The need for compromise in innovation communication in the age of AI

Innovation might not be as slow as it seems, and companies need to do a better job of telling us that.

Sep 11, 2025

Pre-2020s, we used to see >2-3x jumps in innovation around performance and capability in, as well as significant design changes to, personal computing devices.

You'd hear things like chips getting 3x better, or cameras being 4x bigger. You'd notice that each iteration of a particular device would have such different and interesting designs and functionality. That era, particularly the 2010s, with companies like Apple and Samsung pushing the bounds in these regards, had the vastest array of design choices that made every new launch exciting. There was always something to look out for.

One of the things this level, breadth and pace of innovation brought was zero tolerance for meagre, marginal, iterative YoY updates to devices.

As the physical limit to chip technologies began to come close to getting hit, particularly towards the end of 2010s, when companies started realizing that Moore's Law was 'dying', a lot of companies tried to use marketing ploys to sell next-to-negligible changes to their products as more massive than they really were. At first, there was a lot of push back, but at some point, they realized how much they could get away with giving the barest minimum.

I am writing this mostly because of the latest Apple event. Even though, in my opinion, these companies are honestly still pushing the bounds of innovation - take a good look at how Apple has managed to squeeze all the components of the iPhone Air to the top section, making you think to future devices that would take advantage of that level of compactness - they seem more like they are pushing the bounds of how much they can charge without you flinching/reacting. They're banking on our docility.

Another company that is kind of guilty of this is Nvidia. It seems like they are heavily prioritizing AI software innovation and how much frames you can generate, rather than taking significant considerations in raw, rasterized compute performance. Don’t get me wrong, this counts. They just aren’t properly communicating that, and the consumers are furious.

The only thing I'll give to these companies is that truthfully, Moore's Law is dying; we're reaching angstrom lengths - this is very much undeniable. The problem is that the consumer's perspective on computing innovation and advancement has mostly been whole-number-multiples of jumps in performance, or eye-catching design changes, of hardware (components) rather than software.

This is one of the reasons why Nothing has received so much love: because they understand that the consumer-PC fandom wants - craves - hardware (design) innovation.

Another challenge is that these companies might not be able to use these excuses at some point. Not because people will riot—like I said, docility—but due to how quickly AI innovation is speeding up. I read somewhere—think it was a video I saw it, and please don't quote me on this as I don't have very exact figures or proportions - but chip innovation is moving 2x slower than AI is.

(Gen)AI models are seeing massive upgrades multiple times within one year, while traditionally, chips, according to Moore's Law at least, see that same level of improvement over a number of years. That's unsustainable. Unless we start looking into dramatic pivots in this area—looking at photonic computing as the most likely branch of this timeline, as well as some promising materials to replace silicon—and despite some roadmaps I have seen online that seem to postulate that we might still be able to squeeze out some things out of the current chipset architecture we use currently, I sincerely doubt chips will be able to keep up, and that is bad, to say the least.

I think there should be a compromise of acceptance from the two ends of this isle - the consumers and the technology providers.

The former should accept that innovation is hard - harder than they might realize considering that we're literally at the very edge of physical laws here and scientists across the world are doing all they can to make the most of what we have now while looking at possibilities. We have honestly been spoiled or, have pigeon-holed progress as one-dimensional.

This is in no way excusing some companies that are despicable in handling this issue by continuously lying and twisting their improvements and thinking their users are dumb puppets.

The latter should consider that communication is so much better than assuming. We're not babies; instead of banking on our docility, show us that things are not where they ought to be. Show us that we really don't just want design innovation these days - look at the backlash Nothing and Google (for their Pixels) are facing for selling 2022 performance with 2025 premium prices. Show us that/when you're facing challenges and explain your roadmap to getting out of the rut. The problem is companies these days want their consumers to just understand, or they straight up just lie to them. That doesn't solve any problems; it just ignites and drives further user-to-tech-provider vitriol.

We're in for a long ride.

I hope this made you think though. Thanks for reading. 🧡

Discussion about this post

Ready for more?