After the last thirty years of working in technology, innovation in these last few years seems a tad underwhelming.

Is it because governments are starting, for the first time, to seriously consider regulating social media and other IT companies? At least some authors think serious regulation will end the rapid innovation of previous years.  

This doesn’t seem like the right answer — information technology, and social media in particular, remains one of the most unregulated markets on Earth. The understanding, based on Section 230, that social media companies are passive platforms when they publish user content, and yet have all the rights of a publisher when filtering content, is still the “law of the land.” Very few information technology companies are taken down a notch for their working conditions, the financial methods, or their obvious support for one political viewpoint.

Maybe it is because of a lack of talent? Numerous articles, like this one, and this one, talk about the lack of coding skills. Network and security engineering are both facing serious shortages of engineers — and the pipeline into both fields is more of a trickle than a flood.

It is hard to determine what is cause and what is effect in engineering talent, though. Are people not entering critical engineering fields because they do not know about them? Or are these fields perceived as boring? I know network engineering is widely considered boring. It’s hard to attract new talent to a boring career field.

This answer doesn’t seem right, either. If talent is the problem, why is Apple spending lots of money protecting their brand in ever-deeper corners of the Internet rather than focusing on building new pools of talent?

Or maybe this is all just an optical illusion. There is real innovation, but we cannot see it because of the effects of Keith’s Law:

All major innovations are the result of many smaller innovations hidden behind a single result.

Maybe there is a lot of innovation out there, but we are not seeing it until some new product is released. In ten years, it is possible we will look back at these years as a time of small improvements that will not be revolutionary until they are all put together in a single product.

This might be a part of the answer, but it still does not seem like the whole thing. Despite the triumph of Large Language Model (LLM) natural language AI systems, there do not seem to be any new “things” on the market that compare favorably to the invention of steam locomotion or electricity. The overall “invention count” of these last few decades just does not seem to be as grand or life changing as the list of inventions from the late 1800s and early 1900s.

Another possible explanation is tech companies are running out of the raw materials of revolution. Going back to the beginning of the tech revolution, the primary driver has always been acquiring new data. Each advance has either been a new way to collect or process data.

Social media networks began by scouring users’ posts within their service. Search engines began by consuming the entire Internet. LLMs and modern AI consume every work of every human in all of history. The Internet of Things (IoT) consumes data from every car, thermostat, industrial system, and everything else in existence.

If it can be measured, it is either already being measured, or someone is spending time and money figuring out how to measure it. If it produces data, someone is already processing that data to create new data. We are, in fact, close to the point where the velocity of new data is overcome by the velocity of data about the data we already have. We are coming close to knowing more about what we know than knowing about the world.

If this is true — and it certainly appears to be — it only makes sense that the branch of thought that deals with information — information technology — should be slowing down.

But rather than seeing IT as boring, perhaps a better way to think about it is IT is becoming stable. Rather than everyone in the world of IT running as fast as they can on a treadmill, maybe it is time for these technologies to stabilize a little and focus on quality and useability rather than trying to make all the new things all the time.

Stability does not mean boring. It just means taking a break to assimilate and understand instead of rushing to the next big thing before we have time to think about the latest thing we were rushing towards just yesterday.

Maybe IT is slowing down. But maybe that is a good thing.