SDS 247: The Science Fact of Technology, Ai, & Social Media

Podcast Guest: Pablos Holman

March 27, 2019

Today we have an incredible discussion with Pablos Holman about the more nuanced and in depth nature of technology in the vein of human needs. 
About Pablos Holman
Pablos is a hacker, inventor and technology futurist with a unique ability to distill complex technology into practical tools. A world-renowned expert in 3D Printing, Artificial Intelligence, Big Data, Robotics, Automated Manufacturing, & Cryptocurrency, Pablos has contributed to visions for the future of urban transportation, entertainment, education, energy, manufacturing, health care, food delivery, sensor networks, payment systems & cloud computing. He is on the faculty of Singularity University, founder of Bombsheller and an advisor to data.world & Glowforge.
Overview
Pablos Holman has a crazy and diverse background in data science. He got his start in startups which he considers “the best way to do new things.” New technology are super powers, they give you an ability to solve problems in a way that scales. And when you have that kind of superpower, your job is to figure out where to apply it. Pablos points out this is where difficulty comes in. Most larger companies aren’t designed or have the infrastructure for the risk of doing new things. The answer is startups: a globally scaled experiment or ecosystem for teams and ideas.  
Tech entrepreneurship is utilizing technology to start a business and solve a problem in a way that hasn’t been done before. It’s an appealing kind of startup for both professionals and audiences. Pablos likes to warn he won’t fix anyone’s business. Most businesses have a built in immune system against change and innovation means fighting the company’s defense mechanisms off. Silicon Valley, by and large, learned not to fight that and rather invent a parallel industry to disrupt the existing businesses. You see it everywhere: Uber did it to the taxi industry, AirBnB to hospitality, Facebook to media. They built a superior, parallel industry while the old industries get outpaced. 
This reminded me of artificial intelligence and the concept of a utility function as a double edged sword. An AI is designed with a utility function, say to protect a spouse and might decide that the best way to do this is to consider its own creator a thread to the spouse. You don’t beat the AI by fighting its goal but rather changing the utility function. This is what has to be done for businesses and their “immune system” against change. As Pablos pointed out, Sillicon Valley learned not to fight it, but to work around it. But thinking about AI as biology isn’t always the way to go. Rather, it’s more important, Pablos says, to think about AI in terms of human utility functions: who do you protect, what are your goals, are they at odds with each other? It’s much more complicated than people think. 
Maslow’s hierarchy of needs is a wildly accepted psychological way of looking at human behavior and incentives. You’re born in America, those basic needs are, for the most part, solved. You go higher up to family, friendship, community, and if you think about technology as a way to meet needs we’ve only used technology for the bottom half of the pyramid: food, health, shelter. Tech stays very basic but we have higher needs sexually, socially, and culturally. Silicon Valley isn’t even trying to solve those needs and people can feel that. The robots will come and take all the jobs that will excel at, compared to people. But robots can’t come and take the jobs that we’re still working on because it’s hard: teaching, taking care of the elderly, helping people learn new skills. Technology can’t solve that. We have to. And sometimes technology is making the top levels of that pyramid worse. 
There’s addiction curves to new technology, Pablos admits, but also points out it’s not as terrible as people often make it out to be. Too much email, too much social media, too much texting. But eventually we get it under control. It’s usually a period of being inundated with a new piece of technology before taking it from granted and moving on. There’s also positives. Pablos grew up isolated in Alaska and felt alone in his interest in computers and technology in a remote state. But now, information and technology has helped people not feel alone. 
However, social media can be addictive. And 12 year olds are getting on Facebook and Instagram while there’s regulations for who can drink alcohol and smoke. Should we regulate them that way? Pablos says no. Kids are so used to technology now that it doesn’t affect them in the same way. It’s not perfect, but they don’t freak out like we do. It’s not as dangerous for them. Humans tend to focus on something that’s not entirely positive, they get regulated too soon as a reflex, and it hurts society. Pablos compares it to the lack of regulation on alcohol in Europe where children have a much more flexible relationship with alcohol and don’t go through the drinking binges that American teenagers and young adults do in college.  
We’re on the verge of a new way to use data, to use technology, to look at social media. Pablos is excited and doesn’t think having hyperbolic reactions gets us anywhere. We have to use our imagination to harness computational ability, and it’s exciting. 
In this episode you will learn:
  • Startups & parallel industries [8:13]
  • Maslow’s hierarchy of needs in tech [23:13]
  • Regulating social media [40:00]
  • Role of data in the information world [49:50]
  • Why you should go to DataScienceGo [55:40]
Items mentioned in this podcast:
Follow Pablos
Episode Transcript

Podcast Transcript

Kirill Eremenko: This is episode number 247, with the incredible Pablos Holman.

Kirill Eremenko: Welcome to the SuperDataScience Podcast. My name is Kirill Eremenko, Data Science Coach and Lifestyle Entrepreneur, and each week we bring you inspiring people and ideas, to help you build your successful career in Data Science.
Kirill Eremenko: Thanks for being here today, and now let’s make the complex simple.
Kirill Eremenko: Hey everybody, welcome back to the SuperDataScience podcast. Super excited to have you on the show today, and we’ve got a very, very special guest joining us for this episode, Pablos Holman. The famous hacker, inventor and entrepreneur. A person who has worked on projects ranging from a brain surgery tool to hurricane suppression machines. From 3D printing food to lasers shooting down mosquitoes.
Kirill Eremenko: Pablos has over 20 million views on his TED Talks. He was our keynote speaker at DataScienceGo 2018, and today he’s joining us for this episode of the SuperDataScience podcast. So very, very excited to share this episode with you. We’re going to be discussing lots of different topics, ranging from how artificial intelligence is impacting the world, what the Maslow’s hierarchy of needs is, and how that is affected by technology. What roles Data Science and machine learning are playing in the future of the world, and many, many more.
Kirill Eremenko: I can’t wait to share this episode with you. A quick note is that this episode is also available in video on YouTube. So you can go to the SuperDataScience YouTube channel, and watch this whole episode in video, of Pablos and me chatting over the internet.
Kirill Eremenko: So if you like you can go to the SuperDataScience YouTube channel, and watch this whole episode of Pablos and me chatting over the internet. And whichever way you like to watch it, or hear it, whether it’s on YouTube or here on the podcast, this is going to be epic. We’re going to discuss so many cool topics. I had such a fun chat with Pablos.
Kirill Eremenko: So without further ado, I bring to you, Pablos Holman, the famous hacker, inventor and entrepreneur.
Kirill Eremenko: Welcome to the SuperDataScience podcast ladies and gentlemen, super excited to have you on the show, and today we have a super special guest, Pablos Holman. Pablos, how are you today?
Pablos Holman: Hey I’m doing great.
Kirill Eremenko: Awesome, awesome. We were talking before about the drum in the background. That is so cool. Where’s that one-
Pablos Holman: Oh yeah. I’ve got a djembe that’s made of fiberglass.
Kirill Eremenko: Oh nice.
Pablos Holman: Which makes it super light, and super loud, and anyone can play it with just fingertips, and it sounds amazing even if you don’t know how to play the drums. It’s a lot of fun.
Kirill Eremenko: Is that like a side hobby of yours?
Pablos Holman: I guess it used to be. I used to play around with that stuff some. Not so much these days. But it’s fun to make noise.
Kirill Eremenko: Got you, got you. Well Pablos, great to have you on the show. For those who don’t know, Pablos was our keynote speaker at DataScienceGo 2018. It was so much fun to meet you there, and I think everybody had lots of fun at your talk. I love how you started out with a little bit of salsa dancing. Where did you pick up salsa?
Pablos Holman: Probably 14 years now I’ve been obsessed with dancing salsa. So it’s just a good way to … I don’t know. I think it’s a good way to compensate for all the other kinds of things I do in my life, and you can do it anywhere in the world. You don’t need to speak the language, you just show up and dance with whoever’s there. So it’s really good. It’s a lot of fun.
Kirill Eremenko: Nice. Very cool. So what did you think of the event? What did you think of DataScienceGo 2018?
Pablos Holman: Well it was a really, I think very special, soulful kind of event. I was really pleasantly surprise. A lot of times you get a bunch of nerds together, and you don’t get that much energy. I felt like people loved being there. It feels like a real community. Folks are trying to help each other out. They’re super exited to be there.
Pablos Holman: I’ve been to a lot of events that are way less compelling than that, and so I thought that was pretty special. I was so fascinated by the folks there. It felt like possibly … I think there’s no question, it had to be the most diverse audience I’ve been in, in the tech industry. For a tech event, I felt like there were two or three other white guys around, but it was pretty much people from all over the world. It was amazing to me. I love that, and it was so exciting to see that happening.
Pablos Holman: So I don’t know how you did that, got a lot of people from different ethnic backgrounds, bunch of people from different countries, you got lots of women there. That’s not the case at almost every tech event, and it’s a pretty special thing. So whatever you’re doing, keep it up, because everybody else is going to be jealous of what you’ve made there.
Kirill Eremenko: Thank you. Thank you. That’s very cool. Yeah, I remember we were having lunch, you pointed that out, and that’s the first time I actually paid attention. It was so natural to me, and it was like, “Well, that’s how things should be.” And this time round we had people from 23 countries fly into the event.
Pablos Holman: Wow, wow. Cool.
Kirill Eremenko: Really cool, yeah. Well thanks, and definitely looking to grow that, and continue that trend. I think we had 30% female speakers this time, because that whole concept … I read an article recently on why it is important. How gender diversity can be improved in the tech industry.
Pablos Holman: Sure.
Kirill Eremenko: And the reason why women don’t … Like why don’t we see a lot of women in STEM space, in science, technology, engineering and maths, is because historically there haven’t been many role models. There haven’t been many women to look up to. So young women aren’t that confident getting into the space because it’s like-
Pablos Holman: Sure.
Kirill Eremenko: But if we put speakers out there, if we put guests on the podcast, and build those … Or not even build, showcase those role models, that will inspire more women to get into the space.
Pablos Holman: I think that’s definitely a factor, yeah.
Kirill Eremenko: And you’ve probably seen this from your entrepreneur career, that when you have gender diversity in the executive team, company performs much better. That is a fact that has been studied, and there’s lots of research papers on this topic.
Pablos Holman: I’m certain that’s true. I also think if you were operating without that, you’d just be missing out on … it’s cutting your talent pool in half at the start. There’s no way to characterize that as an advantage.
Kirill Eremenko: Yeah, exactly, exactly. All right, well there’s quite a lot of topics I want to cover. By the way, everybody listening, you can find Pablos’ keynote on YouTube. It’s available from the event. So if you want to watch it there, so we won’t go into all the exact same topics, even though there was a couple that I did want to cover.
Kirill Eremenko: One of the most interesting things that … There’s lots of things we chatted about. But one of the really cool things is, in general your background as an entrepreneur.
Pablos Holman: Yeah.
Kirill Eremenko: You do a lot of work in the Silicon Valley, and you’ve done a lot of different projects. From working with Bill Gates, working on the worlds smallest computer, to hacking and stuff like that. You’re TED talks have like 20 million views. In terms of entrepreneurship, I love that idea that you shared where entrepreneurs are not here to fix your business. Entrepreneurs are here to disrupt the world. Can you tell us a bit about how that works in the Silicon Valley?
Pablos Holman: Yeah, I think high tech entrepreneurship maybe in the ’80s, was necessary, because the only way to get a new technology developed and advanced into the world, turned out to be in the context of a startup. It was increasingly difficult to do that in big companies.
Pablos Holman: Then also, if you were working on technology or computers, you were trying to do new things, almost by definition. So if you were that type of person who wants to do new things, well then it turns out that startups are often times the effective way of doing new things.
Pablos Holman: So that’s worked out well enough that you can get a lot of support for startups now, and that’s pretty great. The fundamentals are that you can get these new superpowers from technology. Right? If you invent a new technology, and you bring that into the world, you can get a force multiplier on how you go after solving big problems. You don’t get that force multiplier from anything else, it’s literally the exponent in your equation that you get that gives you this exponential ability, this ability to solve problems in a way that scales. That’s because lots of new technologies, especially computer technologies, have been generally applicable, that you can solve lots of different kinds of problems with them, and they can scale. Meaning you could, you know if you can make one computer do something, it’s not that much harder to make thousands or millions of computers do the same thing.
Pablos Holman: So when you have that kind of superpower, then the job is to go figure out where to apply it, and if you look at what happens in big companies, a lot of times they’re just not structured well for doing something new. Especially these days as publicly traded companies in particular, but really every company has gotten a lot of pressure to show immediate performance results, at least on a quarterly basis, if not more often, for their stockholders. Well they can’t take risks on doing new things too much. They have to do the same thing they already know works, and they have to do it a little faster, a little better, a little cheaper.
Pablos Holman: So because of all that, startups became the context in which we’re able to run a lot of experiments, and try new things, and figure out what works, and get some support for it, and then that’s the necessary process of figuring out what’s going to work. And a startup is kind of an experiment. On a global scale, startups are experiments. Each startup is a million dollar experiment to just try an idea, see if it can work, see if the team works, see if it makes competitive sense in the market. All those things. And then the ones that work, well then we pump a lot of money into scaling them, or we sell them to bigger companies that are good at that kind of thing.
Pablos Holman: But it’s an ecosystem for running experiments. That’s what startups are. So if you’re doing tech entrepreneurship, what that means is it’s different than entrepreneurship in general. Right? If you start a restaurant, well that counts as entrepreneurship. You could go start a restaurant, you’ve got a new business, you’re an entrepreneur, which is possibly just fine, but that’s not tech entrepreneurship. Tech entrepreneurship means I’m using technology to make a business, and solve a problem in a new way, that hasn’t been possible before. Hasn’t been done before.
Pablos Holman: So that’s a particularly appealing kind of startup, because if it works, then you’re going to get this big hit, because you’ve taken that force multiplier, you’ve solved a problem in a way that no one could imagine solving it before. So those are the fundamentals.
Pablos Holman: The reason that I say we won’t fix your business, is that every business that’s already successful has already evolved a kind of immune system. And the job of the immune system is to get rid of any risk. Just like the immune system in your body. And the thing that’s most risky is change. So in any big company, or successful industry, there’s this immune system, and if you go in and you want to innovate, or you want to try something new in that context, you have to fight the immune system off. The immune system has evolved over a long period of time, and it’s really good at it’s job, and you’re just a startup, you suck at your job almost by definition.
Pablos Holman: So what I think is, Silicon Valley has learned not to fight the immune system. Some people haven’t, but by and large what we’ve figured out works better is if we can take all those superpowers that we got from our computers, and we can invent a parallel business, or a parallel industry, and make an end run around you, and around the immune system. That’s what we mean by disruption. And that’s what you can see happening with all those so called disruptive businesses.
Pablos Holman: Uber did that to taxis, or transportation in general. Urban logistics in general. Airbnb did that to hospitality. Facebook did that to the entire media industry. We did not try in any of those cases to go fix an old broken industry. In each of those cases, Silicon Valley just started from scratch and built a new parallel business, or a parallel industry that is superior to the old one. So we’re just going to let the old industries grow old and die. We’re going to make up new ones to replace them.
Kirill Eremenko: Nice. Very, very apt description. And what is the percentage of successful startups that you see?
Pablos Holman: Well we always say that it’s only 10%. But the truth is, we say nine out of ten startups fail. That’s a made up number. They don’t fail, it’s just that … fewer than that fail, but…
Kirill Eremenko: They learn?
Pablos Holman: But they fail to make a 10 or 100 X return for their investors. So if you’re a venture capitalist and you’ve got a startup that’s profitable and is only going to give you a two or three X return on your investment, that’s considered a failure.
Kirill Eremenko: Wow.
Pablos Holman: They’ll shut you down, because that’s not useful for them. They need to be getting bigger hits, and they need to shut you down and move on, and go work on things that could provide that 10 X to 100 X return.
Kirill Eremenko: Opportunity cost for them.
Pablos Holman: Yeah. Because if you’re just going to get two X return on your investment over five years, well you could have put your money in a stock market, or in something else right? So there’s no point doing high risk venture capital with the money.
Pablos Holman: So it’s a little bit misleading. I mean lots of startups, they just fail to live up to the kind of hype that they sort of generated when they were raising money. When you raise money, venture capital for a startup, it’s basically the worst loan in the world. Because if I take 10 million dollars from a venture capitalist, and my startup is successful and the VC is happy, that means he made … I paid him … he got 100 million dollars back, for his 10 million dollar investment.
Kirill Eremenko: Yeah, that makes sense.
Pablos Holman: 1000% interest on your loan. Right? And that’s the success. The failure is, he didn’t get any of his money back, and my startup failed, so I’m down, and I lost years of my life, and that’s when it goes south. That’s the worst loan in the world. But that’s how it works, because nobody else is going to loan you money for your crazy idea.
Kirill Eremenko: Yeah. Interesting. That’s very cool. What you were saying about the immune system of big business, that’s really a fascinating idea, and reminded me of the concept of the utility function in artificial intelligence. The whole idea when we, and if, and probably when we create general artificial intelligence, it will have its own utility function. For instance, let’s say hypothetically, Ben Taylor uses this example, the utility function of the AI is to protect my wife, make sure she’s safe at all costs, at all times.
Kirill Eremenko: And then the AI decides that first of all it could decide that the best way to protect his wife is to actually kill him. That’s a problem in the programming, he didn’t think that through.
Kirill Eremenko: But on the other hand, the idea with the immune system is if you go in and you try to change the utility function of the AI to be something like, “Protect my wife and kids.” Or, “Make a great dinner,” or what ever, then all of a sudden, you changing the utility function of the AI is a threat to it executing its utility function in its own right, and therefore it will prevent you at all costs, from changing the utility function.
Kirill Eremenko: And I thought that was a pretty interesting analogy.
Pablos Holman: Well I think we’re using these analogies from biology to try and give people a way of thinking about the technology. Right? But I think it’s actually almost more important to do the opposite, which is think about the utility function of a human. This is much more important for people to do. What’s the utility function of a human, of an actual man. Is it to protect his wife and kids? Is it to try and improve the community that he’s a part of? Those things might be at odds with each other. Is it to become the alpha male of the tribe? These kinds of things.
Pablos Holman: And some of them are at odds with each other, and certainly at times you see what seem like really tragic choices made by people, because they were operating on a different … what we call a utility function for humans is incentives. Right? Because they were incentivized to solve a different problem than you are. And that can be very hard to relate to. You can see the stark contrast between the choices made by people living in extreme poverty versus you and I. They have different incentives, they have different threat models, they have different problems that they have to worry about.
Pablos Holman: So I think about those things a lot. I think most of the conjecture by humans about artificial intelligence is total bullshit and a waste of human activity, because there’s so much work left to do just figuring out what to do with human intelligence. And until we solve that, we’re not going to have any meaningful contribution to artificial intelligence anyway. So much more important to figure out how do you make humans perform better.
Kirill Eremenko: Yeah. And I think it’s incentive, but it’s also combined with belief system. Like you take Adolf Hitler, he had his own belief system. And you take Alice in the same position as him, who has similar incentives, but in the US or in a democratic world, they have a different belief system, and that really-
Pablos Holman: Oh yeah. But the belief system is part of what creates the incentives. Right?
Kirill Eremenko: True, true. Like self incentives. You incentivize yourself.
Pablos Holman: Yeah, yeah, absolutely.
Kirill Eremenko: Okay. Got you, got you. That ties in really well with … This is the part I loved. You know when you were doing your speech at DataScienceGo, I was thinking of maybe there’d be some questions that I want to ask you afterwards. And I was thinking, “This is ideal. The Maslow’s hierarchy of needs that I would love to talk to him about.” And then you had in your talk, and I realized that that was the first time that I actually encountered it in your TED talk. When I was listening to your TED talk. I encountered the concept of the Maslow’s hierarchy applied to technology and how it works.
Pablos Holman: Right.
Kirill Eremenko: I think this ties in really well. Can you tell us a bit about how you combine the two, tech and Maslow’s hierarchy of needs?
Pablos Holman: Well I think I used that because it’s a pretty generally accepted model from psychology for what humans need to feel fulfilled. And what they need to thrive. So he’s showing … What Maslow’s hierarchy … which is quite old now, but also quite famous. What Maslow’s Hierarchy shows, is that at the fundamental level you need things like food, and you need breathing, and you need shelter. These are basic fundamental needs. And you need sex. Or at least a species does.
Pablos Holman: And then at a slightly higher level, you need things like physical security, and you need some of these … You need a job or something. You need work. You need some of those things. And most people could relate to those things pretty well.
Pablos Holman: Most people in the west don’t have to worry about these kinds of needs. They’re largely solved. You’re born American, most of those needs are solved. You’re going to get food, you’re going to get shelter, you’re going to get enough heat to not freeze to death. We’ve pretty much figured out how to do that for everyone in America. There are certainly some folks on the fringe that don’t get all those things, but by and large, those are solved problems.
Pablos Holman: But Maslow’s hierarchy of needs keeps going, and the higher you get, you get into things like family, friendship, having a sense of community. And the point I’m trying to make is that when you think about how we’re using technology to meet people’s needs, well we’re using technology to meet all their needs on the bottom half of the pyramid. We use technology to feed people, we use technology for medical support and health, and we’re improving that every day. We have all these technologies now that we’re using to give you a healthier life, and to keep more people alive. To feed seven billion people is a big challenge, but you know what? We got there with new technologies. They’re not perfect, and the job’s not done, but we’re able to meet that challenge because of technology.
Pablos Holman: And I call that quantity of life problems, those are the things on the bottom half of the pyramid. We can keep more people alive, and keep them alive longer and things like that. But when you get to the top half of that pyramid, it’s things like a sense of community, sexual intimacy. Things like expressing creativity. Things like sense of purpose and feeling needed by the people around you. None of these things are things that technology is really helping solve right now. These are needs that humans have. And I think for me it explains … I see it when I look at the difference … because I get to travel all over the world. And when I see the difference between my lifestyle, which is arguably pretty affluent. I mean it’s not that I’m rich, it’s just that I’m American, and I live on the West Coast, and I work in tech, so I’m doing all right.
Pablos Holman: You look at that compared to people not even living in extreme poverty, but people basically living almost anywhere else in the world, I’ve got it pretty good. There’s a bit gap between my lifestyle and theirs.
Pablos Holman: But you know what? When I look at … There’s no gap, or maybe even a reverse gap when you look at, well, who’s happier? Right? Who’s got that sense of family? I mean I have a family, but it’s pretty small, and I don’t spend nearly as much time with them as my friends, because I live in a city, and that’s just not what people do anymore.
Pablos Holman: When you go to places in India, people are all surrounded by their family, and it’s pretty amazing. Those cultures really value family. And again, I love my family, but it’s not as big a part of my life as in lots of other cultures. And then you look around, sense of community. Well those people have a … and not all of them. But in lots of other cultures, and lots of other countries, and lots of other places where they may not have the socioeconomic, or call it economic status that I have by default, they’ve still got all the needs at the top of the pyramid, they’re doing at least as good as I am, if not better.
Pablos Holman: An example I used on stage, which is a little extreme, is these orphans that I met in Ethiopia. They live in an orphanage. They don’t have a Mercedes. They’re never going to. But I cannot make the case that I’m any happier than them. And I’m not unhappy. I’ve got it pretty good. But compared to those kids, they know exactly who needs them when they wake up in the morning.
Pablos Holman: So that contrast I think is important to understand, because those are the kinds of things technology is not helping us close that gap. We’re not solving those kinds of problems with technology. We’re not solving for the psychological wellbeing of a human. And I think it’s really important to understand this distinction, partly because Silicon Valley is often acting as if we’re going to solve every problem in the world with an app. We’re not, we’re not even trying. And I think people can feel that. You look around you and people still hurting.
Pablos Holman: So what I’m trying to show is for people who think that robots are going to come and take all the jobs, and there’s nothing left for humans to do, they’re absolutely wrong. The robots are going to come and take all the jobs that humans suck at, compared to a robot. But there’s a whole bunch of jobs, and a whole bunch of work that we’re not even getting to because it’s hard, and it doesn’t scale, and it’s slow, and it’s expensive. Work like nursing, teaching kids, taking care of elderly folks who don’t have grandkids to care for them. There’s a lot of things there that we’ve got to do. Work like helping people learn new skills if they do lose their job. Those are things that technology doesn’t solve for you. Those are things that humans have to solve, and we have to do that work ourselves, and figure out how to do it well, long before we can ever expect technology to help solve those problems.
Pablos Holman: So those are the points I was trying to make on stage. I think Maslow’s hierarchy is just hopefully a way of helping people see the difference between those two classes of problems. At least in my mind that’s how they are.
Kirill Eremenko: Yeah, yeah, totally agree. And I think if you look around, I think technology’s actually making those top layers of the Maslow’s pyramid, it’s making them worse from my perspective-
Pablos Holman: Sometimes.
Kirill Eremenko: Like if you look at people who are checking Instagram all the time, and comparing their life to other peoples lives. And most of the people adding filters, posting only the best moments, maybe not even the true representation of their lives, and they’re sharing as if everything is awesome. Then people are getting depressed and sad because my life is not as good as theirs, and then they go and turn and do the same. And so when somebody wakes up and the first thing they look at is their phone, the last thing they look at is their phone. On average in America, people check … or in the developed world, people check their phones 160 times per day. That is ridiculous. The attention span has gone down to the levels of a goldfish, 5 to 15 seconds.
Kirill Eremenko: In my mind, that is causing, even though more technologically advanced, and more opportunities in life, at the same time it’s causing less fulfilled life. How can you feel fulfilled if your attention span is 15 seconds. You can’t even have a-
Pablos Holman: Yeah. Certainly what you’re describing is … I’d say like a lot of people express that same frustration, and on an acute basis it is easy to see that. But I think there’s a couple things that get lost in that conversation at times. Which is that I got on email in 1982. By the early ’90s I was an email addict, before you’d even heard of the internet. And I was … literally I had to check my … Spam hadn’t even been invented yet.
Pablos Holman: I would check my email, go to lunch. Come home check my email, go out to dinner. Come home check my email, go to a movie. It was … I didn’t have any email, nothing was going on, but I just was addicted, I had to feel like, “Oh no, I might be missing something.” I had to check it to know. And there was no mobile internet, so I had to go home to check my email.
Pablos Holman: And then by the time we got Instant Messenger with ICQ in the later ’90s, you got addicted to that. With IRC, with chat. We got addicted to that, and it was hard to leave your computer. But everyone of those things follows an addiction curve. You overdo it at first, you lose a couple of weeks of your life to Instagram, checking it every 40 seconds. But then you eventually figure out, “Okay, this isn’t actually adding that much to my life.” And you start to figure out what’s an appropriate level of interaction. Most of us are not addicted to email anymore, because we got it under control. We’re not addicted to Instant Messaging anymore, because we’ve mostly got it under control. We text maybe a little bit too much sometimes, we Instagram a little bit too much sometimes, but eventually you build that immunity. You get it under control, and you figure out, what’s the level where it’s adding to my life, and what’s the level where it’s making my life worse?
Pablos Holman: And so that’s the job of a human, that’s just your job, that’s what you’ve got to do is you’ve got to get that under control. Same thing as like drinking too much Coca Cola, or all these other healthy lifestyle habits, it’s your problem. You’ve got to figure out how to get those under control.
Pablos Holman: So I think there’s a personal responsibility thing there that often gets lost in the conversation. Second, all the positives get thrown under the bus. Not thrown under the bus, but these technologies do bring something good to our lives, and it takes us about 15 seconds to go from, “Holy cow, that’s amazing,” to taking it for granted.
Pablos Holman: When I was a kid, I grew up in Alaska. And I was interested in computers. There was nobody for a thousand miles in any direction who was more interested in computers than me, or knew more about it than me that I could learn from. I literally had to go to the library and try and get them to get books about computers from mysterious places in the world I’d never heard of. Like Cupertino. Like I don’t know. So it was hard, and I was alone. And being a computer nerd isn’t particularly cool now, it really wasn’t cool then.
Pablos Holman: So you feel like an outsider, and you feel like an outcast. And lots of people feel that way growing up. Maybe if you’re into football you don’t feel that way, but for most people they feel kind of alone with their own set of interests. What’s amazing is, ever since probably the mid ’90s or so, with Usenet, and then beyond there, you can be any kind of fringe wacko, into weird shit, and you can go online and find a thousand other people who are into the same thing as you, and you will not feel alone.
Pablos Holman: And it’s not just deviant stuff. You could be interested in collecting Beanie Babies, or you could be interested in … I know people in America who got interested in cricket. Cricket is a uniquely not American … nobody in America plays cricket except for folks from India who came here to work for Microsoft, and they get together on weekends and play cricket I guess.
Pablos Holman: But the point is, whatever weird thing you’re into, or not weird, I don’t know, you can find a community online. It’s amazing. And everybody gets to do that, and I think in that sense it’s hard to feel as alone as was normal when I was a kid. You can go find that community that will be supportive. I see it all the time.
Pablos Holman: We have this company that is an apparel company that makes clothes in all sizes, not just normal sizes. So it goes up to 6XL, which is pretty big. And I met people from that community who helped us develop the sizing. And they’re pretty big people. And they come visit us, and I’ve never even seen people this big, much less a bunch of them at once, because they normally don’t leave the house. But you know what’s amazing is they have this sense of community online. I mean we found those communities, and they helped us develop the sizing. And they’re all in it together. One of them here is called The Fatshionistas. And there’s another one called the Pacific North West Fatitude. They’re groups that have found other people of the same kinds of problems shopping for clothes that they have, and the same kinds of social struggles that are unique to them, and it’s amazing. It’s really an amazing thing. That’s just another example.
Pablos Holman: But the point is, whatever your situation is, you can find a group to be a part of. So I think that’s one of the beautiful and amazing things that these technologies have brought us. And yeah, you can kind of … If what Instagram is showing you is a bunch of unreasonably beautiful people, well you know what? You can unsubscribe. You can unfollow them. And you can go follow whoever you want. And so again, I think of it as a personal responsibility thing. You don’t get to blame the technology if it’s brainwashing you. It’s your choice what to do with it. You can unfollow anybody you want, it’s your choice. If you don’t like the way Facebook is feeding you fake news, you don’t have to use Facebook every day. You can actually delete that app if you want.
Pablos Holman: A lot of people don’t realize, the most valuable part of Facebook is the events. Well there’s a Facebook events app you can load on your phone. Load that one, and then delete the Facebook app. Make your life better.
Kirill Eremenko: Very, very interesting. Very interesting. I’m going to ask you this question. It’s a bit playing devil’s advocate here. I agree definitely, personal responsibility. However, technology is advancing so much faster than we’re evolving. We still have the million year old brain that’s used to the savannas of Africa. That’s linear and local, technology is global and exponential.
Kirill Eremenko: And totally in terms of personal responsibility, 100%, but what are your thoughts on this, that social media, it’s proven that it invokes the same releases of dopamine, just as alcohol and gambling. And we have restrictions on age for both of those, alcohol and gambling, whereas any child at the age of 12 can get on social media, and they don’t have that sense of personal responsibility. It’s completely out of their control. And right away they’re getting these releases of dopamine every time they get a message, or they check their phone.
Kirill Eremenko: What are your thoughts on that? Should government restrict the access to social media, or even computer games, for protect citizens. Or it’s completely in our hands?
Pablos Holman: Well here’s the thing, and there’s a couple kind of pieces to this. One of them, my daughter’s 12. She’s had an iPad since they first came out. She’s had an iPhone for years. She’s basically immune to the blinging of her phone. Because it blings so much that there’s no dopamine left.
Pablos Holman: And it’s not that she doesn’t … she does use these things quite a bit. But she uses them to go find the things that she’s interested in. She doesn’t get that same dopamine hit that you and I got. Her social value is not measured by what she posts there.
Pablos Holman: So I think of it as, she’s growing up with those things, they’re normal to her. She’s developed a relationship with them, that’s more appropriate for her and her generation. I’m not saying that it couldn’t be better.
Pablos Holman: But alcohol is not regulated the same way everywhere. In lots of places in Europe it’s not the government’s job to regulate when your kids can drink alcohol. It’s legal. But families do that. They figure out what’s appropriate. Kids grow up with a more normal and comfortable relationship with alcohol. It’s not taboo. They don’t turn 21 and go crazy. Like, “I can finally drink all I want.” That’s what happens to American kids. And I’m not sure that that’s proven to be the best approach.
Pablos Holman: One of the big problems that we have with technology that you constantly see humans struggling with, is these notions that like, “This isn’t all positive, it’s making our kids have the attention span of a goldfish,” and these kinds of things. “We’ve got to stop it.”
Pablos Holman: And so what they do is they try to regulate these things too soon. I think regulatory, or some kind of approach to managing these things that can be addictive, or that can get out of control, that can be helpful. Whether it’s families doing it, or parents doing it, or governments doing it, or businesses doing it, yeah you want to do those things, but not until you understand the problem, and you understand what solution works for what your values are. For what you care about and the outcomes that you want.
Pablos Holman: And so my kids’ school has decided no electronics. She’s in middle school. So there’s no electronics for kids in her grade at school at all. I mean they do use computers and things, but the school provides them, you don’t get to bring your phone and be on Instagram during school. Which sounds fine to me. That’s a new policy, they used to have kids bring in electronics, and they just figured it started to get out of hand, and they decided that it would be better for the school if they just had no electronics. Which I’m fine with, because they figured out in the context what works for them. I didn’t need the government to do that. Her school chose to do that, other schools might choose different things, and I think that’s fine. That’s kind of the definition of diversity. Figure out what works for your community.
Pablos Holman: In her school the high school kids do have electronics. So as she gets a little older, she’ll probably be allowed to have electronics at school, but she’ll learn in that context that she has to be responsible for that, and not overdo it, and not do it at the wrong times, and all those kind of things. And what they figured out is by eighth or ninth grade, kids are ready to work on that problem. At sixth grade, they’ve got other things to worry about.
Pablos Holman: So that makes sense to do. In all cases though, what I believe is that we are not good at guessing what’s going to work. So when you’re talking about the implications of a new technology, one that’s just come out, you’ve got to get some experience with it. People seem to be unaware that when they signed up for Facebook they were probably going to end up with a whole bunch of propaganda shoved at them. Apparently everyone’s acting so surprised. I don’t know how they could have missed it, because I don’t know anybody who’s ever written a check to Facebook, but it’s the thing they use the most every day.
Pablos Holman: Same with Google. Obviously the deal was, “We’re going to give you all these amazing free services, and you’re going to pay us by watching some ads. That’s basically the deal. So acting shocked and appalled now that you’ve watched a bunch of ads that you’re upset about, seems a little disingenuous to me.
Pablos Holman: But the point is, you have to go through that cycle. We’ve been doing it long enough now that we’re starting to get some experience with the failure modes of social media. And if we can do that for a while, then we can eventually figure out, “Okay, what are the ways this goes wrong? What are different ways that we could mitigate that problem?” And if you’re prone to want a babysitter to take care of problems for you, then you might say, “Well, this should be made illegal. Or Facebook should be forced to solve this problem for me.” But it’s usually … never seems to occur to people that they could actually solve this probably themselves by not using Facebook.
Pablos Holman: So I don’t know. It’s also in America we have this weird situation where everyone thinks that Facebook threw the election, by letting Russians post fake news that swayed peoples … I don’t know a single person who thinks they voted for the wrong guy, or the wrong person for president, because of fake news. Everybody thinks that they voted for the right person, and everybody who voted for the other person must have been swayed by fake news. There’s not actually, as far as I can tell, anybody who voted for the wrong person in their own mind.
Pablos Holman: And so this is the kind of thing that … They’re making these unfounded logical leaps and saying, “Well, artificial intelligence is coming, and it wants to kill us all, so we should regulate it into oblivion.” And this is so ridiculously unfounded. We do not have artificial intelligence, it does not exist. We have no idea how long it’ll take for those breakthroughs to come and for us to get something that’s even like what we’re imagining. And we have no evidence that it’ll be malicious other than the fact that some humans are. And I think it’s just a ridiculous set of conversations to be having when we could be aiming ourselves at solving actual problems.
Kirill Eremenko: Well thank you. That’s very reassuring to hear coming from you, with all your expertise in entrepreneurship and AI. We’re getting close to the end of the podcast and I have to ask this question. Data. What role does data play in technology? We’ve been talking about technology a lot, and how it disrupts the world. There’s a lot of data scientists listening to this right now. What can you say to them in terms of what is the role of data in this whole new world that we’re entering?
Pablos Holman: Well you have a pretty smart audience that knows a lot. I guess what I’d say is, for a decade we’ve been at a point were we could start to use what we call big data. And that’s just that our computers got big enough and fast enough and had enough memory that they could handle a large amount of data.
Pablos Holman: But it also comes from sensors, we have almost the ability to create something that measures almost anything now. And every day we get new types of sensors. And so those sensors are collecting data, and then we have networks to bring all that data back to our giant supercomputers where we can analyze that data on a large scale.
Pablos Holman: All of this is unprecedented for humans. And so we’re in a situation where we get to … And when I keep saying new superpowers, well this is what I’m talking about. This is a superpower we get. We were evolved to make decisions using gut instinct. We were evolved, like you and I said, I think at lunch, to make decisions by looking around and saying, “Oh, is that a rock or a bear. Either way I should just run from it and decide later.” That’s gut instinct. That’s what we’re evolved to do. And we’re good at that, but it’s only good to a point, and it’s only good for certain types of decisions. And it was good for surviving bears and rocks, but now we have a superpower which can help us make better decisions. That’s what big data is, and that’s what all the tools that are coming to support it are.
Pablos Holman: For a decade we’ve been doing big data, which was basically like shoving billions of rows into excel. But now because machine learning is so computationally intensive, we haven’t really been able to do much with it until the last, call it five years. And so in this time frame we’ve finally gotten to the point where now it’s cost effective to run these arbitrarily complicated machine learning algorithms, and the output of things like neural nets, which are extraordinarily complex algorithms. But we can afford to run them on massive amounts of data in our models, because our computers are so fast, and so powerful, and so cheap.
Pablos Holman: So that’s the superpower. And then what humans need to do is use their brains to imagine, “Well, what could we do with it?” I mean I’m interested in working on embedding new technologies all the time. But if we didn’t invent a single new technology for the rest of our lives, machine learning is so powerful, there’s so many places we can go with it, that we can stay busy for the rest of our lives applying machine learning to everything.
Pablos Holman: And that’s what I believe. People are distracted by these ideas that machine learning is somehow leading us to an AGI, and somehow going to obsolete humans. I don’t believe that is true at all. There’s lots of work for humans to do, and I think machine learning is powerful enough that this is the community that can get ahead of it. It can take that and start imagining what we can do with it.
Pablos Holman: I tell you, we are not computationally constrained. The computers we have today make every computer I’ve had in my life a toy. And the computers you’re going to have in the next five or ten years, are going to make today’s super computers look like my first Sony. Like a little toy with giant buttons, that are in primary colors. You have a piece of shit computer right now, no matter how big your supercomputer is.
Pablos Holman: And we need to use our imagination to figure out, what are we going to do with all that computational ability? Well the toolkit the data scientists have is where it starts. And as you all well know, there’s lots of ways that it can go wrong. You can start with bad data. You could have … I mean anybody who’s worked in excel knows, if you screw up the formula in one cell, the whole spreadsheet can go south. Well, that’s the exact same kind of thing that can happen in any kind of scenario that you’re building using data.
Pablos Holman: So we need a human to think these things through, and to learn from experience, and design models that make sense, and create a market for data that makes sense to play in, so that people can get access to those things.
Pablos Holman: So I think it’s incredibly rich future for data scientists. I think there’s so much to do that’s important. And I couldn’t be more thrilled about the way things are going, and I think you guys are all lucky to be a part of it. We’re in a renaissance you guys, and you on the front lines are getting to be part of it.
Kirill Eremenko: Fantastic. Thanks Pablos, that’s amazing wrap up. Just one final question. You were at DataScienceGo, you saw what it’s like, what would you say to people who are … Because we have DataScienceGo 2019 coming up also at the end of September. What would you say to those that are listening to this podcast and on the fence, they’re not sure should they go, should they not go. What would your message-
Pablos Holman: Oh. Well like I said at the beginning, this is a pretty special community, a pretty unique event. I get to go to lots and lots of different events in tech, and even outside of tech, all over the world, and I can’t imagine a better opportunity. You can just go to this event and you will feel like part of something, and you’ll learn a lot, and you’ll meet other people who are excited, and it’s a pretty special opportunity I’d say.
Pablos Holman: So definitely get there.
Kirill Eremenko: Thank you. Thank you Pablos, really appreciate you taking some time to come on the show, and I’m sure people will get value out of this. Thank you so much.
Pablos Holman: Yeah, my pleasure. Thank you.
Kirill Eremenko: So there you have it ladies and gentlemen. That was the famous hacker, inventor and entrepreneur, Pablos Holman. I hope you enjoyed this chat as much as I did, and got lots of key insights and takeaways. And of course there are plenty of more topics that we wanted to cover off, but we’re limited on time. So if you’d like to get more of Pablos’ inspiration and ideas then I highly recommend his TED talks, which you can find all over the internet, and have 20 million plus views. And also his keynote speech at the DataScienceGo conference, which you can find on the SuperDataScience YouTube channel. Available there for you to watch. You’ll feel as if you were at DataScienceGo. We were all sitting on the edges of our seats when he was delivering his talk. Highly, highly recommend that to you.
Kirill Eremenko: And speaking of DataScienceGo, as you heard from Pablos himself, this is a non typical … this is not your standard Data Science conference, this is something we’re building. It’s a community we’re putting together. This year is happening in San Diego on the 27th, 28th and 29th of September. So check it out. Make sure if you can come that you do come. Because you want to be part of something bigger, you want to network with people, you want to meet people like Pablos Holman, and lots of our speakers. We’re going to have over 30 speakers at the DataScienceGo event this year. So make sure you are there.
Kirill Eremenko: You can find your tickets at www.datasciencego.com. You can find all the line up of speakers, the agenda, and lots, and lots of other things that you will experience at the event, and I personally would love to meet you there.
Kirill Eremenko: So head on over to www.datasciencego.com and get your ticket today. And I look forward to seeing you back here on the SuperDataScience podcast next time. And until then, happy analyzing.
Show All

Share on

Related Podcasts