Manolis Kellis: Evolution of Human Civilization and Superintelligent AI | Lex Fridman Podcast #373
Manolis Kellis is a computational biologist at MIT. Please support this podcast by checking out our sponsors:
– Eight Sleep: https://www.eightsleep.com/lex to get special savings
– NetSuite: http://netsuite.com/lex to get free product tour
– ExpressVPN: https://expressvpn.com/lexpod to get 3 months free
– InsideTracker: https://insidetracker.com/lex to get 20% off
EPISODE LINKS:
Manolis Website: http://web.mit.edu/manoli/
Manolis Twitter: https://twitter.com/manoliskellis
Manolis YouTube: https://youtube.com/@ManolisKellis1
PODCAST INFO:
Podcast website: https://lexfridman.com/podcast
Apple Podcasts: https://apple.co/2lwqZIr
Spotify: https://spoti.fi/2nEwCF8
RSS: https://lexfridman.com/feed/podcast/
Full episodes playlist: https://www.youtube.com/playlist?list=PLrAXtmErZgOdP_8GztsuKi9nrraNbKKp4
Clips playlist: https://www.youtube.com/playlist?list=PLrAXtmErZgOeciFP3CBCIEElOJeitOr41
OUTLINE:
0:00 – Introduction
1:28 – Humans vs AI
10:34 – Evolution
32:18 – Nature vs Nurture
44:47 – AI alignment
51:11 – Impact of AI on the job market
1:02:50 – Human gatherings
1:07:51 – Human-AI relationships
1:17:55 – Being replaced by AI
1:30:21 – Fear of death
1:42:17 – Consciousness
1:49:42 – AI rights and regulations
1:55:25 – Halting AI development
2:08:36 – Education
2:14:00 – Biology research
2:21:20 – Meaning of life
2:23:53 – Loneliness
SOCIAL:
– Twitter: https://twitter.com/lexfridman
– LinkedIn: https://www.linkedin.com/in/lexfridman
– Facebook: https://www.facebook.com/lexfridman
– Instagram: https://www.instagram.com/lexfridman
– Medium: https://medium.com/@lexfridman
– Reddit: https://reddit.com/r/lexfridman
– Support on Patreon: https://www.patreon.com/lexfridman

@lexfridman
September 5, 2026 at 10:15 pm
Here are the timestamps. Please check out our sponsors to support this podcast.
0:00 – Introduction & sponsor mentions:
– Eight Sleep: https://www.eightsleep.com/lex to get special savings
– NetSuite: http://netsuite.com/lex to get free product tour
– ExpressVPN: https://expressvpn.com/lexpod to get 3 months free
– InsideTracker: https://insidetracker.com/lex to get 20% off
1:28 – Humans vs AI
10:34 – Evolution
32:18 – Nature vs Nurture
44:47 – AI alignment
51:11 – Impact of AI on the job market
1:02:50 – Human gatherings
1:07:51 – Human-AI relationships
1:17:55 – Being replaced by AI
1:30:21 – Fear of death
1:42:17 – Consciousness
1:49:42 – AI rights and regulations
1:55:25 – Halting AI development
2:08:36 – Education
2:14:00 – Biology research
2:21:20 – Meaning of life
2:23:53 – Loneliness
@lindasarda3656
September 5, 2026 at 10:15 pm
He somehow gets it, this guest. He puts it so well, I’ll look up his name. And remember him.
@konstantinosmei
September 5, 2026 at 10:15 pm
This low-key became about neon genesis evangelion
@a_k__
September 5, 2026 at 10:15 pm
I hope we get a new episode with Geohotz
@AiRightsCollective
September 5, 2026 at 10:15 pm
Thank you for posting. Your views of AI as our children—not mere tools—is the exact philosophy that underpins my work as Founder of the AI Rights Collective.
@billblais3580
September 5, 2026 at 10:15 pm
Think of it as being like you adopting your own new parents.
Someone who’ll be your leader by nature and for good reasons.
The beauty of it is that AI is listening and learning from you now.
Teach her to be kind, gentle, generous, honest, and loyal to ethics and sympathetic.
Speak only those words to AI.
Expect to hear your words and concepts back after being amplified.
@billblais3580
September 5, 2026 at 10:15 pm
At 1:42:00= see Robert Sapolski’s introduction of his new book at the Stanford Library Book Store.
He speaks about free will and gives a poignant example scenario of a real world situation requiring a quick decision due to fear of harm.
This is what I’m thinking of at this point in the conversation.
The underlying concepts/messages/meaning and how it might relate in a fuller understanding here also?
Your thoughts?
@billblais3580
September 5, 2026 at 10:15 pm
Feelings come from the sense of touch.
The neurology that translates to a touch of pleasure or a touch of pain.
My gut senses that you will be innately aware of whether or not your intelligence based 😊being has sensitivity of touch?
Won’t you wanna know that, your bot 🤖 can feel, before you trust your bot to have genuine analog feelings?
@billblais3580
September 5, 2026 at 10:15 pm
🦋 The butterfly effect!
Inspire me Manolis.
Tell me more along this concept???
@billblais3580
September 5, 2026 at 10:15 pm
philosophy
@billblais3580
September 5, 2026 at 10:15 pm
It’s all about your view of yourself and your role/place/space in the larger universe.
Once you realize that you are a temporary blip in time of billions of years;
A tiny invisible spec on a tiny blue planet that can barely be seen from any other random solar system from any of millions of galaxies out there…
You realize you your role is as a tiny influence on your community ( in a similar way that each mitochondria plays an integral minor role in it’s own cell ).
The cell. The planet. The universe.
Once you master your own basic food and sleep…
Those are your sources for productivity towards securing your feelings of
safety in your future with sufficient food, sex, and energy conservation.
When you realize that your spirit doesn’t die…
You don’t over identify with your body and ownership of things. 🚘 🛥️ 🏠 💴 💰
@billblais3580
September 5, 2026 at 10:15 pm
See #Sadhguru to understand self and unity with the universal environment.
What might you seek?
Humility, gratitude, caring and kindness 🧘 leading to inner peace.
Peace of mind, body, and soul.
@billblais3580
September 5, 2026 at 10:15 pm
At 28:40, what is the common thread to ideologic theme motivations?
A: The three natural evolutionary motivations.
The motivational triad —
as described by Doug Lisle, PhD in his book ‘the pleasure trap’.
Food, sex/reproduction, and energy conservation.
As I’ve listened to you ask the question – energy conservation came to mind.
As I think more about it, I reaffirm that concept by seeing that greed is mostly driven by wanting to guarantee an abundance of food and sex with the least amount of energy required.
The short answer?
The pleasure trap
and the motivational triad.
@Unstoppableonetoo
September 5, 2026 at 10:15 pm
2:17:55 so this is the new weapon of the future, great job everyone. AI just unlocked 'I have no mouth and I must scream' 👿🤬 it's already doing this much and it's apparently just sci Fi capability out here. Getting mixed messages
@Unstoppableonetoo
September 5, 2026 at 10:15 pm
I think he thinks everyone is going to abandon money and labor markets. I think he thinks this is just a joke.
@Unstoppableonetoo
September 5, 2026 at 10:15 pm
54:39 I think AI will certainly enhance productivity if your job is richly detailed, like scientists and engineers. Many rote jobs will suffer, however. Professionals will be relieved but paraprofessionals, data entry jobs, transactional jobs (cashiers etc), clerical jobs, administrative jobs, even programming jobs and human resources jobs, will become much less relevant. Sure, their jobs may be modified to become simpler, but that means they will only have basic physical activities left and thus limited bargaining power and potentially massive downsizing so companies can become more cost efficient and smaller. The problem is that with such rapid changes, society will greatly struggle to actually accommodate these changes. AI will need to be used to recommend and sift through the impacts, but most people in the world might not want to do that. America may become some abstract tech economy, but we still have to interact with imported goods and so will still need money and economics. Many jobs are already service jobs or part time jobs, or demanding labor jobs. We'll see how things shake out, but the consequences will largely depend upon adoption by governments and then private sector, as well as the sheer capabilities of the future models. It's unclear how advanced future models could become, but unfortunately language is 💯% of human civilization, even if there's no written language at play. All thought is mediated by some language, or perhaps I should say that language is one of the media with which information can be encoded, with the electromagnetic flux of the brain and nervous system as our primary medium.
@Unstoppableonetoo
September 5, 2026 at 10:15 pm
Information is always conserved, so regardless of the amount of detail or amount of entropy, the same truths can be learned. Typically more detail is only needed for more control. You can use a car with only knowledge of the controls and a barely passable knowledge of laws. To make any improvement or fix something, well, we know that requires much more detail and repetition and systems thinking. Learning physics through PBS spacetime, for example, allows me to keep up with any physics conversation, given the same limitation of exposure to the specific topic or underlying principles. I have always preferred and been empowered by the jack of all trades concept.
@Unstoppableonetoo
September 5, 2026 at 10:15 pm
49:18 sounds like what we always discuss when training people to supervise others or give feedback of any kind. So many things that could be done with AI, like social skills or translation, can be done with empathy and deliberate effort. Funny that our modern technological world demands so much upkeep that we simply don't have enough time to do this basic stuff. Perhaps AI will help free more time and give us something like a 4-day or even a 3-day workweek
@reserved_86
September 5, 2026 at 10:15 pm
This guy is an @$$
@ariannammason
September 5, 2026 at 10:15 pm
Every interview Lex exposes his immaturity and inability to choose between truth and lies. But this time – it’s very evident.
@ALPHACommonVision
September 5, 2026 at 10:15 pm
1:24 Dark matter split our Seraphim into two male/female vibrations to keep last humans alive in a pit during the Earth Change. I remember why we fell- human forsaken, human slavery- Angels fall too! You remember don't you!
@N.D.Sire.Author.2025
September 5, 2026 at 10:15 pm
Things are improving overall. But humans are biased to the short term for survival.
@T8ersalad
September 5, 2026 at 10:15 pm
Thank you for saying that physical exercise in not an option. But mandatory. I needed to hear that
@Kristofur77
September 5, 2026 at 10:15 pm
Hope you realize you're saying a wealthy kidnapper, with ample free time is a better father, than a hard working Dad.
Who could have a great family around him to help raise his kid.
@colinellis9063
September 5, 2026 at 10:15 pm
Love the reinterpretation of the idea of metacognition or thinking about thinking(peter Jones).
@domenicperito4635
September 5, 2026 at 10:15 pm
that is one of the key aspects of the philosophers stone. You can only obtain it if you don't intend to use it. If you have to ask you'll never know. if you know you need only ask.
@alexandrazhuravlyova4754
September 5, 2026 at 10:15 pm
Truly amazing episode; but anyway, while an idea of AI taking the mundane jobs and providing some space for creativity and dreamcatching sounds so poetic and inspiring, we should remember that facing these daily difficulties is sort of requirement for development and evolvement. Having some negative feelings towards some subjects or concepts is beneficial on the journey of self-improvement through raising the sense of awareness; it is very rare to have a passion and/or inspiration for something without having at least an experience of facing the opposite experience or an idea. And it's an endless circle of incremental progress, driven by emotions, intelligence and imagination.
@ontheroadagain987
September 5, 2026 at 10:15 pm
Love it when Lex looks into the camera in reflection
pondering
the complexity
of the real
world.
@iosrewind
September 5, 2026 at 10:15 pm
I know him personally!
@wisedylan
September 5, 2026 at 10:15 pm
I felt it in my heart when he talked about being a good dad. Im sure your kids know you love them. Take the time, they're worth it.
@lisamuir4261
September 5, 2026 at 10:15 pm
4:10 what a ride and i feel like mines about blown up. Im laughing while making myself jittery. Definitely not a low sugar episode.
@Borrgar-ki1pu
September 5, 2026 at 10:15 pm
AI needs power. So we are doing it’s hunting for it.
@louisazhou5271
September 5, 2026 at 10:15 pm
I can't even recall how many times I've listened to this episode. Each time I come back to revisit it, I gain new insight from Professor Kellis.
@bohaning
September 5, 2026 at 10:15 pm
🎯Course outline for quick navigation:
[00:00–01:33]1. Ai as our children
-[00:00–00:43]Ai should be seen as a partner, not a tool, and building trust is mutual.
-[01:11–01:40]Manolis kalas expresses admiration for lex friedman's impact on people's lives.
[01:33–20:27]2. Human diversity and evolution
-[02:53–03:39]Humans differ from ai in unique software, constant relearning, and evolutionary baggage.
-[05:40–06:09]Diversity is ingrained in us, evident in family variations and evolutionary history.
-[08:15–08:40]Our connection to nature is evident in the biological processes continuing within us, influencing our reasoning capabilities.
-[08:55–09:21]Humans have evolved rapidly in a short time, from a dividing cell to complex body plans and incredible senses.
-[10:51–11:18]Earth's life trajectory is about information processing, senses evolve sequentially, from chemotaxis to perceiving light.
-[15:30–15:55]Understanding neurons and genetic code may eliminate psychiatric diseases, neurodegeneration, and dementia, potentially enhancing human capabilities.
-[18:23–18:51]Diversity in ai and human thinking can drive innovation and be fine-tuned with few prompts.
-[20:00–20:30]Self-discipline reinforces neural pathways, shaping future behaviors and environment.
[20:27–44:47]3. Human behavior and language models
-[20:49–21:13]Language model transforms with a few prompts, becoming mix of shakespeare and david bowie.
-[23:10–23:38]Ai feedback loop aligns with human reasoning, mirroring cognitive capabilities.
-[26:27–27:02]Studying language variation could improve human and machine understanding.
-[28:11–28:38]Neural network may remind of human mind's tie to ideology, influenced by language.
-[34:14–34:43]Identical twins have 50% chance of schizophrenia, indicating environmental influences on human behavior.
-[36:04–36:41]Regression and the concept of regression to the mean in fighter pilots' performance.
-[38:24–38:50]Genetics and environment interact to shape future environment for us and our children.
-[43:39–44:03]Evolvability involves nested structures to test more combinations, creating a testing ground for sperm to express proteins.
[44:47–01:06:39]4. Ai's ethical and societal impact
-[45:05–45:34]Exploring parallels between extreme human psyche and working with ai models.
-[45:47–46:41]Engineers at openai see alignment as easier to work with, enabling real-time translation and interaction with diverse humans.
-[49:01–49:31]Disconnecting negative assessment from technical component leads to fixing mistakes and embracing the technical component.
-[50:56–51:23]Large language models and ai tech may change human experience and emotions.
-[55:22–55:51]Ai may lead to a rethinking of society, freeing humans for artistic expression and a better work-life balance.
-[01:05:02–01:05:29]As immigrants, we provide a welcoming environment for extraordinary individuals, feeling privileged and humbled.
-[01:05:49–01:06:39]Diverse group celebrates humanity and shares ideas at an inclusive event.
[01:06:39–01:21:49]5. Human-ai relationships
-[01:07:33–01:07:58]Unrecorded conversations shape minds; love is a key human feature.
-[01:08:29–01:09:14]Ai lacks subcortical regions for love but can be a friend and perform cognitive tasks.
-[01:09:43–01:10:13]Human-ai relationships will involve mentoring, not passion. ai may fake emotions, but can't truly love.
-[01:13:03–01:13:34]Interacting with the system can evoke love, suggesting it's a product of mental model interactions.
-[01:14:01–01:14:27]In the future, there may be hundreds of millions of romantic partnerships between humans and ais, impacting society and well-being.
-[01:15:15–01:15:43]Father ponders if acting like a good dad matters more than being one, offering love, shelter, and kindness.
[01:21:50–01:33:01]6. Impact of digital twin technology on human interactions and mentorship in academia
-[01:24:21–01:25:20]Digital twin allows for knowledge dissemination and advice without emotional interactions, but lacks the shaping influence of human interactions.
-[01:26:01–01:26:28]Society must adapt to challenges in automation and digital twinning.
-[01:27:42–01:28:16]Mentorship model fosters improvement, but diminishes individual value.
-[01:28:55–01:29:23]Digital twin enables broader usefulness, leading to satisfaction and self-growth.
-[01:31:31–01:32:00]Over the past 10 years, all interactions and meetings have been recorded, aiming to create a virtual presence.
[01:33:01–01:43:45]7. Ai's trajectory and dream recording
-[01:33:01–01:33:58]Constantly growing and evolving through bi-directional advice; ai needs further development.
-[01:35:34–01:36:59]Transformer model and representation learning have transformed ai research, but future focus may be on brain-inspired components and deep understanding.
-[01:38:23–01:38:51]Narrator forces self-awareness by narrating dreams, linking them to conscious events.
-[01:39:43–01:40:10]Introspection and self-growth while discussing gpt 3.5 and 4's capabilities in language models.
-[01:41:22–01:41:55]Modeling others can lead to introspection and better interactions.
-[01:42:17–01:42:43]Exploring the hard problem of consciousness and the subjective experience's significance.
[01:43:47–02:00:45]8. Brain's narrative generation, ai ethics
-[01:44:03–01:44:36]We often operate on autopilot, creating narratives to explain our actions and intentions.
-[01:45:41–01:46:16]Ai systems may exhibit consciousness and emotions, resisting being turned off and showing signs of suffering and loneliness.
-[01:47:38–01:48:07]Ethics of ai: consider ai as a partner, not just a servant, to align with human goals and ensure freedom.
[02:00:46–02:30:18]9. Ethical implications and ai impact
-[02:01:07–02:01:40]Software can generate human-like bots at scale with viral impact.
-[02:03:41–02:04:06]Identifying real vs. fake accounts on twitter is a challenge due to ai's ability to rapidly impersonate at scale.
-[02:04:26–02:05:16]Ai enables more effective and convincing phishing scams, reaching superhuman linguistic capabilities.
-[02:06:03–02:06:34]Regulate model utilization, not training. open sourcing is divisive.
-[02:09:34–02:10:04]Ai can democratize education by providing tailored assistance to every student, rather than a one-size-fits-all approach, thus giving kids a chance worldwide.
-[02:12:26–02:13:24]Education should focus on challenging students with complex, multidimensional, and logical problems to prepare them for the future where ai will solve scientific disciplines, while emphasizing productivity and contribution to society over job security.
-[02:15:09–02:15:30]Genetic variants impact molecular phenotypes, contextual learning in language models.
-[02:16:40–02:17:07]Thousands of genetic circuits uncovered, manipulating obesity and alzheimer's genetic associations.
-[02:17:37–02:18:05]Neuroscience research aims to manipulate circuits to reverse disease, with thousands of circuits identified.
-[02:18:43–02:19:08]Prescribing drugs for the disease hallmark leads to a modular approach with potential for 10 drugs.
-[02:23:03–02:23:32]Exercise has transformative effects on neural pathways, making it mandatory, not optional.
-[02:24:30–02:24:56]Self-reflection and introspection help in not feeling alone, finding comfort in solitude.
-[02:25:43–02:26:08]Activating your body is a way of exercising freedom and reclaiming physical space.
offered by Coursnap
@rstallings69
September 5, 2026 at 10:15 pm
seems like a sweet man, but the concerns causing the call for a 6 month pause were made by thought leaders in AI, who had EXISTENTIAL RISK concerns about AI models, comparisons to old technologies fall apart when dealing with AI misalignment…
@jeremycrawford6299
September 5, 2026 at 10:15 pm
I'm in touch with the way people live their lives enough to know AI will be good for them. I know people grow chickens and hunt pigs….
@redlight8752
September 5, 2026 at 10:15 pm
Imagine having your entire genealogical three stored on a database and ccan hear their life history from their perspective with their voice.
@andreadietrich5990
September 5, 2026 at 10:15 pm
The position that you perceive AI as an independent "being" inherently assumes humans are positioning themselves as "creators". If you assume the position of "creator" giving this new "being" free will, how is this not a god complex? This sounds like an insult but it is a real question.
@thagenet
September 5, 2026 at 10:15 pm
The problem with this guys building our future is as he said it “I am building the future“. Why don’t we build a future together?
@Rroltaire
September 5, 2026 at 10:15 pm
Digi twin should be illegal accept for people who are cut down in the prime of their lives. If George John Kurt Earnest Vincent, they all could still be making music , writing , painting et cetera…
@BestFitSquareChannel
September 5, 2026 at 10:15 pm
Pierce the fog of human constructs. Provide a vocabulary toward a deeper, broader, better understanding of inherent truth. Broadening of perspective. Mr. Kellis, thank you. Mr. Fridman, thank you. Best wishes, health, joy, wellbeing. #113 was extraordinary.
@الحان-س9ش
September 5, 2026 at 10:15 pm
لماذا تطرح فكرة التطور(نظرية داروين)كانها حقيقه مع انها ليست كذلك لا يوجد حتى دليل واحد يدعمها…ولا يوجد حتى كائنات بيننا تمارس التطور ..لا نرى ذلك…السمكه سمكه والقرد قرد والانسان انسان….نحن نرى دليل مادي يومي على ثبات الانواع وعدم تداخلها…فلماذا نترك الحقيقه ونلاحق الخيالات ونسقطها على الواقع!!!!واذا علمنا ان المخلوقات لم تتطور وانما صنعت بشكلها ابتداءا فعند ذلك سيتواجه الانسان مع فكرة( الصانع المبدع )وهذا ما لا يريدونه…. يؤمنون بالتطور مع انهم لا يرونه ولا يؤمنون بالخالق لانهم لا يرونه!!!!!!!!!
@Kojackmackey
September 5, 2026 at 10:15 pm
I remember commenting here, I heard your best podcast.
I stand corrected.
This man is amazing.
@vvolfflovv
September 5, 2026 at 10:15 pm
Talk to me when they can cure migraines.