Big tech giants like Amazon, Google, Meta and the tech industry as whole have experienced tremendous growth over the past years. However, the recent dip in their stock prices, disappointing results, and challenges they face have led to concerns about the sustainability of their dominance. Amazon has seen a decline in its operating income and has come under fire for its treatment of workers, resulting in calls for increased regulation. Meanwhile, Google has faced increased competition in its advertising business and regulatory scrutiny over its privacy practices, with a lawsuit by the US threatening to break up its ad-technology business. Meta has suffered from a decline in user engagement and the company's latest attempt at creating a virtual reality platform, the Metaverse, has also failed to gain traction, despite billions of dollars spent on development. Layoffs are now widespread and it seems the industry is stagnating.
Why has this happened? Besides the current poor economic outlook, there might be a long-run trend here. It could be argued that big tech companies haven’t been innovating as much as they used to because they are out of ideas and we are now seeing the effects. As early as 2010, venture capitalist Peter Thiel argued that a bet on Google is a bet against technological innovation in search and that other big tech firms aren’t really ‘technology' firms, because they sit on large amounts of cash instead of reinvesting a large share of their earnings into the development of new technology. In the meantime, little has changed, with Google currently sitting on more than 100 billion US dollars in cash. Why don’t they spend it? Thiel claims these companies have simply run out of ideas.
Computing and the End of Moore’s Law
An interesting thesis, but does it hold any water? To investigate, we should find ways to measure innovation and technological progress. A recent declaration by Nvidia's CEO, Jensen Huang, that Moore's law is dead, sheds some light on the current rate of progress in computing. Moore’s law states the number of transistors on a chip will double roughly every two years. Since a lot of tech companies rely on advances in computing, the end of Moore’s law could mean a slowdown in innovation.
Sure, maybe we are reaching the end of progress in classical computing, but we will think of something else. After all, that is what tech companies are for, right? Some are hopeful about developments in quantum computing. However, as of today only a handful of quantum algorithms have even been developed. Furthermore, the quantum computer is a very niche machine, only useful for a small subset of calculations, and no fault-tolerant quantum computer has ever been built. Regardless, quantum computers are not a new idea by any means, having first been proposed in the 80s.
Artificial intelligence has seen a resurgence in the past decade that shows no signs of slowing down, especially in light of the current ChatGPT craze. However, it must be noted that developments in AI have historically relied not on radically better models, but on more powerful computers. Training a large neural network is a process that consumes a tremendous amount of energy and computing power. Despite this, artificial neural networks, which date from the 1940s, are still the most popular approach.
But perhaps the problem lies a level deeper. Are we running out of ideas in the sciences?
It would not be controversial to say that progress in fundamental physics has stalled in the last 40 years or so. However, stagnation is not limited to physics. There is a sense of stagnation in other fields too, as researchers are finding it increasingly harder to come up with new ideas. Economist Ben Jones believes the reason for this lies in what he calls the ‘burden of knowledge’. As knowledge accumulates over time, each generation has to spend more and more time learning it. This causes new scientists to waste their best years not making groundbreaking discoveries, but learning what came before. Furthermore, the increase in knowledge requires deeper specialisation, which can make it harder to find new insights that combine knowledge of different fields of science.
Regulation and Government Funding
Another reason may be the increasing push for greater regulation and scrutiny of tech companies, in light of privacy concerns and the growing power of these companies. Governments around the world are grappling with how to balance the benefits of these companies with the potential harm they can cause.
This brings me to the issue of government policy. The US government alone has in the past been responsible for funding the development of many of the technologies we use everyday. The Manhattan project gave us the digital computer. The military developed GPS and what would eventually become the internet. Space exploration has yielded a wide variety of technology, from better solar panels to digital image processing used in MRI machines. The role that governments play in funding research cannot be understated.
True, basic research funding in the US has continuously been rising. But deeper investigation reveals a much less optimistic story. The government’s share of funding has been decreasing and, as of today, less than half of basic research is government funded. Most of the rest comes from corporations, which are more risk averse and short-sighted. To see why this results in stagnation, we should take a look at the role basic research has played in scientific and technological breakthroughs. Mathematical theorems sometimes go unnoticed for decades or more before someone discovers them to be crucial tools in solving certain scientific problems and, by extension, creating new technologies. For instance, non-Euclidean geometry had been considered useless for almost a century, until Einstein showed that space-time can be non-Euclidean in his theory of general relativity. Without this discovery, GPS would not exist and you wouldn’t be able to stalk your friends on Snapchat.
Should we want more innovation?
Now, one might argue that, even if innovation is slowing down, there is no problem at all. That to avoid destroying the planet we need less growth, not more. Additionally, the tendency towards monopoly, concerns over privacy and the tremendous influence some tech firms have on society begs the question: Should we want more innovation in the first place?
However, technological progress has often meant that we could do more with less. In the past we have instead opted to use our technology to, equivalently, do much more with more resources. But this is just a matter of choice. In fact, I would argue technological progress to be indispensable in solving our environmental problems. We just have to be wise in harnessing it.
Founders Fund essay - Peter Thiel (2010)
Burden of Knoweldge - Ben Jones (Northwestern Univ) (2009)
Stagnation hypothesis - Why I came around (2021)