The phone you are reading this article on is way more powerful than the computer systems used to launch the Apollo spacecraft to the moon in 1969.
As we move into an age of connecting our brains to computers, sitting in a car that’s driving itself, or selecting the genes we want our children to have, we’re going to need even more powerful computers.
Who could reimagine entirely new chip designs and operating systems to make this happen? I think African researchers could change the game on this front.
The things our devices are able to do — recognize our voices, turn our faces into emojis, recommend music — require enormous amounts of computing power. This is taking longer to do as the cost of increasing the density on chips goes up. Researchers are working on new chip designs that continue exponential increases in computer power, speed, and cost-efficiency, but a lot of these try to stick to the current way our operating systems are structured.
Moore’s Law has explained the phenomena of rapid advances in computing speed for over 50 years. Gordon Moore, one of the founders of Intel, observed back in the mid-1960s that the number of transistors per square inch on computer chips were doubling every year while the cost reduced by half. This industry took this observation as a law that has gone on to propel chip makers like Intel, AMD, Samsung, and more to keep pace.
To give a sense of the scale we’re talking about – the chips used on the Apollo space ship had a few thousand transistors on them, while the chips in our phones have tens of billions of transistors. As a result, gamers have an incredible experience playing the new Madden and Facebook tags the photo you just posted faster than you can remember the names of the people in the picture.
Moore observed the doubling of transistors on chips taking 12 months. That’s now happening at around a three-year clip. Developing computer chips is an expensive business. For example, it can take $10 billion to construct one plant. Further, the energy required to run the algorithms that power the machine vision algorithms tagging your Facebook photos is a lot, and expensive. Not many organizations can afford these costs at scale — Amazon, Facebook, Baidu, Microsoft, Google, Alibaba, Tencent, and a few more. Having so few companies able to afford running complex algorithms puts a cap on new discoveries and could further widen the wealth gap.
The semiconductor industry has been snapping its development to the pace of Moore’s observation. What’s important to remember is that he was observing something that was already happening organically. Perhaps the time is ripe for a new way of thinking about chip development that isn’t wed to the way things have been done until now.
Today, a lot of African countries are dealing with power constraints and business environments where the development of a $10-billion manufacturing plant is tough to imagine. Tyler Perry and Angela Rye did a fireside chat at Essence Fest where Perry made the observation many have experienced, “Black people know how to take a little and make it go a long way.” Engineering researchers still working in African countries or who have made their way to labs in other parts of the world could bring an understanding of these constraints to bear in reimagining chip designs.
Ideally, this reimagining would take place in Africa. A lot needs to happen for that to become a reality. For one, African countries would need to make the decision to invest heavily in research and development. Kenya has committed to investing 2 percent of its GDP, roughly $1.5 billion in R&D. South Africa invested roughly $350 million in technology R&D in 2015. For this investment to actually move the needle in this space, it can’t be nominal. The constraints in these countries should be an innovation enabler, not a crutch to continue underinvesting in building the future.
I see three benefits from African countries jumping into developing the next wave of chip technology. For one, we would have Africa repositioning itself to be at the center of AI development. Secondly, we would see Africa finding a path around the persistent energy challenges across the continent. Perhaps developing game-changing technology on this front opens doors for rethinking how much energy is actually needed for high-level development on the continent. Third, Africa could build some long-term resilience against falling victim to the battle China and the U.S. are waging on multiple fronts including in AI development.
Changing the game in chip development isn’t going to take place over a short period. Research and development for the current slate of computer chips takes over a decade before we see the technology in our devices. Disciplining ourselves to take the long view and focus on seeing around the corner will be difficult but will be more than worth it for the world and the role Black folks play in it.
In his 2005 book, “The Singularity is Near”, Ray Kurzweil theorizes that technology will lead to the exponential growth of human intelligence, reaching a point where we are capable of superhuman cognition. In Singularity, human minds merge with AI and transcend biology.
If Kurzweil is right and we hit the Singularity in 2045, the investment and discipline would be well worth it.
Kwame Som-Pimpong leverages relentless research, a knack for connecting dots, human-centered design approach, and effective communications strategy to help organizations realize their strategic objectives. Over a 10-year career, Kwame has supercharged grassroots political organizing efforts, assessed the effectiveness of U.S. federal agencies, managed an international program, founded a digital media startup, and advised government agencies on delighting their end-users. He earned a BA in Political Science from Davidson College and Master of Public Administration from the University of Georgia.