The future is exponential | Pascal Finette

Now, you hear us talk about this idea of like the future being exponential, and it's kind of like a weird word, so let me explain. If you go back to like your Math 101 class, a classic exponential trend looks like this. It's a doubling every time period. So you go from 1, 2, 4, 8, 16, 32. When you draw it, it looks like the hockey stick. This is the classic curve every Silicon Valley investor wants to see when it comes to your revenue. Or user growth. But that's an exponential class curve. The most important technology exponential, which you surely have heard of, is Moore's Law. So this is Gordon Moore. Gordon Moore is the co-founder of Intel. And Gordon Moore wrote 50 years ago, wrote an article for IEEE magazine where he described— interesting enough, he described a trend he saw in the past. So he wrote in this article that What he observed is that for the last 10-ish years, semiconductors had a doubling in transistors per square inch on an integrated circuit, and then he made the bold prediction to say, I believe this continues into the future. Now, this held up for 50 years, and it became known as Moore's Law. So what this simply means is that the thing in your pocket, the thing on your lap, becomes more powerful effectively by the day. To give you an idea of how this looks like, this is a computer called ASCIRED. It was invented in, or was built in 1997, cost us $55 million. This computer was at Sandia National Lab, which is a big national lab here in the US. It was used to do atomic weapons testing. This computer, $55 million, the size of a court, Like a tennis court. And it's remarkable because it was the first computer to break through the 1 teraflop barrier. So 1 trillion floating point operations, mathematical operations per second. So massive, massive breakthrough. This was, for a really long time, this was like the pinnacle of computing. 9 years later, Sony introduces the PlayStation 3, $499, 2.1 teraflops. So this is what computing does to you, okay? It goes a different way. You take the Pi Zero, it was released December last year, costs you $5. So this is a full-scale computer, you can do pretty much anything with it, you can run a proper operating system with it, plug it into a monitor. $5. It has 191 megaflops, so 191 million floating-point operations per second. So for the price of a Starbucks latte, you can buy the compute power of 2.5 Cray-1 supercomputers. The Cray-1 was released in 1975. Each one of those Cray-1s had more compute power than we had to put total to put the man on the moon. So now for the price of a Starbucks latte, you get 2.5 times the compute power from NASA had total to put the man on the moon. This is where computing goes, okay? More importantly, Is this. This is a golf ball. You see the dimple in the golf ball. You see a little chip. This is a Freescale chip. It was already released in 2014, so it's fairly old. It's 2 millimeters by 1.6 millimeters. This chip has the raw compute power of an early-stage Pentium processor and costs you 75 cents. This chip is single-handedly, and all its brothers and sisters, is single-handedly responsible why everything, everything, which has an electric cord attached to it will become smart. Because you get it for free. Like, why wouldn't you? I get asked this all the time, like, why would you want to have a smart toaster? That doesn't make sense. And God knows I don't want to have a smart toaster. But I get it for free, so why wouldn't I? This is the reason why every light bulb will become smart. Also, this thing runs on the battery of a watch for about 2 years. So it's extremely energy efficient, which is the reason why anything which doesn't have an electric cord will become smart. So computing becomes incredibly powerful and incredibly small and cheap. And this leads to really interesting outcomes in this world. But it's not just computing. So computing is a really, really strong trend, but it's bigger than computing. So you take something like DNA sequencing, which by the way, If you're following this a little bit, this is going to be the biggest change to humanity we have ever seen outside of computing. So you take cost of DNA sequencing, reading a genome. In 1999, we did this the first time as part of the Human Genome Project. $2.7 billion, 7 years of effort. Total effort, 20 years. So 7 years of core effort, $2.7 billion. Nation states are the only ones who can do this. In 2007, we did this the first time commercially. $350,000, took us a couple of months. In 2014, that price dropped to $1,000, and it takes us an hour. A company out of San Francisco called Illumina brought out a machine which does this. This is by far the quickest drop we've ever seen in any technology in terms of price-performance ratio. Now, the interesting question becomes, If you look at the cost per genome and you look at the trend curve, I mean, it's a very clear, very dramatic drop. The question becomes, what do you do with it? And where does it go? Right? So as an entrepreneur, like, we're at $1,000. $1,000 to read a single genome is still kind of expensive. So where does it go? Well, you ask the experts and they will tell you that price drops to close to zero. reading the genome will become close to free. And then you can ask yourself, what do you do with it? Well, one thing is, every single time you flush the toilet, the toilet will sequence your genome and give you a full health report. This is not science fiction. There's a company working on this. Or, a little bit more on the sinister side, every single one of you is currently shedding cells. You're losing hair cells and skin cells and follicles, etc. The moment you leave this room, my crew in the back and I will hover up and we generate a full genomic fingerprint of every single person in this room. So that sounds a little scary. If you don't believe me, I want to introduce you to a friend of mine. This is Heather Dewey Haggard. She's an artist in New York, and she has an interesting story to tell. DNA might be used against us in the future. New York artist Heather Dewey Haggard— Heather Dewey Haggard— One artist in New York is making 3D models of people's faces, people that she's never met. She calls the project Stranger Visions. The strangers are people whose genetic material she finds on the sidewalks and subways of New York City. How much can I actually find out about you from something that you accidentally leave behind? So you see where the future is going. This is an artist. This is not like a secret national-funded lab. Okay, so interesting stuff. This is how you do this. This is a company out of our portfolio that developed a web-based tool which allows you to do genomic editing. So basically, you click together your GTACs and you save them. You send them to a printer. It creates a process similar to laser printing. It prints it. You put it into a blank cell. You grow your organism. This software was used to do something called the Glowing Plant, which was the largest— one of the largest Kickstarter projects. They raised more than $1 million. They took the genomic information from the firefly, so the information which makes a firefly glow, and put that into a plant and created a plant, a flower, which glows in the dark. Their big game is they want to put this into trees so you can replace streetlights because the trees glow in the dark. So we're moving rapidly into a new world. Let me show you one more example where we see an interesting drop here. So cost of solar energy. So first time we did this, or not the first time, but like the first kind of commercial number. In the late '70s, $80 per kilowatt-hour. That's terrible. It's really, really expensive. In 1987, 10 years after that, the price already dropped by a factor of 10 to $10. In 2015, in California, the price dropped to $0.30. So now we have parity with coal. So energy is now the same price as coal. In 2016, in Dubai, the price dropped to $0.03. This is unsubsidized. So energy becomes free, like literally free. Now you can ask yourself again the big question, like what do you do with free energy? Like a lot of things. If you plot these things out, you see 2 effectively exponential graphs. The one is the dropping price, the other one is the installation of solar panels. There's something called Swanson's Law, which is similar to Moore's Law, which stipulates that with every doubling in solar panels, the price drops by 20%. It has something to do with economies of scale. So now I have a world where once we have reached the tipping point, which we have, solar becomes cheaper and cheaper and cheaper and cheaper, so it gets close to zero. A lot of people in the energy industry— so the guy who runs the largest energy company in Europe is a friend of mine— he will tell you that soon you will not pay for energy anymore. Which is interesting because if you think about it, we're fighting wars over energy at the moment. We send people to to die to fight over scarce resources, whereas in reality, like, these resources are abundantly available. If we were to build a solar panel that size, that would power the world over, notwithstanding that you need to transport the energy, that's a whole different story, but that is all it takes to power our current energy consumption. Thank you.