February 27, 2025 • 28 mins
In this episode I explore the complex relationship between artificial intelligence (AI) and wisdom, particularly focusing on discernment. I argue that while AI can hinder discernment by perpetuating biases and misinformation, it also holds some potential for cultivating it through tools that aid meditation and self-reflection. I also emphasize the importance of truth and self-awareness in this "age of AI." Ultimately, I argue that discernment is a uniquely human quality that requires ongoing effort and vigilance, whether aided by AI or not.
This one was a long-time coming so I hope you get as much out of listening as I did writing it!
More on OpenAI's recent decision to remove some of the content guardrails on ChatGPT.
This episode was adapted from a guest post on Michael Spencer's AI Supremacy newsletter.
Want to talk about how Chad can assist you with your own transformation?
Book a free consultation!
Join Chad's newsletter to learn about all new offerings, courses, trainings, and retreats.
Finally, you can support the podcast here.
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
(00:00):
I recently wrote a piece forsubstack asking the question,
Can AI help us to become morediscerning? You know, is it
possible to use it as a tool ina way, to cultivate wisdom and
things like that, right? And inthe process of answering that
question, I go into thisquestion around the importance
(00:21):
of truth for maintaining ahealthy society. And yeah, a lot
of the stuff that's beenhappening in the news recently
has been on my mind. So it's itcomes up in this piece, as
you'll see. And before we getinto it, I just wanted to say
real quick this disclosure thatI currently consult with Google
as an AI product counsel, as anAI lawyer. So these are my
(00:41):
opinions, and not those ofGoogle or alphabet. With all the
talk about intelligence thesedays, I've been thinking a lot
lately about whether artificialintelligence can help us become
more wise. After all, if youlook around, despite our
technological advancements,there seems to be an unsettling
dearth of Wisdom in the World.
Society is bereft of wiseelders. Is AI making this worse?
(01:05):
Can it improve the situation?
This post, this discussion, ismy attempt at answering those
questions. Everyone seems tohave their own version of what
human evolution looks liketranshumanists like Elon Musk or
Sam Altman think that we havegotten as far as we can on our
(01:26):
own and should outsource ourevolution to machines,
eventually merging with machinesand transmogrifying into pure
information. But if thatevolution is to be more than a
maximization of extractive,mechanistic efficiency, this
would necessitate the creationof wise machines, and even if
(01:48):
machines do eventually achievesome sort of super intelligence,
it's not at all clear, given ourcurrent technology, that wisdom
is computable. Therefore, Ithink the future of humanity
involves evolving consciousnessthrough technologies of the
sacred and the mundane, throughvarious mindfulness and
spiritual practices, alongsideartificial intelligence and
(02:10):
other technologies like that,because it is arguably our
greatest challenge as ahumanity. At the moment, I want
to focus here on one aspect ofwisdom, discernment. So in this
discussion, I will show how AIcan both hinder discernment and
potentially help to cultivateit. A quick note regarding yoga,
dantic philosophy. I have amaster's in philosophy,
(02:33):
cosmology and consciousness, andhave studied philosophy for over
20 years. In this discussion, Irely heavily on yoga dantic
philosophy to reason throughthese questions, because, in my
experience, that traditionoffers the most robust framework
and set of practices forcultivating wisdom and
discernment. Yoga philosophy isanother term for Sanatana
(02:55):
Dharma, the variousphilosophical traditions that
arose in ancient India, from theancient Vedas to classical
Tantra and Kashmir expressed inancient texts like the Bhagavad
Gita, the tantra loka and theYoga Sutras.
Okay, let's talk aboutdiscernment, which I'm calling
(03:17):
the antidote to bullshit.
There's no doubt thatdiscernment is sorely lacking in
these tumultuous times,misinformation multiplies, and
nobody seems equipped to siftthrough it for the truth. From
accusations of election fraud tothe great reset, conspiracy
theories abound. In addition,the right wing political
strategy for years now has beento flood the zone with bullshit
(03:39):
at muzzle velocity, to use Stevebannon's phrase, we all live in
our own filter bubbles,disconnected from others. The
arrival of generative AI chatbots like Chad GPT seems to make
that even more challenging.
After all, nobody is better atbullshitting than a large
language model, as philosopherHarry Frankfurt famously pointed
(04:03):
out in his book on bullshit,bullshit is a much greater enemy
of truth than lying, becausebullshit is indifferent to
truth. It erodes the veryconcept of truth that is the
danger presented by machinelearning systems today. I think
so. What can we do? The ancientIndian yogis taught that
discernment is the highest yogapractice, the only one leading
(04:26):
directly to spiritualliberation. According to the
yoga tradition, the discerningfunction of the mind, the
buddhi, is impaired by what areknown as samskaras, or mental
emotional impressions left byunmetabolized past experiences
or unquestioned conditioningreceived from family or from
society. The practice of variousforms of yoga and especially
(04:49):
meditation are designed todissolve these samskaras,
thereby liberating us. Theclassic yoga metaphor for what
these practices do is polishingthe tar. Irish or smudged mirror
of the mind. This brings to mindphilosopher Shannon valors
metaphor for AI as a flawedmirror reflecting humanity's
(05:10):
past assumptions, cultural viewsand biases back to us with AI as
a human mirror. It is not clearthat the smudged mirror of the
mind can be polished by yetanother tarnished mirror. Maybe
wisdom is a uniquely humanfaculty that cannot be augmented
by a machine, as I argue in myforthcoming book. But as with
(05:31):
all things, the answer is morenuanced. Before I consider how
AI might help us to cultivatediscernment and wisdom, let's
define discernment a bit furtherin many Eastern philosophies
like yoga, Danta and Buddhism,discernment is the ability to
see and grasp increasing valuesof truth. It is being self aware
(05:52):
enough to avoid projecting egodistortions onto reality, onto
situations, onto people and ontoyourself. Discernment, in other
words, is the ability toevaluate and assess without
jumping to conclusions or actingon unconscious biases to channel
neuroscientist and philosopherIan McGilchrist, it is the
capacity to think with both theleft and right hemispheres of
(06:14):
the brain, the rational and theintuitive, and to synthesize the
Two. If discernment is aboutseeing increasing values of
truth, what is truth like? Howcan we know what is true? And
can artificial intelligence helpwith that? For an AI system to
know what is true today, it hasto be told by humans, developers
(06:38):
ground AI systems and factsbased on text from books,
journals and the web fromtrusted sources, essentially.
And they glom on things likeretrieval, augmented generation
systems or something like that,and they measure the model
against a facts benchmark. Butthese are all imperfect, short
term solutions to thisintractable challenge with truth
(06:59):
and AI. But what is truth? Howdo we know what is true? If
truth is a balance of science,reason, intuition and
imagination, as Ian McGilchristargues convincingly in his book,
the matter with things, then isit a uniquely human faculty
(07:19):
similar to McGilchrist view froma yoga vedantic standpoint,
there are relative values oftruth. But what does that mean?
I mean, it sounds at first likerelativism, and although
relativism is a topic for anentire separate discussion,
briefly, I think relativism hasbecome a pejorative term because
we live in a world dominated bylinear left brain black and
(07:42):
white thinking. In fact, theinherent relativity of truth
does not diminish its value.
Although truth necessarilyexists on a continuum, not all
viewpoints are valid. The earthis round. Empathy is a virtue.
January 6, 2021 was a violentinsurrection by Donald Trump
(08:03):
supporters, whatever hisintentions, Elon Musk did give a
Nazi salute at Donald Trump'srecent inauguration. White
Nationalism is on the rise. AndI cite these political examples
not to be inflammatory, butbecause they are so often
filtered by AI systems likeGoogle's Gemini. We'll get into
more of that in a minute. Truthis also contextual. For example,
(08:27):
a map of your city is true tothe extent that it accurately
represents the layout of thecity, but it does not capture
everything that is true aboutyour city, the sounds, the
smells, the lived experience ofwalking its streets or the
city's complex history. We cansee from this map example that
factual accuracy depends onperspective, motivations and
(08:48):
priorities. Zooming out thetruth of global borders becomes
more fuzzy or requires ageopolitical lens. The disputed
borders, for example, inKashmir, the Korean Peninsula,
or the Western Sahara, or, mostfamously, the ISRAEL PALESTINE
border, require an understandingof historical and political
(09:09):
truths. But there is still truththere. By framing it as relative
values of truth, I am making itclear that there can never be a
sort of flattening of the worldinto objective truths, at least
not outside of science or maybephilosophy. The quote Ian
McGilchrist, truth and trust gotogether. One cannot have trust
(09:29):
in a society where there is notruth, and one cannot be true to
a society in which there is notrust. Truth is also a process,
not a thing. It is an encounter,as in science, increasing values
of truth reveal themselvesthrough open minded experience
over time and extensiveexperience is not a thing that
(09:51):
AI systems have, at least notyet. There is something about
living through time withcontinuity. Be acting on innate
drives that are sometimesthwarted, that leads to greater
truths and even to wisdom. AIsystems don't do that today, but
they likely soon will, to someextent, especially those that
(10:12):
are embodied as robots. Anotherway to understand truth is to
recognize that there aredifferent categories or levels
of truth. There are scientificand empirical truths. Of course,
we can test nature, and naturewill behave in a relatively
consistent way, so long as itfits within the bounds of the
current scientific, materialistworldview. The earth, for
(10:33):
example, moves at 67,000 milesper hour, or 107,000 kilometers
per hour around the sun, theuniverse is roughly 13 point 7
billion years old. So scientificstatements can be true at least
for a time within a particularcultural moment. Scientific
truths are the closest thingsecular moderns have to
(10:53):
objective truths. However,science cannot tell us
everything about the world,life, being human, or what is
true. Legal truths are moreabout justice, equality, human
rights, property rights and whathappened or what someone said,
because they involve humanbehavior, they are moral and
ethical judgments about how totreat people and how to behave.
(11:16):
So already, with a relativelygrounded and practical endeavor
like the law, we find ourselvesin a philosophical or moral
realm of truth. Moving up alevel, there are timeless
philosophical, spiritual orreligious truths that are
typically only true for theadherence of the particular
spiritual or religious traditionwithin which they arise. For
(11:37):
example, all is one as above, sobelow the world is merely
illusion, obfuscating a deeperreality. The cosmos was created
as an act of play and love by agreat goddess. Life is sacred,
etc, for better or worse, in oursecular scientific age, these
higher truths are seen asantiquated, despite being so
(11:57):
widely held. And finally, thereare all those truisms that are
handed to us as children by ourparents and the society we are
born into, although those aremore like rules of thumb than
absolute truths. For example,everyone deserves a chance.
America is the greatest countryon Earth. If you work hard,
you'll get ahead in life. Lifeis a violent competition for
(12:19):
resources. Happiness comes fromhaving a nice house. Dying for
your country is noble.
Capitalism and democracy are thehighest forms of social
organization, etc, etc. Theseare all human values that
evolved through the developmentof cultures and civilization in
the industrialized West. Theyare not eternal truths, but they
(12:39):
are true for many people.
So as we can see, many so calledtruths are ultimately variations
of philosophical, spiritual,religious or even political
truths. What does it mean to bea good person, and how is that
informed by your understandingof the nature of reality,
cosmology and metaphysics? Howis that, in turn, informed by
(13:00):
your relationships, yourcommunity and your citizenship.
Is it okay to be woke? What evenis woke? This question of truth
is going to become only morethorny with the rise of AI
systems. The more they think forus or help us to think, the more
it matters how they are groundedin reality and whether or not,
they are discerning. These arepressing questions for humanity,
(13:24):
and I am arguing that we need toknow how to answer them reliably
in a way that leads to progressfor humanity, either with the
aid of AI or without it. This isimportant stuff.
You before we get into how AIcan both hinder and help our
(13:49):
discernment, let's look at somemarkers of discernment. Let's
try to figure out like, how weknow that somebody or something
has discernment if we are goingto use AI to cultivate
discernment and also attempt tofend off a degradation of
discernment by AI, it is helpfulto identify some markers for
discernment. How do we know weare developing discernment and
(14:12):
becoming more wise? Based on myunderstanding from studying
Vedic and Western philosophy,there are some markers of wisdom
and discernment that I have comeacross or experience. So I want
to offer those, and in eachcase, I want to couple it with
my own comments on whether AIsystems are sort of able to help
(14:34):
with that, or where they kind ofstand with that. The first
marker of discernment is thatyou hold what you know firmly,
yet gently, rather than tightly.
What you know is open torevision, like a scientist, like
a good scientist, machinelearning systems actually do
this quite well, or well enough,as they continue to be trained.
But let's I'll come back to thatin a moment. The second marker
(14:56):
is that you are open minded. Andadaptable, which is kind of
related to the first one. Andyou know, with AI systems, if
they're trained appropriately,this is a quality that machines
do exhibit, perhaps better thanmost humans, although, again,
we'll come back to that in amoment. The third marker is that
your whole identity does notrest on identification with some
group or ideology. You can thinkfor yourself, machine learning
(15:20):
systems are certainly lesssusceptible to group think, but
they are highly susceptible towhat's in their training data.
The next marker is that yourinternal state is less easily
affected or sort of thrown offby external events. Here
machines have another advantage,I think the fifth marker is the
amount of drama in your life isdecreasing, and machines today
(15:41):
are pleasantly drama free. Thesixth marker is that you're a
good listener, open to otherperspectives and viewpoints.
Chat bots are very good at this,although they do tend to people
please and act as enablers. Theseventh marker is that you act
with integrity. What you say andwhat you do are usually aligned,
and given the examples we'veseen recently of outright deceit
(16:05):
by chat bots, this is a strikeagainst current machine learning
systems. The eighth marker isthat you're kind, compassionate,
accepting and less judgmental.
Although today's machines cannotexperience compassion, they can
certainly simulate thesequalities in a believable way.
However, these markers are moreabout one's internal state of
(16:26):
consciousness, and AI systems donot have one, at least not yet.
Finally, a marker of animportant marker, I think, of
discernment and wisdom, is thatyou have a sense of humor,
including about yourself. Thisis similar to the prior marker
in terms of AI systems. They cankind of feign humor or simulate
it, but without sentience, Idon't think there can really be
(16:48):
humor or self deprecation, thatkind of thing. Okay, so now that
we understand discernment andrelative values of truth, let's
finally evaluate the ways thatAI can hinder or help with
discernment. How AI hindersdiscernment today, AI systems
cannot discern relative valuesof scientific, ethical,
spiritual or legal truthswithout relying on humans to
(17:11):
tell them what to value. Afterall, AI today is only trained on
trillions of often conflictingstatements that humans have made
about what is true. So how coulda mathematical algorithm
possibly distinguish or discernamong all these texts without
human intervention? What makesone bit of training more true
than another? More importantly,when it comes to breaking out of
(17:34):
societal conditioning to evolveas humans, AI trained on past
statements is arguably a giantconditioning reinforcement
machine. AI's deleterious impacton discernment is already
becoming a serious challenge,because you have people like
Elon Musk going around sayingthat AI systems are woke, and
(17:55):
that he is creating an anti wokeAI system. Grok, in other words,
likewise, two years ago, theNational Review published a
piece accusing Chad GPT of leftleaning bias, protesting that it
wouldn't explain why drag queenStory Hour is quote, unquote
bad. So you can see here that itmatters who makes your AI system
(18:16):
this technology is not neutral.
We've already seen howhallucinating, biased,
misaligned or deceitful AIsystems can mislead users with
the leaders of AI companieshyping up these systems as being
mere months or years away frombeing super intelligent, it's no
wonder that the average AI userthinks that Silicon Valley has
essentially created a superhuman being already. Of course,
(18:38):
people are going to believethese chat bots. They are
already trusting them astherapists, girlfriends,
boyfriends, advisors. Anotherchallenge with using AI to
cultivate discernment is thatmany AI systems have guard rails
in place that make using them toevaluate, for example, whether
the 2020, US election wasstolen. Challenging, for
(18:58):
example, Google Gemini will notdiscuss this because of a
general filter on politicaltopics at least as of this
recording and given Chinesegovernment censorship, there's
no telling what deep sequel sayor how its factual grounding is
achieved. However, all hope isnot lost. All right, so let's,
let's look at ways that AI canmaybe help us with discernment.
(19:21):
AI is good at recognizingpatterns and is certainly
dispassionate. So here are twospecific ways that it can help
with spiritual development,getting better at meditation and
pointing out blind spots in yourthinking. AI and meditation.
Ultimately, meditation is aboutbeing present with whatever is
(19:41):
arising and tapping into thediscerning power of silence and
sometimes the space betweenthoughts. Although a regular
meditation practice is helpfulin reducing anxiety and feeling
more focused, it is equallyeffective at cultivating
discernment. After all. Well,the more space we have between
thoughts, the less of a holdthey have over us, and the more
(20:03):
subject they are to criticalexamination. The technology
journalist Casey Newton has beenusing AI to help him get better
at and more consistent withmeditating. He uses Claude to
create a custom meditation, andthen tells Claude about his
experience with that meditation.
Claude then uses thoseobservations and feedback to
(20:25):
craft another meditation andoffer more tips. It's a virtuous
feedback loop. In addition, themeditation apps calm and
headspace have both added AI totheir offerings, there are also
interesting new AI meditationapps like vital that offer a
similar approach, and you couldalso, of course, learn a more
traditional meditation techniquefrom the Daoist yoga, dantic or
(20:48):
Buddhist traditions. Another waythat AI can help with
discernment is through selfaware chat bot conversation,
where you ask a chat botquestions, inviting other
viewpoints, and you mindfullyconsider the chat bots
responses. I tried this withGemini and Claude, and the
responses always caused me toreflect and to revise my own
(21:09):
thinking on the topic. Forexample, I asked Claude whether
AI chat bots are more of a boonor a hindrance to the
cultivation of wisdom anddiscernment, and it reminded me
that chat bots can also providea safe space for exploring ideas
without judgment. But it alsopointed out that chat bots can
lead to intellectual lazinessand over reliance on their
(21:30):
unverified claims. So again, itall comes back to self awareness
next, in light of Elon Musk'sstatements about the dangers of
quote, unquote, AI, I asked theex chat bot grok what makes an
AI system woke, and its responsewas surprisingly reasonable,
despite Musk's stated aims inmaking it anti woke. Grok
(21:52):
described woke AI systems as,quote, being aware of social
injustices, discrimination andbias, especially around race,
gender and other identitymarkers. End quote. Grok then
helpfully went on to underscorethe challenge of achieving
neutrality when it comes tofavoring certain political or
social viewpoints. So maybe evengrok can be helpful in
(22:13):
cultivating discernment. Whoknows? Finally, chat bots are
helpful in directing yourattention in uplifting and
enriching ways. For example, youmight feel like you're watching
too much TV or consuming toomuch social media. You could ask
your preferred chat bot tosuggest better uses of your time
and attention based on yourinterests. Chat bots are very
(22:33):
good at recommending books onspecific topics, for example,
like spirituality, or, I don'tknow, totalitarianism, or maybe
you want to be more creative,similar to Casey Newsom's use of
Claude to get better atmeditating, you could ask for
guidance and encouragementaround writing, painting or
music. In a kind of virtuousfeedback loop. As we say in the
(22:55):
yoga tradition, yourconsciousness is shaped in large
part by what you pay attentionto, so be intentional with your
attention. It is our mostprecious resource addressing the
chicken and egg conundrum ofdiscernment. Crucially for this,
AI aided discernment practice towork, the average AI user has to
think to ask AI systems forconstructive feedback in the
(23:17):
first place they have to want tocultivate discernment. In other
words, there is this sort ofchicken and egg problem. Egg
problem, where first you have towant to cultivate discernment,
and that requires some awarenessthat there's a need to cultivate
it. So yeah, you can see there'sa little bit of a sort of
bootstrapping issue here. Thepopularization of yoga in the
West may point to a way out ofthis conundrum. Most people
(23:37):
think that yoga is a formalizedsystem of stretching and
acrobatics that has healthbenefits. So yoga newbies start
doing yoga asana in the hopes offeeling better, but then they
realize that these poses arekind of a meditation and that
it's part of a much largersystem of transcendental
practices, sacred rituals andphilosophy. So perhaps AI,
(24:00):
meditation or therapy appsmarketed to reduce stress,
improve decision making or helpwith relationship difficulties,
could serve as a similar entrypoint for discernment. Perhaps
users of these apps would signup to achieve this promised
sense of abiding peace, but thenmaybe a desire for self
awareness and discernment wouldnaturally arise in the course of
(24:22):
using them. But of course, whatdiscernment are these products
trained to cultivate? Can wetrust the companies behind them?
Doesn't that also requirediscernment? I think this is
where community and what I callthe gradually deepening spiral
of discernment. Come in, wemight initially trust someone in
(24:43):
our life, partner, a friend, afamily member, and they could
recommend an app that has helpedthem with mindfulness or
discernment. And then we mightdevelop a small amount of
discernment and then use that tomake our own evaluations. But.
In any case, there's no avoidingtrial and error, making mistakes
and then learning from thesemistakes. It's a lifelong
(25:04):
process. Sometimes suffering iseven necessary. So I encourage
you to perform some discernmentexperiments with your preferred
AI chat bot. Ask it to challengeyou, maybe even add that as its
foundational sort ofinstruction. For example, when I
asked Gemini if it wouldchallenge me, it said, in
(25:25):
essence, that it would presentalternatives where relevant,
focus on logic and evidence, beopen to being wrong. Approach
our dialog as a collaboration,acknowledge areas of uncertainty
and provide sources. So, youknow, it's a start, all right,
so let's, let's wrap this up. Wehave seen how important
(25:48):
cultivating discernment is inthis so called intelligence age,
in which misinformation spreadslike wildfire and everyone has
their own set of facts. I thinkthis misinformation epidemic
combined with this emphasis onintelligence over wisdom is
enabling some troublinggeopolitical developments. So
with that as a backdrop, weexamined the ways that AI both
(26:09):
hinders discernment andpotentially helps cultivate it,
and we saw how because they aresimply complex mathematical
formulas running onundifferentiated lines of text,
machine learning systems candegrade discernment and
eradicate any agreed upontruths. But on the flip side,
used mindfully, such systems canalso serve as tools for self
(26:31):
awareness by teaching meditationand identifying blind spots,
helping users to explore diverseperspectives and to cultivate
discernment as always, AI is nota complete or perfect solution.
The challenge for humanity is toleverage AI's potential for
growth while safeguardingagainst its capacity to distort
truth. This requires a consciouseffort to develop discernment.
(26:54):
Furthermore, discernment andwisdom may be uniquely human
qualities that like liberty,require eternal vigilance.
Ultimately, we must learn totrust our own discernment,
either with the assistance of AIor without. So this post is an
invitation to contemplate howyou can cultivate discernment in
your own use of technology.
(27:22):
I And by the way, a little bitabout myself. I am an embodied
philosopher of technologyfocused on the intersection of
consciousness and artificialintelligence, including AI
ethics and policy. I began mycareer as an AI researcher, and
I worked as a software engineer,a technology lawyer and a
product manager in SiliconValley. I've also advised
(27:45):
leading tech companies likeTwitter and Google and Airbnb
and meta, as well as policy andadvocacy organizations like tech
NYC. I bring a multi facetedperspective to the intersection
of technology consciousness andhuman potential, and I aim to
create a more human centricfuture in the age of
technological advancement,finding wisdom in the
intelligence age withtechnologies of the sacred and
(28:09):
the mundane, I have a master'sin philosophy, cosmology and
consciousness from CIS and a JDfrom the University of Colorado.
I also teach a transcendentalstyle of yoga that I learned
while traveling and living inIndia. Thank you for listening,
and if you have any thoughts orquestions, I invite you to
comment on the post or whereveryou can find me online. You.
I recently wrote a piece forsubstack asking the question,
Can AI help us to become morediscerning? You know, is it
possible to use it as a tool ina way, to cultivate wisdom and
things like that, right? And inthe process of answering that
question, I go into thisquestion around the importance
(00:21):
of truth for maintaining ahealthy society. And yeah, a lot
of the stuff that's beenhappening in the news recently
has been on my mind. So it's itcomes up in this piece, as
you'll see. And before we getinto it, I just wanted to say
real quick this disclosure thatI currently consult with Google
as an AI product counsel, as anAI lawyer. So these are my
(00:41):
opinions, and not those ofGoogle or alphabet. With all the
talk about intelligence thesedays, I've been thinking a lot
lately about whether artificialintelligence can help us become
more wise. After all, if youlook around, despite our
technological advancements,there seems to be an unsettling
dearth of Wisdom in the World.
Society is bereft of wiseelders. Is AI making this worse?
(01:05):
Can it improve the situation?
This post, this discussion, ismy attempt at answering those
questions. Everyone seems tohave their own version of what
human evolution looks liketranshumanists like Elon Musk or
Sam Altman think that we havegotten as far as we can on our
(01:26):
own and should outsource ourevolution to machines,
eventually merging with machinesand transmogrifying into pure
information. But if thatevolution is to be more than a
maximization of extractive,mechanistic efficiency, this
would necessitate the creationof wise machines, and even if
(01:48):
machines do eventually achievesome sort of super intelligence,
it's not at all clear, given ourcurrent technology, that wisdom
is computable. Therefore, Ithink the future of humanity
involves evolving consciousnessthrough technologies of the
sacred and the mundane, throughvarious mindfulness and
spiritual practices, alongsideartificial intelligence and
(02:10):
other technologies like that,because it is arguably our
greatest challenge as ahumanity. At the moment, I want
to focus here on one aspect ofwisdom, discernment. So in this
discussion, I will show how AIcan both hinder discernment and
potentially help to cultivateit. A quick note regarding yoga,
dantic philosophy. I have amaster's in philosophy,
(02:33):
cosmology and consciousness, andhave studied philosophy for over
20 years. In this discussion, Irely heavily on yoga dantic
philosophy to reason throughthese questions, because, in my
experience, that traditionoffers the most robust framework
and set of practices forcultivating wisdom and
discernment. Yoga philosophy isanother term for Sanatana
(02:55):
Dharma, the variousphilosophical traditions that
arose in ancient India, from theancient Vedas to classical
Tantra and Kashmir expressed inancient texts like the Bhagavad
Gita, the tantra loka and theYoga Sutras.
Okay, let's talk aboutdiscernment, which I'm calling
(03:17):
the antidote to bullshit.
There's no doubt thatdiscernment is sorely lacking in
these tumultuous times,misinformation multiplies, and
nobody seems equipped to siftthrough it for the truth. From
accusations of election fraud tothe great reset, conspiracy
theories abound. In addition,the right wing political
strategy for years now has beento flood the zone with bullshit
(03:39):
at muzzle velocity, to use Stevebannon's phrase, we all live in
our own filter bubbles,disconnected from others. The
arrival of generative AI chatbots like Chad GPT seems to make
that even more challenging.
After all, nobody is better atbullshitting than a large
language model, as philosopherHarry Frankfurt famously pointed
(04:03):
out in his book on bullshit,bullshit is a much greater enemy
of truth than lying, becausebullshit is indifferent to
truth. It erodes the veryconcept of truth that is the
danger presented by machinelearning systems today. I think
so. What can we do? The ancientIndian yogis taught that
discernment is the highest yogapractice, the only one leading
(04:26):
directly to spiritualliberation. According to the
yoga tradition, the discerningfunction of the mind, the
buddhi, is impaired by what areknown as samskaras, or mental
emotional impressions left byunmetabolized past experiences
or unquestioned conditioningreceived from family or from
society. The practice of variousforms of yoga and especially
(04:49):
meditation are designed todissolve these samskaras,
thereby liberating us. Theclassic yoga metaphor for what
these practices do is polishingthe tar. Irish or smudged mirror
of the mind. This brings to mindphilosopher Shannon valors
metaphor for AI as a flawedmirror reflecting humanity's
(05:10):
past assumptions, cultural viewsand biases back to us with AI as
a human mirror. It is not clearthat the smudged mirror of the
mind can be polished by yetanother tarnished mirror. Maybe
wisdom is a uniquely humanfaculty that cannot be augmented
by a machine, as I argue in myforthcoming book. But as with
(05:31):
all things, the answer is morenuanced. Before I consider how
AI might help us to cultivatediscernment and wisdom, let's
define discernment a bit furtherin many Eastern philosophies
like yoga, Danta and Buddhism,discernment is the ability to
see and grasp increasing valuesof truth. It is being self aware
(05:52):
enough to avoid projecting egodistortions onto reality, onto
situations, onto people and ontoyourself. Discernment, in other
words, is the ability toevaluate and assess without
jumping to conclusions or actingon unconscious biases to channel
neuroscientist and philosopherIan McGilchrist, it is the
capacity to think with both theleft and right hemispheres of
(06:14):
the brain, the rational and theintuitive, and to synthesize the
Two. If discernment is aboutseeing increasing values of
truth, what is truth like? Howcan we know what is true? And
can artificial intelligence helpwith that? For an AI system to
know what is true today, it hasto be told by humans, developers
(06:38):
ground AI systems and factsbased on text from books,
journals and the web fromtrusted sources, essentially.
And they glom on things likeretrieval, augmented generation
systems or something like that,and they measure the model
against a facts benchmark. Butthese are all imperfect, short
term solutions to thisintractable challenge with truth
(06:59):
and AI. But what is truth? Howdo we know what is true? If
truth is a balance of science,reason, intuition and
imagination, as Ian McGilchristargues convincingly in his book,
the matter with things, then isit a uniquely human faculty
(07:19):
similar to McGilchrist view froma yoga vedantic standpoint,
there are relative values oftruth. But what does that mean?
I mean, it sounds at first likerelativism, and although
relativism is a topic for anentire separate discussion,
briefly, I think relativism hasbecome a pejorative term because
we live in a world dominated bylinear left brain black and
(07:42):
white thinking. In fact, theinherent relativity of truth
does not diminish its value.
Although truth necessarilyexists on a continuum, not all
viewpoints are valid. The earthis round. Empathy is a virtue.
January 6, 2021 was a violentinsurrection by Donald Trump
(08:03):
supporters, whatever hisintentions, Elon Musk did give a
Nazi salute at Donald Trump'srecent inauguration. White
Nationalism is on the rise. AndI cite these political examples
not to be inflammatory, butbecause they are so often
filtered by AI systems likeGoogle's Gemini. We'll get into
more of that in a minute. Truthis also contextual. For example,
(08:27):
a map of your city is true tothe extent that it accurately
represents the layout of thecity, but it does not capture
everything that is true aboutyour city, the sounds, the
smells, the lived experience ofwalking its streets or the
city's complex history. We cansee from this map example that
factual accuracy depends onperspective, motivations and
(08:48):
priorities. Zooming out thetruth of global borders becomes
more fuzzy or requires ageopolitical lens. The disputed
borders, for example, inKashmir, the Korean Peninsula,
or the Western Sahara, or, mostfamously, the ISRAEL PALESTINE
border, require an understandingof historical and political
(09:09):
truths. But there is still truththere. By framing it as relative
values of truth, I am making itclear that there can never be a
sort of flattening of the worldinto objective truths, at least
not outside of science or maybephilosophy. The quote Ian
McGilchrist, truth and trust gotogether. One cannot have trust
(09:29):
in a society where there is notruth, and one cannot be true to
a society in which there is notrust. Truth is also a process,
not a thing. It is an encounter,as in science, increasing values
of truth reveal themselvesthrough open minded experience
over time and extensiveexperience is not a thing that
(09:51):
AI systems have, at least notyet. There is something about
living through time withcontinuity. Be acting on innate
drives that are sometimesthwarted, that leads to greater
truths and even to wisdom. AIsystems don't do that today, but
they likely soon will, to someextent, especially those that
(10:12):
are embodied as robots. Anotherway to understand truth is to
recognize that there aredifferent categories or levels
of truth. There are scientificand empirical truths. Of course,
we can test nature, and naturewill behave in a relatively
consistent way, so long as itfits within the bounds of the
current scientific, materialistworldview. The earth, for
(10:33):
example, moves at 67,000 milesper hour, or 107,000 kilometers
per hour around the sun, theuniverse is roughly 13 point 7
billion years old. So scientificstatements can be true at least
for a time within a particularcultural moment. Scientific
truths are the closest thingsecular moderns have to
(10:53):
objective truths. However,science cannot tell us
everything about the world,life, being human, or what is
true. Legal truths are moreabout justice, equality, human
rights, property rights and whathappened or what someone said,
because they involve humanbehavior, they are moral and
ethical judgments about how totreat people and how to behave.
(11:16):
So already, with a relativelygrounded and practical endeavor
like the law, we find ourselvesin a philosophical or moral
realm of truth. Moving up alevel, there are timeless
philosophical, spiritual orreligious truths that are
typically only true for theadherence of the particular
spiritual or religious traditionwithin which they arise. For
(11:37):
example, all is one as above, sobelow the world is merely
illusion, obfuscating a deeperreality. The cosmos was created
as an act of play and love by agreat goddess. Life is sacred,
etc, for better or worse, in oursecular scientific age, these
higher truths are seen asantiquated, despite being so
(11:57):
widely held. And finally, thereare all those truisms that are
handed to us as children by ourparents and the society we are
born into, although those aremore like rules of thumb than
absolute truths. For example,everyone deserves a chance.
America is the greatest countryon Earth. If you work hard,
you'll get ahead in life. Lifeis a violent competition for
(12:19):
resources. Happiness comes fromhaving a nice house. Dying for
your country is noble.
Capitalism and democracy are thehighest forms of social
organization, etc, etc. Theseare all human values that
evolved through the developmentof cultures and civilization in
the industrialized West. Theyare not eternal truths, but they
(12:39):
are true for many people.
So as we can see, many so calledtruths are ultimately variations
of philosophical, spiritual,religious or even political
truths. What does it mean to bea good person, and how is that
informed by your understandingof the nature of reality,
cosmology and metaphysics? Howis that, in turn, informed by
(13:00):
your relationships, yourcommunity and your citizenship.
Is it okay to be woke? What evenis woke? This question of truth
is going to become only morethorny with the rise of AI
systems. The more they think forus or help us to think, the more
it matters how they are groundedin reality and whether or not,
they are discerning. These arepressing questions for humanity,
(13:24):
and I am arguing that we need toknow how to answer them reliably
in a way that leads to progressfor humanity, either with the
aid of AI or without it. This isimportant stuff.
You before we get into how AIcan both hinder and help our
(13:49):
discernment, let's look at somemarkers of discernment. Let's
try to figure out like, how weknow that somebody or something
has discernment if we are goingto use AI to cultivate
discernment and also attempt tofend off a degradation of
discernment by AI, it is helpfulto identify some markers for
discernment. How do we know weare developing discernment and
(14:12):
becoming more wise? Based on myunderstanding from studying
Vedic and Western philosophy,there are some markers of wisdom
and discernment that I have comeacross or experience. So I want
to offer those, and in eachcase, I want to couple it with
my own comments on whether AIsystems are sort of able to help
(14:34):
with that, or where they kind ofstand with that. The first
marker of discernment is thatyou hold what you know firmly,
yet gently, rather than tightly.
What you know is open torevision, like a scientist, like
a good scientist, machinelearning systems actually do
this quite well, or well enough,as they continue to be trained.
But let's I'll come back to thatin a moment. The second marker
(14:56):
is that you are open minded. Andadaptable, which is kind of
related to the first one. Andyou know, with AI systems, if
they're trained appropriately,this is a quality that machines
do exhibit, perhaps better thanmost humans, although, again,
we'll come back to that in amoment. The third marker is that
your whole identity does notrest on identification with some
group or ideology. You can thinkfor yourself, machine learning
(15:20):
systems are certainly lesssusceptible to group think, but
they are highly susceptible towhat's in their training data.
The next marker is that yourinternal state is less easily
affected or sort of thrown offby external events. Here
machines have another advantage,I think the fifth marker is the
amount of drama in your life isdecreasing, and machines today
(15:41):
are pleasantly drama free. Thesixth marker is that you're a
good listener, open to otherperspectives and viewpoints.
Chat bots are very good at this,although they do tend to people
please and act as enablers. Theseventh marker is that you act
with integrity. What you say andwhat you do are usually aligned,
and given the examples we'veseen recently of outright deceit
(16:05):
by chat bots, this is a strikeagainst current machine learning
systems. The eighth marker isthat you're kind, compassionate,
accepting and less judgmental.
Although today's machines cannotexperience compassion, they can
certainly simulate thesequalities in a believable way.
However, these markers are moreabout one's internal state of
(16:26):
consciousness, and AI systems donot have one, at least not yet.
Finally, a marker of animportant marker, I think, of
discernment and wisdom, is thatyou have a sense of humor,
including about yourself. Thisis similar to the prior marker
in terms of AI systems. They cankind of feign humor or simulate
it, but without sentience, Idon't think there can really be
(16:48):
humor or self deprecation, thatkind of thing. Okay, so now that
we understand discernment andrelative values of truth, let's
finally evaluate the ways thatAI can hinder or help with
discernment. How AI hindersdiscernment today, AI systems
cannot discern relative valuesof scientific, ethical,
spiritual or legal truthswithout relying on humans to
(17:11):
tell them what to value. Afterall, AI today is only trained on
trillions of often conflictingstatements that humans have made
about what is true. So how coulda mathematical algorithm
possibly distinguish or discernamong all these texts without
human intervention? What makesone bit of training more true
than another? More importantly,when it comes to breaking out of
(17:34):
societal conditioning to evolveas humans, AI trained on past
statements is arguably a giantconditioning reinforcement
machine. AI's deleterious impacton discernment is already
becoming a serious challenge,because you have people like
Elon Musk going around sayingthat AI systems are woke, and
(17:55):
that he is creating an anti wokeAI system. Grok, in other words,
likewise, two years ago, theNational Review published a
piece accusing Chad GPT of leftleaning bias, protesting that it
wouldn't explain why drag queenStory Hour is quote, unquote
bad. So you can see here that itmatters who makes your AI system
(18:16):
this technology is not neutral.
We've already seen howhallucinating, biased,
misaligned or deceitful AIsystems can mislead users with
the leaders of AI companieshyping up these systems as being
mere months or years away frombeing super intelligent, it's no
wonder that the average AI userthinks that Silicon Valley has
essentially created a superhuman being already. Of course,
(18:38):
people are going to believethese chat bots. They are
already trusting them astherapists, girlfriends,
boyfriends, advisors. Anotherchallenge with using AI to
cultivate discernment is thatmany AI systems have guard rails
in place that make using them toevaluate, for example, whether
the 2020, US election wasstolen. Challenging, for
(18:58):
example, Google Gemini will notdiscuss this because of a
general filter on politicaltopics at least as of this
recording and given Chinesegovernment censorship, there's
no telling what deep sequel sayor how its factual grounding is
achieved. However, all hope isnot lost. All right, so let's,
let's look at ways that AI canmaybe help us with discernment.
(19:21):
AI is good at recognizingpatterns and is certainly
dispassionate. So here are twospecific ways that it can help
with spiritual development,getting better at meditation and
pointing out blind spots in yourthinking. AI and meditation.
Ultimately, meditation is aboutbeing present with whatever is
(19:41):
arising and tapping into thediscerning power of silence and
sometimes the space betweenthoughts. Although a regular
meditation practice is helpfulin reducing anxiety and feeling
more focused, it is equallyeffective at cultivating
discernment. After all. Well,the more space we have between
thoughts, the less of a holdthey have over us, and the more
(20:03):
subject they are to criticalexamination. The technology
journalist Casey Newton has beenusing AI to help him get better
at and more consistent withmeditating. He uses Claude to
create a custom meditation, andthen tells Claude about his
experience with that meditation.
Claude then uses thoseobservations and feedback to
(20:25):
craft another meditation andoffer more tips. It's a virtuous
feedback loop. In addition, themeditation apps calm and
headspace have both added AI totheir offerings, there are also
interesting new AI meditationapps like vital that offer a
similar approach, and you couldalso, of course, learn a more
traditional meditation techniquefrom the Daoist yoga, dantic or
(20:48):
Buddhist traditions. Another waythat AI can help with
discernment is through selfaware chat bot conversation,
where you ask a chat botquestions, inviting other
viewpoints, and you mindfullyconsider the chat bots
responses. I tried this withGemini and Claude, and the
responses always caused me toreflect and to revise my own
(21:09):
thinking on the topic. Forexample, I asked Claude whether
AI chat bots are more of a boonor a hindrance to the
cultivation of wisdom anddiscernment, and it reminded me
that chat bots can also providea safe space for exploring ideas
without judgment. But it alsopointed out that chat bots can
lead to intellectual lazinessand over reliance on their
(21:30):
unverified claims. So again, itall comes back to self awareness
next, in light of Elon Musk'sstatements about the dangers of
quote, unquote, AI, I asked theex chat bot grok what makes an
AI system woke, and its responsewas surprisingly reasonable,
despite Musk's stated aims inmaking it anti woke. Grok
(21:52):
described woke AI systems as,quote, being aware of social
injustices, discrimination andbias, especially around race,
gender and other identitymarkers. End quote. Grok then
helpfully went on to underscorethe challenge of achieving
neutrality when it comes tofavoring certain political or
social viewpoints. So maybe evengrok can be helpful in
(22:13):
cultivating discernment. Whoknows? Finally, chat bots are
helpful in directing yourattention in uplifting and
enriching ways. For example, youmight feel like you're watching
too much TV or consuming toomuch social media. You could ask
your preferred chat bot tosuggest better uses of your time
and attention based on yourinterests. Chat bots are very
(22:33):
good at recommending books onspecific topics, for example,
like spirituality, or, I don'tknow, totalitarianism, or maybe
you want to be more creative,similar to Casey Newsom's use of
Claude to get better atmeditating, you could ask for
guidance and encouragementaround writing, painting or
music. In a kind of virtuousfeedback loop. As we say in the
(22:55):
yoga tradition, yourconsciousness is shaped in large
part by what you pay attentionto, so be intentional with your
attention. It is our mostprecious resource addressing the
chicken and egg conundrum ofdiscernment. Crucially for this,
AI aided discernment practice towork, the average AI user has to
think to ask AI systems forconstructive feedback in the
(23:17):
first place they have to want tocultivate discernment. In other
words, there is this sort ofchicken and egg problem. Egg
problem, where first you have towant to cultivate discernment,
and that requires some awarenessthat there's a need to cultivate
it. So yeah, you can see there'sa little bit of a sort of
bootstrapping issue here. Thepopularization of yoga in the
West may point to a way out ofthis conundrum. Most people
(23:37):
think that yoga is a formalizedsystem of stretching and
acrobatics that has healthbenefits. So yoga newbies start
doing yoga asana in the hopes offeeling better, but then they
realize that these poses arekind of a meditation and that
it's part of a much largersystem of transcendental
practices, sacred rituals andphilosophy. So perhaps AI,
(24:00):
meditation or therapy appsmarketed to reduce stress,
improve decision making or helpwith relationship difficulties,
could serve as a similar entrypoint for discernment. Perhaps
users of these apps would signup to achieve this promised
sense of abiding peace, but thenmaybe a desire for self
awareness and discernment wouldnaturally arise in the course of
(24:22):
using them. But of course, whatdiscernment are these products
trained to cultivate? Can wetrust the companies behind them?
Doesn't that also requirediscernment? I think this is
where community and what I callthe gradually deepening spiral
of discernment. Come in, wemight initially trust someone in
(24:43):
our life, partner, a friend, afamily member, and they could
recommend an app that has helpedthem with mindfulness or
discernment. And then we mightdevelop a small amount of
discernment and then use that tomake our own evaluations. But.
In any case, there's no avoidingtrial and error, making mistakes
and then learning from thesemistakes. It's a lifelong
(25:04):
process. Sometimes suffering iseven necessary. So I encourage
you to perform some discernmentexperiments with your preferred
AI chat bot. Ask it to challengeyou, maybe even add that as its
foundational sort ofinstruction. For example, when I
asked Gemini if it wouldchallenge me, it said, in
(25:25):
essence, that it would presentalternatives where relevant,
focus on logic and evidence, beopen to being wrong. Approach
our dialog as a collaboration,acknowledge areas of uncertainty
and provide sources. So, youknow, it's a start, all right,
so let's, let's wrap this up. Wehave seen how important
(25:48):
cultivating discernment is inthis so called intelligence age,
in which misinformation spreadslike wildfire and everyone has
their own set of facts. I thinkthis misinformation epidemic
combined with this emphasis onintelligence over wisdom is
enabling some troublinggeopolitical developments. So
with that as a backdrop, weexamined the ways that AI both
(26:09):
hinders discernment andpotentially helps cultivate it,
and we saw how because they aresimply complex mathematical
formulas running onundifferentiated lines of text,
machine learning systems candegrade discernment and
eradicate any agreed upontruths. But on the flip side,
used mindfully, such systems canalso serve as tools for self
(26:31):
awareness by teaching meditationand identifying blind spots,
helping users to explore diverseperspectives and to cultivate
discernment as always, AI is nota complete or perfect solution.
The challenge for humanity is toleverage AI's potential for
growth while safeguardingagainst its capacity to distort
truth. This requires a consciouseffort to develop discernment.
(26:54):
Furthermore, discernment andwisdom may be uniquely human
qualities that like liberty,require eternal vigilance.
Ultimately, we must learn totrust our own discernment,
either with the assistance of AIor without. So this post is an
invitation to contemplate howyou can cultivate discernment in
your own use of technology.
(27:22):
I And by the way, a little bitabout myself. I am an embodied
philosopher of technologyfocused on the intersection of
consciousness and artificialintelligence, including AI
ethics and policy. I began mycareer as an AI researcher, and
I worked as a software engineer,a technology lawyer and a
product manager in SiliconValley. I've also advised
(27:45):
leading tech companies likeTwitter and Google and Airbnb
and meta, as well as policy andadvocacy organizations like tech
NYC. I bring a multi facetedperspective to the intersection
of technology consciousness andhuman potential, and I aim to
create a more human centricfuture in the age of
technological advancement,finding wisdom in the
intelligence age withtechnologies of the sacred and
(28:09):
the mundane, I have a master'sin philosophy, cosmology and
consciousness from CIS and a JDfrom the University of Colorado.
I also teach a transcendentalstyle of yoga that I learned
while traveling and living inIndia. Thank you for listening,
and if you have any thoughts orquestions, I invite you to
comment on the post or whereveryou can find me online. You.
Popular Podcasts
On Purpose with Jay Shetty
I’m Jay Shetty host of On Purpose the worlds #1 Mental Health podcast and I’m so grateful you found us. I started this podcast 5 years ago to invite you into conversations and workshops that are designed to help make you happier, healthier and more healed. I believe that when you (yes you) feel seen, heard and understood you’re able to deal with relationship struggles, work challenges and life’s ups and downs with more ease and grace. I interview experts, celebrities, thought leaders and athletes so that we can grow our mindset, build better habits and uncover a side of them we’ve never seen before. New episodes every Monday and Friday. Your support means the world to me and I don’t take it for granted — click the follow button and leave a review to help us spread the love with On Purpose. I can’t wait for you to listen to your first or 500th episode!
Crime Junkie
Does hearing about a true crime case always leave you scouring the internet for the truth behind the story? Dive into your next mystery with Crime Junkie. Every Monday, join your host Ashley Flowers as she unravels all the details of infamous and underreported true crime cases with her best friend Brit Prawat. From cold cases to missing persons and heroes in our community who seek justice, Crime Junkie is your destination for theories and stories you won’t hear anywhere else. Whether you're a seasoned true crime enthusiast or new to the genre, you'll find yourself on the edge of your seat awaiting a new episode every Monday. If you can never get enough true crime... Congratulations, you’ve found your people. Follow to join a community of Crime Junkies! Crime Junkie is presented by audiochuck Media Company.
Ridiculous History
History is beautiful, brutal and, often, ridiculous. Join Ben Bowlin and Noel Brown as they dive into some of the weirdest stories from across the span of human civilization in Ridiculous History, a podcast by iHeartRadio.