It’s been 50 years since Gordon Moore published his paper (dated 19 April 1965) detailing a simple theory that revolutionised computing, making all the technology we see as fundamental to life today possible.
Moore predicted that the number of transistors on a computer chip would double every year. Not only was he correct, but remarkably, this exponential growth trend has continued until this day.
Ahead of this important anniversary, WIRED.co.uk spoke to Intel futurist, Steve Brown, about how Moore’s Law will continue to impact society through alternative future technologies, and some of the challenges it may face as we look to the next 50 years.
WIRED.co.uk: What’s going to be the biggest impact of Moore’s Law over the coming years?
Steve Brown:I wouldn’t say it’s anything specific like the internet two or anything like that. I would say the biggest impact is going to be how pervasive computing and smart becomes.
Initially in the very early days of computing before Moore’s Law took hold, a computer was something that filled a huge room and was the province of academics and researchers. You’d go into a room and bend over and worship in front of a giant machine and now it is something that is available to billions of people and it’s our pockets and purses. The next logical step of Moore’s Law is that it just becomes part of everything. If I had to articulate it as what’s the big thing, it would be just that computing merges into our world and dissolves into it and becomes invisible. It’ll be in our walls and in every object and it will just be part our expectations of every object — it will be smart and connected.
What alternative technologies do you see developing as part of this pervasive computing?
I think the things that are most compelling are figuring out how to deal with the challenges of the 21st and 22nd centuries. Mass overcrowding, people trying to live together harmoniously in cities of 20, 30, 40 million people, trying to feed everybody, trying to get everybody enough water. I think those are the big challenges we face as a society and Moore’s Law will be a key tool to helping solve those.
So for example if you’re in a very busy city, it doesn’t makes sense for everyone to own a car. Having self-driving cars, that are innately more shareable because I can just summon one on my phone and it will arrive and pick me up and take me where I need to go and drop me off and go and pick up somebody else. That is going to be a key part of any transportation solution for a future metropolis.
Feeding people on the planet — we’re going to have to get much more efficient with the way we do agriculture. At the moment, the way that farmers water their crops is they spray water near indiscriminately in the air and it lands on the crops. Some plants are getting more water than they need and some are getting less than they need.
Being able to till tiny computer elements into the soil so that you can instrument, the soil, the field can tell you basically what it needs. Each plant can tell you, I need a bit more water or I need a bit more fertiliser and then to have your robot watering units tap into that […] so you’re much more accurate and much more granular in the way you’re thinking about solving problems. You could use a lot less water, a lot less fertiliser, waste a lot less, have a lot less run off. So that would be an example of something very futuristic that solves a very serious human problem.
Will this smart, pervasive computing reduce the emphasis on actually owning technology?
I think we are already seeing people, particularly millennials, move from embracing the notion of ownership to much more thinking about the access economy or the shared economy. I think that it is already happening partly because of awareness, because of wanting to use resources and wanting to live more sustainable lives, but technology absolutely helps with that. As things become digital, enabled by Moore’s Law, whether that is music or video or books they are innately more sharable. Spotify is made possible by Moore’s Law, Netflix is made possible by Moore’s Law, so that’s how things can be shareable. But also when you start to put intelligence inside them, that also makes objects more shareable
The self-driving car is a great example of that because I can now start to use the intelligence in the car to act as like a gateway. If I pay my bills, I can use this thing. What that does is it turns products from products into services, and that innately fuels this access economy, so yes I think in some ways the future of Moore’s Law is to reduce ownership.
How will this transform urban living in the future?
In almost every way. When you have people living in very tight confines, you’re going to have to have good transport policy and that’s going to mean using technology to improve transportation networks. Not just self-driving cars, but having transport policies that link self-driving cars with autonomous public transport, with other modes of transport, linking them together to make it really efficient way to move people from A to B.
I think you’re going to see retail outlets getting smaller as you crush more and more people into a small space and theyre all going to be living together in close proximity. Technology is going to be there to create a big store feel in a small store environment, so virtual technology to enable you to have more choice and to have digital delivery — to have next-hour delivery instead of next-day delivery. Technology is going to help with all of that, whether its supply chain, or whether its small virtual stores, or other types of experiential stuff.
Will it be able to help solve housing problems and overcrowding in cities?
That’s another area where technology can find application. There’s a growing small homes movement around the world, particularly in the US, with people living quite happily in 200-300 square feet — that sort of size. They’re doing that because they’re using very careful design, but they’re also starting to think about using technology to make their space flexible so they can easily transform. There’s some interesting robot furniture thats been posited by some universities, where the furniture actually changes its shape.
The possibilities of Moore’s Law are enormous and when you can start to make any object in our lives smart and sensing and connected to the cloud, suddenly you change the nature of that object forever and it’s going to help us in all kinds of ways.
There’s been a lot of discussion about how progress in technology will impact jobs — will Moore’s Law ultimately be responsible for people losing their jobs to robots?
Thats a very valid question and one we should all be concerned about. So a couple of comments, the first one is technology has always replaced jobs and probably always will, but typically it replaces jobs that aren’t really much fun for the human beings doing them. It tends to be the lower value jobs, so think about agriculture 200 years ago, you were dragging a piece of rusty metal through the soil. And if you did that it was back-breaking work, ten or twelve hours a day and you were lucky if you lived to 40.
Now that’s all been mechanised. The farmer sits in an airconditioned cabin bouncing around on top the combine harvester and life is good. And all the people that used to till the land are now web designers and work in retail and making lattes. Life is better for those people, and I think it’s always going to continue.
Jobs will be replaced. If self-driving cars truly become a reality, then taxi driving is going to be gone 20 years from. It’s always going happen, the question is are we going to be removing jobs at the same pace we are creating new jobs because of Moore’s Law and having new technological developments. That is the big open question.
Has Moore’s Law defined the generations that have lived through the progress it has made possible, or has it only continued at the pace it has because of human intervention?
Moore’s Law would not have happened without people. The great thing about Moore’s Law is it is not a law, it was an observation that Gordon made that the people at Intel decided to make into a law and its become their north star. They come into work everyday thinking how do we outdo ourselves and it’s become this call to action. It’s about belief more than anything and so fundamentally Moore’s Law is driven by human belief and human ingenuity and people pushing against laws of physics. And so yes Moore’s Law is fundamentally shaped and driven by people, but you’re right — and this is what makes it so deep and interesting — Moore’s Law has shaped us all. The modern lives that we live would not be possible without the advancements of Moore’s Law.
When it comes to wearable technology, the main problem is not producing chips that are small enough, but producing batteries that can keep up with them. Does this somewhat negate some of the progress of Moore’s Law?
Moore’s Law has been great at doubling the transistors every two years — it’s an amazing exponential law in computing, but batteries have not kept up. Batteries tend to improve seven percent a year and so when you’re seeing a linear improvement with batteries and an exponential improvement in computing it makes life challenging.
There do look to be some nice potential breakthroughs in the world of batteries though — some of them using peptides, some of them using aluminium and carbon batteries. I think with batteries, you get these slight incrementals and and then you get big breakthroughs once in a while. But the job of Moore’s Law is to make these transistors ever more efficient so that we can continue to deliver more performance and more capability, but with the same battery life and the same battery.
There’s been a lot of focus on the capabilities of Edison and Curie over the past few years, but is the main achievement in Moore’s Law really just about making devices smaller and smaller?
Making transistors smaller doesn’t just give you smaller devices, it gives the designers a choice. When you make transistors half the size, you have a choice. You can either take last year’s chip and make it physically half the size, which makes it half the cost to manufacture, or you can use the same size chip and pack in twice as many transistors, which gives you more capability and more graphics and all that stuff. That’s why over the last 50 years of ml, you’ve seen incredibly small chips and devices made possible, but you’ve also see incredibly large and powerful chips which are powering the internet.
Many of the chips that are running the datacentres at Google and Facebook and Netflix and all those companies, those chips have billions of transistors in them — up to 8 billion transistors now. So it doesn’t just mean you have smaller and smaller devices, it also means you have more and more capability and performance. That’s a choice that designers have and that’s why you see from Intel we have our Xeon and Xeon five chips that the very high end that go into our supercomputers and then we have things at the very low end like Edison and Curie that are going to go into wearables and other such things.
If there was one thing that was going to come along and disrupt Moore’s Law over the next few years, what would it be?
It still keeps going. All the time that I’ve been at Intel we’ve been hearing Moore’s Law is going to die in ten years and it still keeps going. To some extent it’s like walking down a foggy road — you can see a certain amount and as you keep walking down the road you can see further and further, but you only have a certain amount of visibility. The same is true of Moore’s Law. We’ve always had about ten years visibility and we’ve always thought after ten years it’s going to get difficult. We don’t know we can go any further and then we go another couple of generations and we go okay, we can see a path there.
Now you can’t keep cutting something in half indefinitely. It’s going to inevitably change, but I think we are always going to be able to find new ways to innovate. Whether it is using a different type of technology or different materials, there will be new ways to continue to provide more and more computing abilities in a smaller and smaller space.