The End of Computing
Disclaimer: These thoughts are my own and not representative of my employer. These ramblings would have gone to a discord server i’m in but instead all of you get to see them. Lucky you! I didn’t edit or proofread this so expect it to sound like a long winded rant.
There’s this infamous concept called the end of history. it was proposed by Francis Fukuyama, and the main concept is that we have reached a point in human history in which we have gone through all major changes. Change would still happen of course but at a smaller rate. There would be no more walls falling, no further world wars. Regardless of how I personally feel about this concept in its original application (I am not a fan) I keep thinking about how it applies to the tech industry. I can’t stop myself from thinking that we’ve hit a similar point with computing. The revolutions of the personal computer, internet, and modern smartphone have had me wondering for years what the next evolution of technology would be. Smart glasses? Smart watches? Virtual reality? The Metaverse? Self driving? Crypto? AI? Every year this focus changes and changes again. All trends which materialized into nothing.
I am decently optimistic about the possibility of VR / LLM’s, but I used to be optimistic about smart glasses and crypto many years ago so I probably shouldn’t be trusted when it comes to that. Even now I find myself often more pessimistic about the technologies I used to be optimistic towards, as I find their implementations lacking / incomplete. Regardless, I think we’ve reached the end of computing. The big shifts already happened with the mac / windows, ios / android, and even if every one of the aforementioned trends actually do materialize into the next computing platform they arent going to transform society any more than the personal computer, internet, or smartphone did.
Even though I see myself optimistic about LLM’s such as OpenAI’s GPT, I see their potential implementations less as instruments of creation but instead refinement. I have been fascinated with how the tech industry specifically got self driving so very wrong, with its expanding problem space. While that never materialized into anything actually useful, its became a lodestar which led to the creation of new crash detection technologies. Assistance, not autonomy. I see LLM’s as destined for a similar path. They can write emails, but those emails are boring and lack personality. However I actually expect them to exist as a copilot to enhance and critique your writing, rather than simply replace it. Imagine a word processor with a built in copilot to your right, telling you that you need to improve your conclusion or that you’re rambling about LLM’s in the middle of a blog post that was supposed to be about phones or whatever.
I think while there’s new exciting technologies and new possibility, even if LLM’s materialize into the AGI that twitter grifters are foaming at the mouth to predict the timetable of, they won’t matter. We’ve already got a device in our pocket that connects us with everyone and can access the entirety of human knowledge. The best case outcome of this technology is just a mathematical model that reads summarizes search results for you. It’s transformative, but not world changing. Same with Smartwatches, its just a phone on your wrist. Virtual Reality, for as much as I adore it, has a massive fight ahead of it as long as their biggest friction is people who own these headsets not even wanting to put them on. The Metaverse as a concept already exists and is Discord, and anyone trying to build 3D virtual spaces is missing that 2D ones are better. Crypto? Can go on for hours about why that wont work.
A quote I keep thinking about is something Nilay Patel said on the vergecast once, paraphrasing here, “The iPhone was such a successful product that it reshaped society around it”. He’s right. The iPhone and Android reshaped society. The internet reshaped society. Personal computers reshaped society. LLM’s write essays and recombines information. I’m impressed, but not floored yet. I think it can finally be the foundational technology people have been waiting for years to build a new generation of products on… but i’ve been wrong before.