Please support this podcast by checking out our sponsors:
- Brooklinen: https://brooklinen.com and use code LEX to get $25 off + free shipping
- Indeed: https://indeed.com/fridman to get $75 credit
- ExpressVPN: https://expressvpn.com/lexpod and use code LexPod to get 3 months free
- Theragun: https://theragun.com/lex to get 30 day trial

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
0:43 - Will AGI suffer from depression?
5:20 - Love is an escape from the muck of life
11:20 - What questions would you ask an alien?
20:04 - How to pivot careers to computer science
27:12 - What will robots look like in the future?
30:00 - Disagreement with Einstein about happiness
35:56 - How I pick podcast guests
45:32 - How to stay optimistic about the future
53:15 - Major topics I changed my mind on
1:00:45 - Benefits of keto diet
1:10:08 - Darkest time in my life

CONNECT:
- Subscribe to this YouTube channel
- Twitter: https://twitter.com/lexfridman
- LinkedIn: https://www.linkedin.com/in/lexfridman
- Facebook: https://www.facebook.com/LexFridmanPage
- Instagram: https://www.instagram.com/lexfridman
- Medium: https://medium.com/@lexfridman
- Support on Patreon: https://www.patreon.com/lexfridman

The following is an AMA episode where I answer a few questions that folks asked on Patreon, YouTube, and other social networks.

I'll try to do these episodes on occasion if it's of interest to anyone at all.

Brooklyn and Sheets, Indeed Hiring website, ExpressVPN, and Theragun muscle recovery device.

So the choice is sleep, employment, privacy, or muscle recovery.

And if you wish, click the sponsor links below to get a discount and to support this podcast.

And now, onto the questions and the answers.

The question is, Lex, I'm a young man that has battled with depression.

Do you think when trying to develop a human-like AI, we will reach a stumbling point where the AI themselves suffer from depression and other complex mental issues?

Do you think it will be a simple fix like rewriting a piece of code or a new patch or update?

Or maybe when trying to create something human-like with high fidelity, you need to leave in the possibility of the AI suffering from such complex mental issues that a human can.

What are your thoughts generally and philosophically about AI suffering from depression?

I think that suffering is a deep fundamental property of consciousness.

I would like to probably say quite a bit about depression.

I have friends who suffer from depression, but that's for another time.

That's for when we talk about depression in humans.

I think depression is just 1 flavor of suffering that is part of the human condition.

I see it as a kind of dark side street on the path to intelligence.

So in terms of robot suffering, if we are to create systems that are truly intelligent, in the way that they're able to interact in intelligent and deeply meaningful ways with other humans, is going to have many of the properties, many of the characteristics of the human condition, of the full human experience.

There's of course a part in us humans that longs to remove all that is cruel in this world.

You know, that's why people that believe in God often the biggest question is of why does God allow there to be suffering in the world?

There's this longing to understand why is there so much unfairness in this world?

And so building on that, there's an inclination to then in our systems engineer something that is void of those things that we cannot understand why that's part of the human condition.

But I think it is intricately part of the experience that is to be human.

And I think if we were to build intelligence systems that are interacting with humans, there has to be, in some ways, properties of consciousness baked in, and if we're to have properties of consciousness baked in, We have to have the full mystery and uncertainty of the human experience, which yes, includes all the different flavors of suffering of which depression is part.

I think the yin and the yang in all of its versions, the ups and downs of moods, but also the more sort of rational, intellectual interpretations of different concepts that are less sort of dramatic, all have to oscillate back and forth.

I think that's where the interesting aspect of interactions happens.

Just like when I have conversations in the podcast, the interesting stuff happens when there's disagreements, when there's a bit of turmoil, when there's a push and pull, when there's a changing of minds, or even just a morphing of your own opinions about something, your own thoughts.

So I really do think all of that mess of humanity has to be engineered in, into AI systems that are interacting with humans and are trying to create meaningful interactions with those humans.

There's of course a huge amount of AI systems that are going to be more intelligent than humans at particular tasks.

Those do not need to have those properties of the human experience like suffering and all those kinds of things.

But for the ones that move among us, I think unfortunately depression has to be part of the experience, or the possibility of depression has to be part of the experience.

Of course I tend to focus on the positive aspects of the human experience, like love, beauty, joy, all those kinds of things, but it's the yin and the yang.

They're lifelong partners, unfortunately, I think.

Now, of course, all of this is just hypothesis, and most of my answers to all of these questions are going to be just my own thoughts, but I am thinking about all of this from an engineering perspective and maybe I'll have more to say in the future about how we actually build these kinds of things into our AI systems that interact with humans.

Thanks for the great question, it's a tough 1.

Question is, Lex, I was wondering if you would be willing to talk about your immigrant experience.

I myself started off as an international student studying and working in America, Not from Russia, I'm from India, but there was a constant push and pull that I experienced given my life circumstance.

I would be curious to hear how you assimilated.

Your statements about do you feel like you belong hit hard for some reason.

Maybe it's because of late at night, maybe because I'm a bit over caffeinated.

Maybe what pops to mind to focus on is the aspect of loneliness, the aspect of belonging.

I think a lot of us in the early teenage years go through that process of feeling like an outsider, an outcast of different kinds.

I think it hit me the hardest personally because I was a popular kid in Russia and when we moved here, I went to the opposite of being popular.

Or feeling like that, I felt like an outcast.

The place I moved to in America had more of an emphasis, maybe it's a cultural thing, of emphasizing material possessions over 2 things that were deeply meaningful to me, which is human connection, like friendship, and also knowledge, like mathematics and scientific discovery, all those kinds of things.

It's just the emphasis of what was valued was different.

And that for me was a catalyst to feel like a total outcast.

As opposed to being this person who looks out into the world and enjoys the beauty of the world, I kind of went to this brooding phase of, first of all, learning the English language, but starting to read books, more philosophical books.

The first 1 I remember reading in English was The Giver, that sort of helped me start thinking about this world.

I was so fortunate to be so in love with people for so long and have close friends in Russia that I didn't notice in my childhood how deeply alone we all are.

So for me, the immigrant experience involved in a small way, at least the first, realizing that hard human truth that we all are born alone, live alone, die alone.

Even when we're in the arms of somebody we love, we're still somehow fundamentally alone with our thoughts, with our hopes, with our fears, trapped in this conscious meat vessel between our ears.

I think the immigrant experience for me was the catalyst to realizing and being terrified and also liberated by the idea that I'm alone in this world.

And at the same time was the realization that this beautiful feeling I felt from the connection to other humans was this gift that took me away from this dark realization.

So it's almost that love is a kind of escape from the reality of life, from the muck of life.

And so the journey began in that way, to think about this world in this way, both the burden of being alone, coupled with the frequent escape from that feeling by being lost in the company of friends, loved ones.

So Early on, coupled with this love of the human mind and curiosity about the human mind was the love of programming and actually building little programs and engineering systems, of course, building robots in college and so on.

I think the gift of the immigrant experience of feeling like the outcast was the love of experiencing the deep connection with others, like a deep appreciation of it when it's there.

I guess because it was taken away, because I was ripped out of it through moving here, I got to really appreciate it and start becoming cognizant of it to where I can start looking for it and being more grateful when I do have it.

And At the same time, a kind of curiosity started boiling up of the perspective on artificial intelligence systems from that kind of longing for a connection.

So as opposed to looking at robots or AI systems or even just programs that accomplish a particular task, can these programs accomplish the same richness of task and richness of experience that I came to appreciate as a human being?

So when I talk about love, there's echoes of that in my longing of the kind of experiences I would like to create in artificial intelligence systems that was born out of the immigrant experience, of the loss of childlike innocence experience, of all of it combined, of starting to read books and thinking deeply about this world experience, all of that coupled in.

I really think sometimes, unfortunately, the first step of deep gratitude is loss.

So for me, I lost quite a bit during that time.

And through that loss, I was able to discover the things that I truly appreciate about life.

Question is, if you were able to ask an alien some questions, what would they be?

This is a really good question, and I find it to be actually a really good thought experiment.

Let me put out some candidate questions out there and see what sticks.

So first I'll probably ask for advice for the human species as a whole, for our civilization, of what we might do to survive and prosper for a long time to come, assuming the alien is from a civilization that's far older than ours or far wiser.

I think there could be some really interesting, clear statements about the things we're doing here on Earth that are getting us into trouble from an alien perspective.

So I think that's the number 1 thing, and maybe I'll bring up along those lines, bring up questions of great filters.

Like, you know, if you look at the history of your civilization, when did you almost destroy the entirety of your species?

It would be informative from a historical perspective to see, like, for us it's currently what the nuclear age and the few moments in the history that could have resulted in an all out nuclear war.

It'd be interesting to see if they mention something about AGI, something about viruses or wars or just things that we don't even think about.

So I guess question number 1 would be like some basic life advice.

Hoping that this alien is like a Naval type character who can like in a crisp, short way give some profound advice.

Second, I would probably ask, now this is a very selfish conversation because it's just following along the things on top of my head that follow my curiosity.

I would ask about the difference between their civilization and ours.

I would ask whether they have some of these things that make us human, like love.

Like do you guys have love where you come from?

You know, I suspect it's possible to have mortality not even be a concept that makes any sense to an alien species, that of course everybody's immortal, and there might be some kind of enforced selection mechanism like evolution in general.

I would ask about consciousness, try to tease apart the question of this thing of subjective experience.

Is this some kind of self-centered, weird, over-dramatized quirk of evolution that we have that's not actually special at all, and then we make a kind of big deal about it.

That's some kind of useful feature of our brain to think of ourselves as individuals that's completely silly.

It'd be interesting to try to tease apart whether they have consciousness and what form their intelligence takes that is distinct from consciousness in the way that we think of humans as being conscious entities that are also able to do intelligent things.

It'd be interesting to sort of tease that apart of how their alien minds work.

So that includes intelligence, consciousness, love and death, all the greatest hits.

Okay, then I would probably go to physics.

I would look into the alien's eyes, if they have eyes, and try to determine if we can actually even find the same language of mathematics of physics or sciences.

In general, we'll probably ask about the big mysteries of physics and science of what's outside our universe.

Why is there something rather than nothing?

And what's outside the stuff we think of as stuff?

I'd be hesitant to ask the why questions, but I'll try a few out to see maybe there is a good answer to the why questions of like, why did it start?

Like, why is there something rather than nothing?

Then I would probably ask slightly more detailed about what's the universe made of?

Like, what's up with this dark matter and dark energy stuff?

Like what are the basic building blocks of reality?

And what are the laws of physics that govern that reality?

So I would of course ask, kind of sneak in there, just like casually, Can you maybe give a few hints of how to unify?

First of all, are we on the right track in terms of quantum mechanics and general relativity?

And then how do you unify all the laws of physics?

Maybe sneak in there in a different angle, trying to ask about the singularity in the black hole, or maybe what happens at the very beginning of the Big Bang, like where those laws are all unified.

Maybe try to get a sense of what are the kind of physics required to fully describe these events.

I think the physics discussion would be a good time to ask, is there a God?

Maybe not use the G word, but instead say, is there a kind of a centralized designer or team of designers that have like launched the universe and are actively managing the universe.

And of course, another version of asking that, I would probably talk about the simulation of looking at the universe as we see it, as a computation, as a computer that's doing information processing, see if that rings a bell to the alien, if there's a connection to that.

In general, I would ask about what kind of computers you have and also what kind of computer games, that'd be really useful.

But that's like usual icebreakers, of course I'm not mentioning those, that's just like chatter at the bar.

So I guess outside the big physics questions, I would ask the more engineering-centric questions.

First, my interest, AI, about superintelligence.

How do we build superintelligent systems, ones that are far more intelligent than humans?

How do we travel close to the speed of light or faster than the speed of light, like how did the aliens get to where we're at that we're meeting and talking?

Related to that would be a question of energy.

How do we harness the energy of a sun or multiple suns or all of the suns in our galaxy?

And then also, kind of an engineering question, can we travel through time?

And if we can, how do we build a time traveling machine?

I think a lot of these questions will be appended with a sort of caveat of like, if you know the answer to this question, will I be better off if you told me this answer?

Sometimes knowledge is a burden that leads to self-destruction.

Of course, as the alien gets tired of talking to me at this intergalactic bar, probably gets up, sort of politely, starts walking away, I would definitely ask some questions, you know, from my own personal knowledge bank.

Theoretical computer science, 1 of the big questions all in mathematics.

Just give me the answer, I'll work from there, okay?

We'll figure out the rest, just the answer, so yes or no.

Probably won't ask him for investment advice.

Probably thinks that the whole concept of money is silly, but I might ask about Bitcoin.

And of course, we'll probably ask, is Elon Musk 1 of you guys or a different species?

Do you know which galaxy, which group of planets it came from.

Is there others like it that visit and build companies?

This AMA has suddenly become ridiculous, but I think this is a really nice thought experiment and I'll think about this a little bit more.

I'm sure there is a list of really precise questions that could most efficiently unlock the mysteries before the human race that are both useful for our progress and useful for our survival.

Question is, what advice would you give an intermediate life stage 36 year old who wants to career pivot from medical technology and research to computer science?

So first, by computer science, I think you mean the broad field that includes software engineering, machine learning, robotics, just computing in general, maybe with less emphasis on the mathematical side, like theoretical computer science.

I think the best advice on this that I could give is find a simple project to get excited about and allow yourself to get really excited by it.

And I should say there's a big emphasis on the simple.

I believe that most people, if they allow themselves, can derive a huge amount of joy for creating some simple little things.

Even if it's following a tutorial, if you just allow yourself to experience the joy of creation, it's there for you.

That's 1 of the magical things about computer science, is it allows you to create things that are almost like entities on their own.

So I think a career in computer science starts first with allowing yourself to be passionate and getting that, stoking that flame and allowing it to build.

So it's not about any of the practical, like which job do I get, what thing I work on, it's just really giving yourself over to the simple passion of creating stuff.

I think there's just a quick set of steps that I think I followed early on that I would also recommend you at least consider following is first is basic software engineering so finding maybe Python or JavaScript like super popular accessible programming language and build just like a Hello World program or something just a little bit more complicated, but not much more.

Beyond that is using that newly acquired set of tools of programming, build something that automates something you do on the computer.

Maybe another way to phrase that is just like scripts that are helping you in your interaction with the computer.

So maybe finding different files in your computer that you try to look for often, or reorganizing things in an automated way, like folder structures, or maybe renaming files.

Like I have a script that finds all the files that have spaces in the file name and it renames them after confirmation to underscores, all those kinds of things.

There's a bunch of little helpful scripts I have all over the place.

And those are just really joyful because you get to use them every day and it's something that you've created that made your life a little bit easier.

For me at least, that's a source of joy that helps feed that love of programming, of just being a part of the computing of the computer science world.

And I've been doing that really my whole life.

It started with C and C++, but now it's a lot of other languages, primarily Python and, yes, JavaScript.

Next is a branching into 2 separate little worlds in computer science of algorithms and then like data science.

I think both are full of beautiful things to fall in love with.

The thing you can really enjoy with algorithms is learning how to build more and more efficient algorithms.

On the data side is learning how to process different data sets, how to clean them up, how to reorganize them and do different kind of statistics on them, processing on them, so we're not even talking about machine learning yet, it's just being able to visualize those data sets, all those kinds of stuff, just working with data.

And now we're starting to talk about a career, because there's a lot of jobs that have to do with the use of computing techniques to process, visualize, interpret, aggregate, analyze data.

So I guess you would call that field data science.

So that's a really cool career trajectory, and there's so many cool things to get into with I think a very reasonable small learning curve that you can really, if you push yourself, do within weeks, maybe months, not years.

And once you become comfortable with the data science world, you can start building on top of that quite naturally, doing some boilerplate machine learning, supervised learning projects, and then building out into more specific, more useful, more novel, cutting edge applications of machine learning, reinforcement learning, that whole world.

Maybe even taking that into physical systems of actually building robots.

And I should backtrack, it sounds like I'm building towards something super complicated, but it's not.

All of these can be really small projects.

Even robotics projects, you can build a little robot that does some basic tasks, maybe does some basic computer vision, And it's a nice way to learn on the robotics side and better systems programming.

So it's just getting more comfortable with hardware and seeing like if that's something you're interested in.

Or on the data science side, where you're sticking much more to the software.

Both of those, you now start to figure out what is the exciting career possibility.

I think 2 things, I would even see them as skills are important here, passion and Google.

I see passion as a skill because it's allowing yourself to be excited.

So it's finding things you could be excited about and allowing yourself to be excited, and seeing that as an actually essential part of progress is allowing yourself to be excited, and the reason I mentioned Google is because I find that in a lot of fields, but especially in computer science, so software engineering, machine learning, there's so many amazing resources out there that the key skill actually ends up being is how good are you at discovering the exact page and resources that is allowing you to take the next step in your journey of exploration of learning.

And that's fundamentally a skill of how do I Google the right thing, what pages do I click on, and all those kinds of things.

I think it sounds almost kind of ridiculous to say that that's a skill, but that is 1 of the most essential skills of the modern day student, lifelong student, is how to Google.

Allow yourself to fall in love with the project and keep taking the next step, the next step, the next step with the help of a good search engine and a bit of curiosity.

Question is, what form factor of robots are you most excited about for the future?

Bipeds, quads, arms, humanoids, maybe something else more obscure?

This is a really tough question because I really like robots.

I think that love is born in software and the hardware stuff just makes it a little more fun.

So I think the things I'm really excited about, even in terms of form factors, is in the software.

I think much of the exciting developments in robotics is actually in simulated worlds currently, and I think that'll be true for quite a while to come.

And so I think in terms of human-robot interaction, the robots that'll be really exciting are the ones that live in virtual worlds, like in virtual reality or even just on a screen.

So I think what we would see more and more is entities, human-like entities, or entities that allow us to anthropomorphize a consciousness, a spirit onto them living in the digital world.

I think that's what I'm really excited about.

And of course, slowly those entities taking a form in the physical space in terms of, I think, probably the humanoid form.

Unfortunately, though very difficult to engineer and create a realistic and natural or fulfilling experience with, I think it's still probably the most, to me, exciting form.

Although I do really like Boston Dynamic Spot, the robot dog, from a kind of having a pet perspective is a really exciting form.

Again, very difficult to do stuff in the physical space.

It's a huge engineering challenge that, as far as I can tell, is several orders of magnitude more difficult than the same challenge in digital space.

So I just see the digital simulated robotics advancing much quicker and having a much larger scale impact on the world, especially if we start seeing more and more virtual worlds being created.

And that doesn't necessarily mean virtual reality or like augmented reality.

It just means ability and mediums within which you can interact with artificial intelligence systems in the digital space.

And I do see that as a form factor, which is entities in digital space having a humanoid or a semi-humanoid form, something that we can anthropomorphize, something we can connect with on a human level.

Question is, on the topic of suffering and growth, is happiness a healthy pursuit, or do you agree with Einstein's view on happiness as the aspiration of a pig?

Okay, let me quickly look up the Einstein quote here that you referenced about a pig and happiness.

Okay, Einstein writes, "'I have never looked upon ease and happiness "'as ends in themselves.

"'This critical basis I call the ideal of a pigsty.

The ideals that have lighted my way and time after time have given me new courage to face life cheerfully have been kindness, beauty, and truth.

Without the sense of kinship with men of like mind, without the occupation with the objective world, the eternally unattainable in the field of art and scientific endeavors, life would have seemed empty to me.

The trite objects of human efforts, possessions, outward success, luxury, have always seemed to me contemptible.

I think I usually agree with Einstein, especially when He talks philosophy on most things, and I do here as well, in terms of material possessions and all those kinds of things.

But I think he unfairly attacks the word happiness and also pigs.

So let me disagree with Einstein and try to defend the word happiness and also maybe defend pigs, if I can somehow figure that out.

So the word happiness, I think, is 1 of those words that could mean a lot of things to a lot of people.

And I think, In this case, Einstein is using it as almost, or the pursuit of happiness as a kind of synonym for hedonism.

So kind of very narrow definition of what happiness is.

See all Lex Fridman transcripts on Youtube

Lex Fridman: Ask Me Anything - AMA January 2021 | Lex Fridman Podcast