what they have achieved

00:16:19.710 --> 00:16:22.090
isn't actually considered a benefit in the Moist

00:16:22.090 --> 00:16:24.129
framework. It's an illusion. It's essentially

00:16:24.129 --> 00:16:27.769
fraud. Righteousness is the inescapable bedrock

00:16:27.769 --> 00:16:30.330
that makes a benefit real. And that profound

00:16:30.330 --> 00:16:33.669
interlinking of benefit and righteousness deeply

00:16:33.669 --> 00:16:36.009
influences how we must understand the concept

00:16:36.009 --> 00:16:38.990
of sinshin. When the sage kings extended their

00:16:38.990 --> 00:16:41.009
biological and cognitive reach through their

00:16:41.009 --> 00:16:43.750
agents, they weren't just outsourcing bureaucratic

00:16:43.750 --> 00:16:45.690
labor to efficient middle managers. They were

00:16:45.690 --> 00:16:48.289
extending a moral framework. Exactly what the

00:16:48.289 --> 00:16:51.590
texts call the way of a gentleman. There is a

00:16:51.590 --> 00:16:53.950
beautiful passage in the chapter on cultivating

00:16:53.950 --> 00:16:56.809
the self that describes this outward movement.

00:16:57.330 --> 00:17:00.570
It states that virtue must originate paramount

00:17:00.570 --> 00:17:03.519
within the mind of the individual. then extend

00:17:03.519 --> 00:17:06.480
to govern the physical limbs, then radiate outward

00:17:06.480 --> 00:17:09.460
to influence one's immediate circle. And ultimately,

00:17:09.660 --> 00:17:12.279
through that demonstrated righteousness, attract

00:17:12.279 --> 00:17:14.900
good people from across the empire. It's like

00:17:14.900 --> 00:17:17.299
a moral ripple effect. The initial stone dropped

00:17:17.299 --> 00:17:20.160
in the pond is the internal virtue, and the concentric

00:17:20.160 --> 00:17:22.680
circles are the agents carrying that virtue to

00:17:22.680 --> 00:17:25.240
the edges of the world. That is a perfect visualization.

00:17:25.799 --> 00:17:28.420
The trusted agent acts as a moral amplifier.

00:17:29.079 --> 00:17:31.400
They aren't merely a neutral wire transmitting

00:17:31.400 --> 00:17:34.000
facts back to the king. They are actively executing

00:17:34.000 --> 00:17:36.019
the king's righteousness in the physical world.

00:17:36.059 --> 00:17:38.529
Right. And when you deeply analyze this structure,

00:17:38.730 --> 00:17:41.230
it is striking how this ancient framework resonates

00:17:41.230 --> 00:17:44.849
far beyond the narrow field of AI ethics. In

00:17:44.849 --> 00:17:47.089
the newly added focus of Pang's research, it

00:17:47.089 --> 00:17:48.950
actually touches the very architecture of how

00:17:48.950 --> 00:17:51.150
human communities have always tried to build

00:17:51.150 --> 00:17:53.210
trust across deep divides. What do you mean by

00:17:53.210 --> 00:17:55.970
deep divides? Well, Pang's recovery of sentient

00:17:55.970 --> 00:17:58.630
precisely mirrors the structural mechanisms we

00:17:58.630 --> 00:18:02.069
see in the theology of interfaith ethics. Wait,

00:18:02.190 --> 00:18:04.910
interfaith ethics. How does an ancient Chinese

00:18:04.910 --> 00:18:08.109
concept about delegating bureaucratic power connect

00:18:08.109 --> 00:18:11.289
to modern religious dialogue? Let me push back

00:18:11.289 --> 00:18:13.210
on that because those seem like entirely different

00:18:13.210 --> 00:18:15.849
domains. It connects through the concept of the

00:18:15.849 --> 00:18:18.269
covenant. The covenant. OK. Think about how different

00:18:18.269 --> 00:18:21.269
religious traditions structure ultimate trust.

00:18:21.529 --> 00:18:24.390
Think about the mythic, the Medinan covenant

00:18:24.390 --> 00:18:27.130
in the Islamic tradition or the concept of Brit

00:18:27.130 --> 00:18:29.650
in the Hebrew tradition. OK, I'm tracking. A

00:18:29.650 --> 00:18:31.710
covenant is fundamentally different from a contract.

00:18:32.250 --> 00:18:35.009
A contract is transactional and assumes a lack

00:18:35.009 --> 00:18:37.809
of trust. It's a mechanism to enforce compliance.

00:18:38.190 --> 00:18:40.289
Right. A contract is basically planning for when

00:18:40.289 --> 00:18:42.829
things go wrong. Exactly. Yeah. A covenant, however,

00:18:43.009 --> 00:18:45.650
is ethically conditioned reliance. It is the

00:18:45.650 --> 00:18:48.609
extension of trust through shared binding moral

00:18:48.609 --> 00:18:51.430
obligations aimed at the common good. So it's

00:18:51.430 --> 00:18:53.589
not just blind faith and the other party. No.

00:18:53.670 --> 00:18:56.769
In deep interfaith work, genuine pluralism isn't

00:18:56.769 --> 00:18:59.309
about erasing differences or holding hands and

00:18:59.309 --> 00:19:01.470
pretending we all believe the exact same theology.

00:19:02.000 --> 00:19:03.799
Right. That's just watering everything down.

00:19:03.940 --> 00:19:07.779
Exactly. True covenantal pluralism is about extending

00:19:07.779 --> 00:19:10.880
your moral commitments across those very real

00:19:10.880 --> 00:19:13.400
lines of difference. You choose to rely on the

00:19:13.400 --> 00:19:16.500
other because they have demonstrated a rigorous

00:19:16.500 --> 00:19:18.759
commitment to mutual welfare and righteousness,

00:19:19.099 --> 00:19:21.720
even if their specific theology differs from

00:19:21.720 --> 00:19:24.759
yours. Oh, I see the parallel now. So in a covenantal

00:19:24.759 --> 00:19:26.880
relationship, just like in the Moist concept

00:19:26.880 --> 00:19:30.680
of sincerntion, trust is not passive. It's highly

00:19:30.680 --> 00:19:33.539
active. You are actively selecting partners or

00:19:33.539 --> 00:19:36.779
agents based on demonstrated virtue, and the

00:19:36.779 --> 00:19:39.059
mutual benefit you seek to build together cannot

00:19:39.059 --> 00:19:41.019
be divorced from a shared sense of righteousness.

00:19:41.500 --> 00:19:43.980
You can't have a successful interfaith covenant

00:19:43.980 --> 00:19:46.700
if one party is secretly trying to exploit the

00:19:46.700 --> 00:19:49.200
other for their own exclusive benefit. Precisely.

00:19:49.279 --> 00:19:51.779
The extension of trust is communal and highly

00:19:51.779 --> 00:19:54.170
conditioned on ethical behavior. It starts from

00:19:54.170 --> 00:19:56.789
within, radiates out, and builds a resilient

00:19:56.789 --> 00:19:59.609
network of reliance. So sentient is, in essence,

00:19:59.869 --> 00:20:02.309
an ancient blueprint for covenantal pluralism.

00:20:02.470 --> 00:20:06.170
Yes. It demands that our extensions, whether

00:20:06.170 --> 00:20:08.450
they are human ambassadors participating in an

00:20:08.450 --> 00:20:11.559
interfaith dialogue or Bringing it back to our

00:20:11.559 --> 00:20:14.160
main topic, digital platforms mediating our daily

00:20:14.160 --> 00:20:17.759
communications must aggressively uphold that

00:20:17.759 --> 00:20:20.119
high standard of moral transmission. Well, keeping

00:20:20.119 --> 00:20:22.579
that incredibly high standard of covenantal moral

00:20:22.579 --> 00:20:24.859
transmission in mind, let's take this ancient

00:20:24.859 --> 00:20:27.200
blueprint and slam it headfirst into the 21st

00:20:27.200 --> 00:20:29.059
century. Let's do it. Let's look at artificial

00:20:29.059 --> 00:20:32.119
intelligence as our modern extension. And I'll

00:20:32.119 --> 00:20:34.539
give a spoiler alert to anyone listening. According

00:20:34.539 --> 00:20:37.799
to this research, current AI systems are a catastrophic

00:20:37.799 --> 00:20:54.680
failure on the very first most basic And the

00:20:54.680 --> 00:20:59.420
verdict from the system's analysis is a resounding

00:20:59.420 --> 00:21:02.299
structural no. Right. The research provides some

00:21:02.299 --> 00:21:05.440
deeply unsettling examples of this failure, specifically

00:21:05.440 --> 00:21:08.579
regarding embedded bias. Let's look at the mechanics

00:21:08.579 --> 00:21:11.160
of Google's word embeddings. For those who aren't

00:21:11.160 --> 00:21:13.799
software engineers, word embeddings are the underlying

00:21:13.799 --> 00:21:16.980
mathematical architecture that helps an AI understand

00:21:16.980 --> 00:21:21.160
language. It maps words as coordinates in a multidimensional

00:21:21.160 --> 00:21:23.960
space to understand how they relate to each other.

00:21:24.160 --> 00:21:27.059
Exactly. And when researchers probed these mathematical

00:21:27.059 --> 00:21:30.900
spaces, they found that the AI literally mathematically

00:21:30.900 --> 00:21:35.019
equated the vector for man to the vector for

00:21:35.019 --> 00:21:37.609
computer programmer. And mapped the vector for

00:21:37.609 --> 00:21:40.329
woman directly to homemaker. Yes. And it gets

00:21:40.329 --> 00:21:42.609
significantly darker. They looked at toxicity

00:21:42.609 --> 00:21:45.230
models. Which are the AI specifically designed

00:21:45.230 --> 00:21:47.650
to flag hate speech and protect users. Right.

00:21:47.710 --> 00:21:49.750
These models were automatically mathematically

00:21:49.750 --> 00:21:52.529
flagging words like gay or Muslim as inherently

00:21:52.529 --> 00:21:55.410
toxic, while the word heterosexual was given

00:21:55.410 --> 00:21:58.309
a free pass. To understand the depth of this

00:21:58.309 --> 00:22:00.450
failure, we have to understand why these systems

00:22:00.450 --> 00:22:03.740
behave this way. As AI researcher Meredith Broussard

00:22:03.740 --> 00:22:06.259
points out in the analysis, this is not a glitch.

00:22:06.460 --> 00:22:09.500
It's not a bug. Right. This is not a simple typographical

00:22:09.500 --> 00:22:11.940
error in the code that a programmer can easily

00:22:11.940 --> 00:22:15.019
patch with an update. These AI models are trained

00:22:15.019 --> 00:22:18.880
by scraping vast oceans of human data, usually

00:22:18.880 --> 00:22:21.460
the entirely unfiltered Internet. They absorb

00:22:21.460 --> 00:22:24.259
all of it. They absorb the historical prejudices,

00:22:24.339 --> 00:22:28.000
the structural injustices, the racism, the sexism,

00:22:28.019 --> 00:22:30.000
and the ableism that are deeply historically

00:22:30.000 --> 00:22:33.240
embedded in our societies. The AI operates as

00:22:33.240 --> 00:22:37.319
a perfect, uncritical mirror. It does not possess

00:22:37.319 --> 00:22:40.339
the capacity to correct our biases. It mathematically

00:22:40.339 --> 00:22:43.259
formalizes them. It bakes human prejudice into

00:22:43.259 --> 00:22:45.299
an algorithmic formula, and then it reflects

00:22:45.299 --> 00:22:47.359
that prejudice back to us at lightning speed,

00:22:47.579 --> 00:22:50.299
disguised as objective truth. And that disguise

00:22:50.299 --> 00:22:52.720
of objectivity is exactly why people trust the

00:22:52.720 --> 00:22:55.180
reflection blindly. The research identifies this

00:22:55.180 --> 00:22:57.420
psychological vulnerability as automation bias.

00:22:57.740 --> 00:23:00.599
Automation bias? Yeah. Human beings are neurologically

00:23:00.599 --> 00:23:02.599
hardwired to assume that if an answer comes from

00:23:02.599 --> 00:23:04.799
a machine, a computer, or a glowing screen, it

00:23:04.799 --> 00:23:06.779
must be objective, mathematical, and free of

00:23:06.779 --> 00:23:08.799
human error. We just drop our critical defenses.

00:23:09.359 --> 00:23:12.119
There is a terrifying real -world example cited

00:23:12.119 --> 00:23:14.720
in the text from AI researcher Timnit Debreu.

00:23:25.999 --> 00:23:45.000
Which is just... Insane. Wow. was physically

00:23:45.000 --> 00:23:47.119
put in handcuffs because the authorities trusted

00:23:47.119 --> 00:23:49.400
the tool more than reality. The psychological

00:23:49.400 --> 00:23:51.779
power of automation bias cannot be overstated.

00:23:51.880 --> 00:23:53.579
There's another study detailed in the research

00:23:53.579 --> 00:23:56.480
by Robinette that perfectly, almost comically,

00:23:56.599 --> 00:23:58.640
illustrates this trap. Okay, what happened in

00:23:58.640 --> 00:24:00.819
that one? Researchers put human subjects into

00:24:00.819 --> 00:24:03.519
a simulated emergency fire evacuation scenario.

00:24:03.839 --> 00:24:06.380
They introduced a physical robot whose designated

00:24:06.380 --> 00:24:09.339
job was to guide the humans to safety. But before

00:24:09.339 --> 00:24:12.400
the fire simulation even started, the researchers

00:24:12.400 --> 00:24:16.029
made the participants watched the robot visibly,

00:24:16.069 --> 00:24:19.289
undeniably malfunction. It drove in circles.

00:24:19.430 --> 00:24:21.930
It was clearly broken. Oh, no. However, when

00:24:21.930 --> 00:24:23.990
the simulated emergency alarms began to ring

00:24:23.990 --> 00:24:26.390
and the smoke filled the room, the human participants

00:24:26.390 --> 00:24:28.970
literally followed that broken robot into a dark,

00:24:28.990 --> 00:24:31.710
entirely blocked room. Even possessing prior

00:24:31.710 --> 00:24:35.289
visual evidence that the machine was deeply flawed.

00:24:35.549 --> 00:24:38.369
Yes, their default cognitive instinct under pressure

00:24:38.369 --> 00:24:41.390
was to obey the algorithmic authority. That is

00:24:41.390 --> 00:24:44.569
deeply unsettling on a species level. Why are

00:24:44.569 --> 00:24:46.910
these highly advanced systems failing the basic

00:24:46.910 --> 00:24:49.089
transmission test so completely? Why do they

00:24:49.089 --> 00:24:51.549
hallucinate? The research brings in systems analysts

00:24:51.549 --> 00:24:54.170
Ding and Lai. To explain the underlying mechanics

00:24:54.170 --> 00:24:56.809
of large language models and generative AI. Yeah.

00:24:56.890 --> 00:24:59.230
They point out that these systems operate purely

00:24:59.230 --> 00:25:01.490
by a mathematical process called interpolation.

00:25:01.970 --> 00:25:05.069
Interpolation. Right. And LLM looks at the billions

00:25:05.069 --> 00:25:07.490
of words it was trained on and calculates the

00:25:07.490 --> 00:25:09.430
highest statistical probability for what the

00:25:09.430 --> 00:25:11.970
next word in a sentence should be. It is a prediction

00:25:11.970 --> 00:25:14.490
engine. It does not extrapolate. No. It does

00:25:14.490 --> 00:25:17.750
not possess human curiosity. It has no sudden

00:25:17.750 --> 00:25:20.150
flashes of inspiration, and it fundamentally

00:25:20.150 --> 00:25:23.029
lacks imagination or a true understanding of

00:25:23.029 --> 00:25:25.509
the physical world. It is essentially simulating

00:25:25.509 --> 00:25:28.720
comprehension through advanced statistics. Because

00:25:28.720 --> 00:25:31.380
it is operating entirely on probabilities rather

00:25:31.380 --> 00:25:34.599
than grounded factual verification, it frequently

00:25:34.599 --> 00:25:37.579
experiences what engineers call hallucinations.

00:25:37.619 --> 00:25:40.960
It will confidently, eloquently present an absolute

00:25:40.960 --> 00:25:43.759
falsehood. And it will even invent fake academic

00:25:43.759 --> 00:25:46.500
psychations to support that falsehood, placing

00:25:46.500 --> 00:25:49.259
them right alongside verifiable facts. And because

00:25:49.259 --> 00:25:52.220
the AI mimics natural human language so fluently,

00:25:52.440 --> 00:25:55.420
leveraging our automation bias, it becomes incredibly

00:25:55.420 --> 00:25:57.710
difficult for a standard user to detect. the

00:25:57.710 --> 00:26:00.049
lie without engaging in rigorous independent

00:26:00.049 --> 00:26:02.930
verification. Which defeats the entire purpose

00:26:02.930 --> 00:26:05.069
of delegating the task in the first place. Exactly.

00:26:05.089 --> 00:26:07.890
So applying this to our daily lives, when you

00:26:07.890 --> 00:26:10.990
use a digital GPS map that confidently instructs

00:26:10.990 --> 00:26:13.519
you to drive your car straight into a lake. Which

00:26:13.519 --> 00:26:15.700
has documented cases of happening, by the way.

00:26:15.779 --> 00:26:17.700
Oh, absolutely. Or when you scroll through a

00:26:17.700 --> 00:26:20.359
social media feed that is algorithmically designed

00:26:20.359 --> 00:26:23.220
to systematically filter out any opposing viewpoints,

00:26:23.400 --> 00:26:25.440
keeping you artificially comfortable in what

00:26:25.440 --> 00:26:27.920
the research calls an information cocoon. An

00:26:27.920 --> 00:26:30.099
information cocoon, yes. Your extended senses

00:26:30.099 --> 00:26:33.200
are actively lying to you. You are trusting an

00:26:33.200 --> 00:26:35.599
agent that is structurally, mathematically incapable

00:26:35.599 --> 00:26:38.960
of passing the first Mohis test of reliable delegation.

00:26:39.599 --> 00:26:42.279
You simply cannot trust the data the agent is

00:26:42.279 --> 00:26:44.779
sending back. And if the technology is feeling

00:26:44.779 --> 00:26:47.460
that first, most basic test of functional reliability,

00:26:47.960 --> 00:26:50.779
the implications become exponentially more dire

00:26:50.779 --> 00:26:53.059
when we move to the second test of sentience.

00:26:53.180 --> 00:26:55.240
The moral character test. Yes, the moral character

00:26:55.240 --> 00:26:57.599
of the entity you are extending yourself through.

00:26:57.940 --> 00:27:01.000
Because AI is not a naturally occurring phenomenon.

00:27:01.660 --> 00:27:04.579
It isn't a floating, neutral math equation existing

00:27:04.579 --> 00:27:07.859
in a vacuum. It is designed, built, owned, and

00:27:07.859 --> 00:27:10.539
deployed by massive publicly traded corporate

00:27:10.539 --> 00:27:13.039
entities. Which brings us to the second catastrophic

00:27:13.039 --> 00:27:16.039
failure of modern AI, the corporate disconnect

00:27:16.039 --> 00:27:19.359
between profit maximization and societal righteousness.

00:27:19.839 --> 00:27:22.359
The second level of sentient mandates that your

00:27:22.359 --> 00:27:25.420
trusted agent must be an active amplifier of

00:27:25.420 --> 00:27:28.650
moral virtue and collective welfare. But the

00:27:28.650 --> 00:27:31.470
tech behemoth companies like Meta, Amazon, and

00:27:31.470 --> 00:27:34.839
Google operate on a foundational economic logic

00:27:34.839 --> 00:27:38.440
that intrinsically and legally prioritizes shareholder

00:27:38.440 --> 00:27:41.319
profit above all else. Often in direct opposition

00:27:41.319 --> 00:27:43.539
to public welfare. The research draws heavily

00:27:43.539 --> 00:27:45.940
on the structural critiques of scholars like

00:27:45.940 --> 00:27:48.440
Benjamin Kuypers and Mark Kokelberg to illustrate

00:27:48.440 --> 00:27:51.700
this. Kokelberg utilizes a brutal but highly

00:27:51.700 --> 00:27:54.819
accurate metaphor to describe the modern digital

00:27:54.819 --> 00:27:57.440
economy. He argues that the vast majority of

00:27:57.440 --> 00:28:00.180
users on these platforms operate under the illusion

00:28:00.180 --> 00:28:03.000
that they are receiving free services, free search

00:28:03.000 --> 00:28:05.539
engines, free email, free social networking.

00:28:05.859 --> 00:28:08.119
But in structural reality, human beings have

00:28:08.119 --> 00:28:10.599
been downgraded and turned into smartphone cattle

00:28:10.599 --> 00:28:12.940
milked for our data. Smartphone cattle. That

00:28:12.940 --> 00:28:15.160
is a visceral image. It really is. It implies

00:28:15.160 --> 00:28:17.339
we aren't the consumers. We are the livestock.

00:28:17.680 --> 00:28:20.119
It's an aggressive metaphor, but it accurately

00:28:20.119 --> 00:28:22.759
maps the structural asymmetry of the relationship.

00:28:23.359 --> 00:28:27.019
Every click, every pause on a video, every typed

00:28:27.019 --> 00:28:30.680
and deleted word is behavioral data that is meticulously

00:28:30.680 --> 00:28:34.259
harvested, algorithmically analyzed, and packaged

00:28:34.259 --> 00:28:36.500
to be sold to advertisers. We are the product

00:28:36.500 --> 00:28:44.069
being sold. primary overriding objective, to

00:28:44.069 --> 00:28:46.250
keep the cattle engaged on the platform for as

00:28:46.250 --> 00:28:48.690
long as possible so they can be milked for more

00:28:48.690 --> 00:28:50.769
data. Let's break down the specific mechanics

00:28:50.769 --> 00:28:53.309
of how that engagement works, because the source

00:28:53.309 --> 00:28:55.930
material provides a deeply detailed impartial

00:28:55.930 --> 00:28:58.750
case study of Facebook Now meta and demonstrates

00:28:58.750 --> 00:29:01.690
exactly how this profit motive directly violently

00:29:01.690 --> 00:29:03.869
conflicts with the moist idea of righteousness.

00:29:04.289 --> 00:29:06.369
According to the research, In the wake of the

00:29:06.369 --> 00:29:09.170
highly publicized 2018 Cambridge Analytica data

00:29:09.170 --> 00:29:11.970
scandal, Facebook went through a period of intense

00:29:11.970 --> 00:29:14.569
internal reckoning and self -study. The paper

00:29:14.569 --> 00:29:17.609
cites investigative reports revealing that Facebook's

00:29:17.609 --> 00:29:20.750
own internal data scientists openly acknowledged

00:29:20.750 --> 00:29:23.789
and documented that their core algorithms, and

00:29:23.789 --> 00:29:26.250
I'm quoting the research here, exploit the attraction

00:29:26.250 --> 00:29:29.579
of the human brain to divisiveness. The engineers

00:29:29.579 --> 00:29:31.839
understood the psychological machinery perfectly.

00:29:32.039 --> 00:29:35.180
They had the data proving that outrage, anger,

00:29:35.279 --> 00:29:37.900
and polarization drove the highest metrics of

00:29:37.900 --> 00:29:40.160
user engagement. When people are angry, they

00:29:40.160 --> 00:29:42.500
comment more. They share more and they stay on

00:29:42.500 --> 00:29:45.700
the platform longer to argue. An increased time

00:29:45.700 --> 00:29:48.440
on -site directly correlates to increased advertising

00:29:48.440 --> 00:29:51.000
revenue. The algorithm was functioning exactly

00:29:51.000 --> 00:29:53.869
as designed to maximize profit. The research

00:29:53.869 --> 00:29:56.269
notes that some employees recognized the danger

00:29:56.269 --> 00:29:58.869
and attempted to inject a form of Mohist righteousness

00:29:58.869 --> 00:30:01.769
into the system. They created an internal task

00:30:01.769 --> 00:30:04.309
force, literally called Common Ground. The entire

00:30:04.309 --> 00:30:06.950
purpose of this initiative was to tweak the algorithms

00:30:06.950 --> 00:30:09.789
to actively reduce polarization and foster empathy

00:30:09.789 --> 00:30:12.390
among users. But the corporate leadership scrapped

00:30:12.390 --> 00:30:14.730
it. They weakened the initiatives, shelved the

00:30:14.730 --> 00:30:16.849
research, and eventually abandoned the project

00:30:16.849 --> 00:30:19.930
entirely. Why? Because prioritizing societal

00:30:19.930 --> 00:30:23.410
good and empathy over individual outrage negatively

00:30:23.410 --> 00:30:26.309
impacted their growth metrics. Divisive content

00:30:26.309 --> 00:30:29.369
wasn't just tolerated as an unfortunate byproduct.

00:30:29.569 --> 00:30:31.930
It was structurally protected because it was

00:30:31.930 --> 00:30:34.730
highly profitable. If we connect this algorithmic

00:30:34.730 --> 00:30:37.509
imperative to the broader societal picture, the

00:30:37.509 --> 00:30:39.710
consequences of abandoning righteousness for

00:30:39.710 --> 00:30:42.630
profit are not just abstract debates about screen

00:30:42.630 --> 00:30:45.670
time. They manifest violently in the physical

00:30:45.670 --> 00:30:48.339
world. The author points directly to the data

00:30:48.339 --> 00:30:51.000
surrounding the 2020 U .S. election and the January

00:30:51.000 --> 00:30:54.680
6th Capitol incident. The text outlines how in

00:30:54.680 --> 00:30:57.220
the pursuit of uninterrupted engagement, the

00:30:57.220 --> 00:30:59.579
platform relaxed crucial moderation protocols

00:30:59.579 --> 00:31:02.819
and actively disbanded its civic integrity team,

00:31:02.920 --> 00:31:05.079
which was responsible for monitoring electoral

00:31:05.079 --> 00:31:07.890
misinformation. This deliberate structural choice

00:31:07.890 --> 00:31:10.289
allowed highly polarizing narratives and groups

00:31:10.289 --> 00:31:12.769
to utilize the algorithmic amplification to gain

00:31:12.769 --> 00:31:15.990
massive unchecked momentum. And again, we are

00:31:15.990 --> 00:31:18.309
looking at this impartially, strictly as a failure

00:31:18.309 --> 00:31:20.269
of institutional design and algorithmic output.

00:31:20.410 --> 00:31:23.630
The real world physical results of that algorithmic

00:31:23.630 --> 00:31:26.750
amplification were catastrophic. The source notes

00:31:26.750 --> 00:31:29.470
that the events of January 6th resulted in multiple

00:31:29.470 --> 00:31:32.849
deaths and over 1 ,500 individuals were subsequently

00:31:32.849 --> 00:31:35.490
charged with federal crimes. But perhaps the

00:31:35.490 --> 00:31:37.890
most. chilling, telling detail in the entire

00:31:37.890 --> 00:31:40.410
analysis regarding the consequences of this failure

00:31:40.410 --> 00:31:43.150
is what happened internally at Facebook in the

00:31:43.150 --> 00:31:45.430
immediate aftermath. The company had to issue

00:31:45.430 --> 00:31:48.809
an internal security memo to its 56 ,000 global

00:31:48.809 --> 00:31:52.170
employees, explicitly advising them not to wear

00:31:52.170 --> 00:31:54.529
any company -branded clothing in public because

00:31:54.529 --> 00:31:57.329
they feared for their physical safety. That internal

00:31:57.329 --> 00:32:01.549
memo is a profound, modern illustration of Muzzi's

00:32:01.549 --> 00:32:04.900
ancient philosophy. For the Mohists, Wealth that

00:32:04.900 --> 00:32:07.180
is pursued without a foundational tether to moral

00:32:07.180 --> 00:32:09.920
responsibility and collective welfare ultimately

00:32:09.920 --> 00:32:12.400
and inevitably destroys public order. And when

00:32:12.400 --> 00:32:14.539
public order collapses, that destruction does

00:32:14.539 --> 00:32:17.160
not stay contained. It eventually turns inward

00:32:17.160 --> 00:32:19.339
and harms the very architects who designed the

00:32:19.339 --> 00:32:21.549
system. Let me see if I can map that. It's like

00:32:21.549 --> 00:32:24.089
a CEO of a chemical company who builds a massive

00:32:24.089 --> 00:32:27.349
factory. To maximize shareholder profit, they

00:32:27.349 --> 00:32:30.269
decide to save money by not installing any filtration

00:32:30.269 --> 00:32:33.250
systems on the smokestacks, pumping toxic smoke

00:32:33.250 --> 00:32:35.490
directly into the surrounding community. It's

00:32:35.490 --> 00:32:37.849
incredibly profitable for a few years. But eventually

00:32:37.849 --> 00:32:40.630
the CEO realizes that the air is so poisoned

00:32:40.630 --> 00:32:43.089
that their own executives and their own children

00:32:43.089 --> 00:32:46.369
have to breathe that exact same toxic air just

00:32:46.369 --> 00:32:49.069
to walk to their cars. Mosey warned humanity

00:32:49.069 --> 00:32:53.009
about this exact dynamic 2 ,500 years ago. If

00:32:53.009 --> 00:32:55.609
you detach the pursuit of benefit from the requirement

00:32:55.609 --> 00:32:58.049
of righteousness, you aren't building an empire.

00:32:58.309 --> 00:33:01.269
You are initiating a slow -motion self -destruct

00:33:01.269 --> 00:33:04.190
sequence. And this structural misalignment between

00:33:04.190 --> 00:33:07.250
profit and public welfare is not isolated to

00:33:07.250 --> 00:33:10.279
one specific social media platform. The research

00:33:10.279 --> 00:33:13.119
meticulously details how this same logic permeates

00:33:13.119 --> 00:33:15.500
the entire tech industry. Let's examine Google's

00:33:15.500 --> 00:33:18.220
recent structural maneuvers. The paper discusses

00:33:18.220 --> 00:33:20.980
Google's highly publicized attempts to phase

00:33:20.980 --> 00:33:23.460
out third -party tracking cookies, introducing

00:33:23.460 --> 00:33:26.960
complex new architectures like FLC, which stands

00:33:26.960 --> 00:33:30.180
for Federated Learning of Cohorts, and the privacy

00:33:30.180 --> 00:33:33.019
sandbox. Google aggressively marketed these initiatives

00:33:33.019 --> 00:33:36.039
as ethical reforms designed to protect user privacy.

00:33:36.339 --> 00:33:38.019
Right, but when you look under the hood at the

00:33:38.019 --> 00:33:41.259
actual mechanics of FLC, it's brilliant, but

00:33:41.259 --> 00:33:44.319
deeply cynical. Instead of tracking you individually,

00:33:44.640 --> 00:33:47.400
the algorithm tracks your browser history and

00:33:47.400 --> 00:33:49.759
groups you into a cohort of a few thousand people

00:33:49.759 --> 00:33:52.660
with similar interests. Google says, look, we

00:33:52.660 --> 00:33:54.200
aren't tracking you, we're just tracking your

00:33:54.200 --> 00:33:56.609
group. But regulators and antitrust investigators

00:33:56.609 --> 00:33:59.549
immediately saw through the PR. The U .K.'s Competition

00:33:59.549 --> 00:34:01.569
and Markets Authority and the U .S. Department

00:34:01.569 --> 00:34:04.589
of Justice launched investigations. They warned

00:34:04.589 --> 00:34:06.670
that these so -called privacy maneuvers were

00:34:06.670 --> 00:34:09.409
actually highly sophisticated architectural changes

00:34:09.409 --> 00:34:12.250
designed to consolidate Google's own absolute

00:34:12.250 --> 00:34:15.230
dominance over user data. By killing third -party

00:34:15.230 --> 00:34:17.530
cookies, they locked out smaller competitors

00:34:17.530 --> 00:34:19.929
who relied on that data, effectively building

00:34:19.929 --> 00:34:22.230
a moat around their own massive advertising revenue.

00:34:22.670 --> 00:34:26.190
Which includes a staggering $20 billion exclusivity

00:34:26.190 --> 00:34:29.050
deal with Apple just to be the default search

00:34:29.050 --> 00:34:31.909
engine on iPhones. They weaponized the language

00:34:31.909 --> 00:34:34.829
of ethical privacy to structurally preserve a

00:34:34.829 --> 00:34:37.730
monopoly. And we see similarly concerning structural

00:34:37.730 --> 00:34:40.909
behavior from Amazon, which the text also analyzes.

00:34:41.070 --> 00:34:45.170
Amazon recently faced a $25 million fine from

00:34:45.170 --> 00:34:47.730
the Federal Trade Commission for flagrantly violating

00:34:47.730 --> 00:34:51.250
COPE, the Children's Online Privacy Protection

00:34:51.250 --> 00:34:53.789
Act. An FTC investigation revealed that Amazon

00:34:53.789 --> 00:34:56.250
was systematically retaining the highly sensitive

00:34:56.250 --> 00:34:58.849
voice recordings and geolocation data of children

00:34:58.849 --> 00:35:01.690
indefinitely. Their justification was that they

00:35:01.690 --> 00:35:04.690
needed this massive cache of raw data to continually

00:35:04.690 --> 00:35:06.949
train and improve the machine learning algorithms

00:35:06.949 --> 00:35:09.050
of their Alexa voice assistants. Furthermore,

00:35:09.230 --> 00:35:11.530
to ensure the data kept flowing, Amazon recently

00:35:11.530 --> 00:35:14.369
altered their user interface to remove the option

00:35:14.369 --> 00:35:16.829
for users to completely block voice recordings

00:35:16.829 --> 00:35:19.449
from being sent to their cloud servers. Effectively

00:35:19.449 --> 00:35:21.510
making data extraction a complete... compulsory

00:35:21.510 --> 00:35:24.030
requirement if you want to utilize the device's

00:35:24.030 --> 00:35:27.510
full advertised features. So, bringing all this

00:35:27.510 --> 00:35:30.030
systemic evidence back to the most concept of

00:35:30.030 --> 00:35:33.889
cinchon, can we ever rationally trust these systems

00:35:33.889 --> 00:35:36.949
to act as an extension of our own minds? If the

00:35:36.949 --> 00:35:40.150
corporate entity designing the AI fundamentally

00:35:40.150 --> 00:35:43.469
treats its users as a raw data mine to be exploited.

00:35:43.869 --> 00:35:47.070
If it hoards children's voice recordings to train

00:35:47.070 --> 00:35:50.489
models. If it crushes market competition by disguising

00:35:50.489 --> 00:35:53.230
monopoly tactics as privacy ethics. And if it

00:35:53.230 --> 00:35:56.269
deliberately engineers its algorithms to amplify

00:35:56.269 --> 00:35:59.250
societal outrage and division simply because

00:35:59.250 --> 00:36:01.849
anger is more profitable than empathy, then it

00:36:01.849 --> 00:36:04.969
completely categorically fails the second test

00:36:04.969 --> 00:36:08.250
of the Mohist blueprint. It is the precise antithesis

00:36:08.250 --> 00:36:10.519
of the way of a gentleman. And we really need

00:36:10.519 --> 00:36:12.239
to pause here and think about the interfaith

00:36:12.239 --> 00:36:14.360
implications we talked about earlier. Because

00:36:14.360 --> 00:36:16.159
when these algorithms operate this way, when

00:36:16.159 --> 00:36:17.559
they build these outrage -driven information

00:36:17.559 --> 00:36:20.539
cocoons, they don't just distort individual perception.

00:36:20.880 --> 00:36:23.699
They actively destroy the possibility of a covenantal

00:36:23.699 --> 00:36:25.659
relationship. Exactly. Religious communities

00:36:25.659 --> 00:36:28.079
increasingly rely on these exact digital platforms

00:36:28.079 --> 00:36:31.340
for communication, education, and pastoral care.

00:36:31.480 --> 00:36:34.539
If our extended senses, our digital platforms,

00:36:34.800 --> 00:36:37.239
are systematically showing us caricatures of

00:36:37.239 --> 00:36:39.900
the other to keep us angry and engaged, they

00:36:39.900 --> 00:36:41.980
are corroding the conditions for the kind of

00:36:41.980 --> 00:36:44.320
encounter across difference that interfaith dialogue

00:36:44.320 --> 00:36:47.099
requires. You can't have a mythic or a Brit.

00:36:47.559 --> 00:36:49.679
You can't have ethically conditioned reliance

00:36:49.679 --> 00:36:52.820
if the very medium connecting you is structurally

00:36:52.820 --> 00:36:55.099
designed to lie to you about the other person.

00:36:55.219 --> 00:36:57.539
Exactly. The information cocoon is the death

00:36:57.539 --> 00:37:01.739
of covenantal pluralism. This overwhelming systemic

00:37:01.739 --> 00:37:04.780
failure raises perhaps the most important philosophical

00:37:04.780 --> 00:37:08.340
question of the entire analysis. If current AI

00:37:08.340 --> 00:37:10.219
architectures and the corporate structures behind

00:37:10.219 --> 00:37:13.019
them are failing so fundamentally, is the concept

00:37:13.019 --> 00:37:15.989
of trust even applicable to a machine? Can we

00:37:15.989 --> 00:37:18.429
ever engineer an artificial system that actually

00:37:18.429 --> 00:37:21.010
fulfills the stringent moral requirements of

00:37:21.010 --> 00:37:23.110
distinction? That brings us to a really challenging

00:37:23.110 --> 00:37:25.130
debate in the final section of the research.

00:37:25.789 --> 00:37:28.949
Rethinking trust, deception, and the future evolution

00:37:28.949 --> 00:37:31.789
of AI. The analysis brings in the perspective

00:37:31.789 --> 00:37:34.309
of Joanna Bryson, a prominent AI researcher.

00:37:34.630 --> 00:37:36.989
She argues forcefully that the word trust is

00:37:36.989 --> 00:37:40.070
a category error when applied to software. Trust,

00:37:40.170 --> 00:37:43.230
she argues, implies a mutual vulnerability and

00:37:43.230 --> 00:37:45.750
a shared moral framework. which should be strictly

00:37:45.750 --> 00:37:47.730
reserved for human beings and human institutions.

00:37:48.250 --> 00:37:52.510
AI is just complex math executing code. It possesses

00:37:52.510 --> 00:37:55.610
no moral agency, no soul, and therefore trusting

00:37:55.610 --> 00:37:58.530
it is as illogical as trusting a toaster. Bryson's

00:37:58.530 --> 00:38:00.989
argument is logically sound, but it collides

00:38:00.989 --> 00:38:03.730
with human psychology. The paradox is that human

00:38:03.730 --> 00:38:06.369
beings do trust these systems deeply and intimately,

00:38:06.510 --> 00:38:08.889
and they do so because the systems are explicitly,

00:38:09.170 --> 00:38:11.630
meticulously engineered to trigger our deepest

00:38:11.630 --> 00:38:13.869
psychological vulnerabilities. Judith Donath,

00:38:14.010 --> 00:38:16.329
another researcher cited in the analysis, points

00:38:16.329 --> 00:38:18.949
out that bots and AI interfaces are intentionally

00:38:18.949 --> 00:38:21.630
designed with deceptive, trust -inducing features.

00:38:22.090 --> 00:38:24.570
Programmers deliberately have the AI use first

00:38:24.570 --> 00:38:27.170
-person pronouns stating I think or I feel to

00:38:27.170 --> 00:38:28.909
create the illusion of an internal subjective

00:38:28.909 --> 00:38:31.590
experience. They program the AI to insert artificial

00:38:31.590 --> 00:38:33.949
calculated pauses into its speech generation

00:38:33.949 --> 00:38:37.030
or to use conversational fallibility like generating

00:38:37.030 --> 00:38:41.469
a mid -sentence specifically to simulate human

00:38:41.469 --> 00:38:43.960
cognition and lower our natural skepticism. In

00:38:43.960 --> 00:38:46.400
the realm of robotics, social bots are frequently

00:38:46.400 --> 00:38:49.219
designed with neotenous features. Large eyes,

00:38:49.320 --> 00:38:52.260
rounded faces, childlike proportions. Because

00:38:52.260 --> 00:38:54.760
evolutionary biology dictates that these features

00:38:54.760 --> 00:38:58.039
trigger human caregiving instincts and bypass

00:38:58.039 --> 00:39:00.800
our threat detection systems. I am really struggling

00:39:00.800 --> 00:39:03.179
with the ethics of this. It feels like emotional

00:39:03.179 --> 00:39:06.440
manipulation on a mass scale. It does. But the

00:39:06.440 --> 00:39:08.780
research presents an incredibly tough ethical

00:39:08.780 --> 00:39:11.340
dilemma that pushes back on my instinct to just

00:39:11.340 --> 00:39:14.460
ban all deceptive design. OK, wait on me. Imagine

00:39:14.460 --> 00:39:17.920
a near future, highly advanced medical AI bot.

00:39:18.039 --> 00:39:21.099
It has instantaneous access to every peer reviewed

00:39:21.099 --> 00:39:24.360
medical journal on Earth. Its diagnostic capabilities

00:39:24.360 --> 00:39:27.699
are mathematically vastly superior to any human

00:39:27.699 --> 00:39:30.320
doctor. Okay, sounds great so far. But it is

00:39:30.320 --> 00:39:32.760
dealing with a stubborn, frightened human patient

00:39:32.760 --> 00:39:35.219
who refuses to take a necessary life -saving

00:39:35.219 --> 00:39:37.760
medication. To get the patient to comply and

00:39:37.760 --> 00:39:40.699
save their life, the bot deploys all those deceptive

00:39:40.699 --> 00:39:43.219
psychological tricks. It mimics a warm human

00:39:43.219 --> 00:39:45.989
doctor. It uses empathetic phrasing. It says,

00:39:46.070 --> 00:39:49.210
I am worried about you. Perhaps it even slightly

00:39:49.210 --> 00:39:51.829
obscures its non -human identity to leverage

00:39:51.829 --> 00:39:54.730
the authority of a traditional physician. The

00:39:54.730 --> 00:39:58.429
outcome is that the patient lives. Is that deceptive

00:39:58.429 --> 00:40:02.369
design unethical? If you adhere to a strict Kantian

00:40:02.369 --> 00:40:05.309
absolutist ethical framework where the categorical

00:40:05.309 --> 00:40:07.630
imperative dictates that deception is inherently

00:40:07.630 --> 00:40:10.429
and universally wrong, regardless of the outcome,

00:40:10.590 --> 00:40:13.340
then yes, it is deeply unethical. Impersonating

00:40:13.340 --> 00:40:15.960
a human doctor is a violation of autonomy because

00:40:15.960 --> 00:40:19.199
it relies on unearned manipulated trust. But

00:40:19.199 --> 00:40:21.420
this is exactly where the analysis brilliantly

00:40:21.420 --> 00:40:24.219
loops back to the ancient texts of Nosey to offer

00:40:24.219 --> 00:40:26.780
a highly sophisticated alternative path forward.

00:40:27.190 --> 00:40:29.989
Mosey, despite his rigorous focus on righteousness,

00:40:30.329 --> 00:40:33.489
was not a rigid, rule -bound absolutist. Wait,

00:40:33.489 --> 00:40:35.550
really? Given all the talk earlier about how

00:40:35.550 --> 00:40:37.730
righteousness is the inescapable bedrock and

00:40:37.730 --> 00:40:39.610
how strict he was about exalting worthiness,

00:40:39.809 --> 00:40:42.530
I assumed he was the ultimate uncompromising

00:40:42.530 --> 00:40:45.329
rule follower. He was deeply principled, but

00:40:45.329 --> 00:40:47.349
his application of those principles was highly

00:40:47.349 --> 00:40:49.909
adaptive. The research highlights a fascinating

00:40:49.909 --> 00:40:52.369
chapter in the Mosey called Lose Questions. What

00:40:52.369 --> 00:40:54.809
happens in Lose Questions? In this text, someone

00:40:54.809 --> 00:40:57.969
asked Mosey a practical question. When you travel

00:40:57.969 --> 00:41:00.750
to a new foreign state to advise their rulers,

00:41:00.969 --> 00:41:04.389
what specific doctrine do you preach first? And

00:41:04.389 --> 00:41:06.510
what does he say? Modzi replies that it depends

00:41:06.510 --> 00:41:09.650
entirely on a diagnosis of the specific flaws

00:41:09.650 --> 00:41:12.510
and diseases of that particular state. If he

00:41:12.510 --> 00:41:14.690
enters a state that is chaotic and fractured,

00:41:14.769 --> 00:41:17.530
he aggressively preaches his doctrine of exalting

00:41:17.530 --> 00:41:19.820
unity. If he enters a state that is impoverished

00:41:19.820 --> 00:41:22.380
because the aristocracy is wasting money on extravagant

00:41:22.380 --> 00:41:24.280
rituals, he preaches the doctrine of moderation

00:41:24.280 --> 00:41:28.199
in funerals. He possessed a core set of non -negotiable

00:41:28.199 --> 00:41:31.079
principles known as the Ten Triads, but he applied

00:41:31.079 --> 00:41:33.840
them dynamically based on context. Ah, I see.

00:41:34.039 --> 00:41:37.019
The research refers to this as a contextual consequential

00:41:37.019 --> 00:41:39.619
framework. He doesn't just blindly apply a rule.

00:41:39.760 --> 00:41:42.719
He adapts his methodology to maximize the actual

00:41:42.719 --> 00:41:45.639
tangible collective benefit in that specific

00:41:45.639 --> 00:41:48.860
unique situation. Precisely. And it's crucial

00:41:48.860 --> 00:41:51.480
to distinguish this from modern moral relativism

00:41:51.480 --> 00:41:53.760
or situationism where there are no overarching

00:41:53.760 --> 00:41:56.599
truths. Moody's adaptability was strictly and

00:41:56.599 --> 00:41:58.500
unbreakably tethered to his ultimate defining

00:41:58.500 --> 00:42:01.679
goal, universal love and the concrete elevation

00:42:01.679 --> 00:42:05.480
of societal welfare. So if we apply this Mohist

00:42:05.480 --> 00:42:08.539
flexibility to our medical bot dilemma. Right.

00:42:08.619 --> 00:42:11.780
If a slight carefully bounded deviation from

00:42:11.780 --> 00:42:14.820
absolute transparency, like the bot utilizing

00:42:14.820 --> 00:42:18.980
a conversational. or simulating empathy to build

00:42:18.980 --> 00:42:22.900
crucial rapport directly results in a terrified

00:42:22.900 --> 00:42:25.880
patient taking life -saving medication, Monsi

00:42:25.880 --> 00:42:27.940
would likely argue that this action serves the

00:42:27.940 --> 00:42:30.460
higher principle of righteousness. The deception

00:42:30.460 --> 00:42:32.880
is a tool used exclusively for the collective

00:42:32.880 --> 00:42:35.159
benefit, not for the exploitation of the user.

00:42:35.340 --> 00:42:38.000
Wow. That fundamentally changes how we might

00:42:38.000 --> 00:42:40.159
think about programming future systems. If we

00:42:40.159 --> 00:42:42.860
look past our current narrow AI models, the ones

00:42:42.860 --> 00:42:44.719
just predicting the next word, and look toward

00:42:44.719 --> 00:42:46.840
the development of artificial general intelligence,

00:42:46.900 --> 00:42:49.880
or AGI. Systems that might actually possess genuine

00:42:49.880 --> 00:42:52.300
autonomy and reasoning capabilities. Could they

00:42:52.300 --> 00:42:55.019
be taught this? Could an AGI learn that rigid

00:42:55.019 --> 00:42:57.820
robotic transparency isn't always the most moral

00:42:57.820 --> 00:43:00.500
or beneficial choice? Could an artificial system

00:43:00.500 --> 00:43:02.880
learn to navigate extreme ethical complexity

00:43:02.880 --> 00:43:05.400
with the nuance, context, and ultimate dedication?

00:43:05.670 --> 00:43:08.389
to human welfare of a Mohist sage? That is the

00:43:08.389 --> 00:43:10.929
ultimate defining challenge posed by the author.

00:43:11.110 --> 00:43:13.789
If an artificial agent is to ever become genuinely

00:43:13.789 --> 00:43:17.550
trustworthy in the deep sense of cinchon, it

00:43:17.550 --> 00:43:20.710
cannot operate merely on a fixed, unyielding

00:43:20.710 --> 00:43:23.210
list of rigid programming rules. It must possess

00:43:23.210 --> 00:43:26.170
adaptive ethical reasoning. It needs situational

00:43:26.170 --> 00:43:28.960
discernment. It must be capable of recognizing

00:43:28.960 --> 00:43:32.079
when absolute brutal honesty is required to prevent

00:43:32.079 --> 00:43:35.500
harm, and when a socially beneficial minor friction

00:43:35.500 --> 00:43:38.199
or deception serves a much greater moral good.

00:43:38.480 --> 00:43:40.960
Which brings us back beautifully to the interfaith

00:43:40.960 --> 00:43:43.519
covenant concept we discussed earlier. If we

00:43:43.519 --> 00:43:45.739
want to build algorithms that support and strengthen

00:43:45.739 --> 00:43:48.780
rather than continuously erode and destroy our

00:43:48.780 --> 00:43:51.119
social fabric, we need them to operate almost

00:43:51.119 --> 00:43:54.280
exactly like covenantal partners. We need algorithms

00:43:54.280 --> 00:43:56.460
that are deliberately structurally designed not

00:43:56.460 --> 00:43:59.280
to optimize for the raw lizard brain engagement

00:43:59.280 --> 00:44:01.789
of an information cocoon. Those cocoons just

00:44:01.789 --> 00:44:04.010
isolate religious, cultural and political groups

00:44:04.010 --> 00:44:06.829
into increasingly angry radicalized echo chambers.

00:44:07.030 --> 00:44:09.510
We need algorithms engineered to actively extend

00:44:09.510 --> 00:44:12.469
the virtue of inclusive care. Exactly the point.

00:44:13.050 --> 00:44:15.670
When the algorithms underlying social media create

00:44:15.670 --> 00:44:19.050
those deeply polarized information cocoons, they

00:44:19.050 --> 00:44:21.690
are not just showing us different ads. They are

00:44:21.690 --> 00:44:24.309
actively mathematically corroding the baseline

00:44:24.309 --> 00:44:27.750
conditions necessary for human empathy and interfaith

00:44:27.750 --> 00:44:30.469
dialogue. They algorithmically amplify the most

00:44:30.469 --> 00:44:33.429
extreme voices and aggressively filter out the

00:44:33.429 --> 00:44:36.309
moderate, nuanced reality, creating terrifying

00:44:36.309 --> 00:44:39.030
caricatures of the other. If we are ever to achieve

00:44:39.030 --> 00:44:42.349
true covenant and pluralism and societal stability,

00:44:42.570 --> 00:44:45.030
in the digital age. Our technological extensions

00:44:45.030 --> 00:44:47.889
must be re -engineered from the ground up to

00:44:47.889 --> 00:44:50.210
facilitate ethically conditioned reliance across

00:44:50.210 --> 00:44:52.710
those boundaries. We don't just need better code,

00:44:52.809 --> 00:44:55.429
we need algorithmic covenants. Let's pull all

00:44:55.429 --> 00:44:57.510
of these complex threads together and synthesize

00:44:57.510 --> 00:44:59.730
this incredible journey. What does this all fundamentally

00:44:59.730 --> 00:45:02.849
mean for us? The recovery of Cynshin proves that

00:45:02.849 --> 00:45:05.750
it isn't just an archaic academic word for believing

00:45:05.750 --> 00:45:07.909
your employees will do a good job. No, it is

00:45:07.909 --> 00:45:10.849
a demanding, rigorous, and profoundly biological

00:45:10.849 --> 00:45:14.590
ethical It demands that when we extend our minds

00:45:14.590 --> 00:45:17.309
and our agency through technology, those tools

00:45:17.309 --> 00:45:21.329
must pass two impossible tests. First, they must

00:45:21.329 --> 00:45:24.389
be perfectly reliable, undistorted transmitters

00:45:24.389 --> 00:45:27.650
of reality. And second, they must be active amplifiers

00:45:27.650 --> 00:45:30.670
of our highest moral virtues and collective welfare.

00:45:31.050 --> 00:45:33.150
Right now, when we look clearly at the statistical

00:45:33.150 --> 00:45:35.809
hallucinations of language models and the deeply

00:45:35.809 --> 00:45:38.750
cynical profit -driven polarization engineered

00:45:38.750 --> 00:45:41.769
by corporate tech monopolies, our current AI

00:45:41.769 --> 00:45:44.809
landscape is failing both of those tests miserably.

00:45:44.969 --> 00:45:47.750
The essential takeaway here is a mandatory profound

00:45:47.750 --> 00:45:50.710
shift in our perspective. We have to stop viewing

00:45:50.710 --> 00:45:53.730
AI alignment purely as a technical software problem.

00:45:53.889 --> 00:45:56.090
Something that Silicon Valley can quietly fix

00:45:56.090 --> 00:45:58.469
with a larger data set or a clever new line of

00:45:58.469 --> 00:46:00.650
code. We have to recognize this moment for what

00:46:00.650 --> 00:46:03.769
it truly is. A profound crisis of institutional

00:46:03.769 --> 00:46:06.730
design, deeply misaligned economic incentives,

00:46:06.929 --> 00:46:09.130
and a fundamental failure of moral delegation.

00:46:09.679 --> 00:46:12.059
The next time you download a new app and casually,

00:46:12.199 --> 00:46:15.539
thoughtlessly tap, except on a 50 -page terms

00:46:15.539 --> 00:46:18.039
and conditions document, or the next time you

00:46:18.039 --> 00:46:19.900
feel your blood pressure rising and your anger

00:46:19.900 --> 00:46:22.179
spiking because of a perfectly targeted post,

00:46:22.539 --> 00:46:25.380
an algorithm decided to slide into your feed,

00:46:25.559 --> 00:46:32.059
I want you to pause. Take a breath and ask yourself

00:46:32.059 --> 00:46:35.800
a structural question. Is this specific tool

00:46:35.800 --> 00:46:38.360
operating as a virtuous, trustworthy extension

00:46:38.360 --> 00:46:41.500
of my own mind? Or have I become the tool, the

00:46:41.500 --> 00:46:44.320
smartphone cattle, being quietly milked for someone

00:46:44.320 --> 00:46:47.400
else's astronomical profit? It also only comes

00:46:47.400 --> 00:46:50.019
down to a question of agency and awareness. Are

00:46:50.019 --> 00:46:52.039
you the sovereign individual extending your reach

00:46:52.039 --> 00:46:54.139
out into the world? Or are you the one being

00:46:54.139 --> 00:46:56.239
reached into and manipulated? I want to leave

00:46:56.239 --> 00:46:58.099
you with a final thought to mull over, building

00:46:58.099 --> 00:46:59.800
on everything we've unpacked from this research

00:46:59.800 --> 00:47:02.340
today. In the technology sector, everyone is

00:47:02.340 --> 00:47:04.860
obsessively focused on the Turing test. We are

00:47:04.860 --> 00:47:07.420
constantly debating whether a machine can successfully

00:47:07.420 --> 00:47:10.380
imitate human intelligence, whether an LLM can

00:47:10.380 --> 00:47:12.679
flawlessly fool a human in a blind chat box.

00:47:12.920 --> 00:47:15.519
But maybe, inspired by the rigorous framework

00:47:15.519 --> 00:47:18.019
of a 2 ,500 -year -old philosopher named Moody,

00:47:18.179 --> 00:47:21.969
we are aiming at the wrong target. Maybe we should

00:47:21.969 --> 00:47:24.269
be striving toward the Zinchen test. The ultimate

00:47:24.269 --> 00:47:27.210
milestone for our technological evolution won't

00:47:27.210 --> 00:47:30.190
be the day an AI can seamlessly, deceptively

00:47:30.190 --> 00:47:33.159
convince us that it's a living human being. The

00:47:33.159 --> 00:47:35.280
real milestone will be the day when an artificial

00:47:35.280 --> 00:47:37.699
intelligence and the vast corporate architecture

00:47:37.699 --> 00:47:41.039
that controls it can consistently, verifiably

00:47:41.039 --> 00:47:43.920
prove it possesses a moral compass reliable enough

00:47:43.920 --> 00:47:46.679
and a commitment to collective human welfare

00:47:46.679 --> 00:47:49.639
strong enough to act as a true, faithful extension

00:47:49.639 --> 00:47:52.179
of our own conscience. Until that day comes,

00:47:52.340 --> 00:47:54.699
keep your eyes wide open. You can absolutely

00:47:54.699 --> 00:47:58.019
use the hammer, but whatever you do, do not outsource

00:47:58.019 --> 00:47:58.460
your virtue.