foo for thought

First, we had MSDOS. Then, GUIs became the norm. Now, AI is very quickly redefining how we use technology. It completely cuts out the middle man to access or manipulate information – the middle man being you, the user. No longer do you have to click and scroll google results for a search, it’s just a ChatGPT away. No longer do you need to look up resources for drafting documents – tools like Gemini operate directly inside the medium (Gmail). All this to say that, we still live in a world of apps – containerized and limited in their ability to perform complex tasks.

I can see AI doing more than redefining workflows in the world of apps, I can see it replacing the world of apps. I talked about this concept in previous blogs, but I think this is going to be the future. No, I know this is going to be the future.

I think iPhones will be obsolete in a matter of years. AI as a user interface has so many benefits over the traditional smartphone/computer, assuming the operating system supports. An AI-powered device could provide the following benefits, for example:

• Accessibility – with the ability to talk to or message the OS, visually or auditorily impaired users won’t be hindered in their ability to interact with the UI. In tech nowadays, we need to be mindful of this and design interfaces around the Narrator tool on operating systems. However with AI, we essentially get this for free
• Language support – Imagine a world where languages less-commonly spoken can get access to the same technology instantly. An AI powered operating system would immediately be accessible to everyone in the entire world, regardless of what language they speak, because the AI will be able to speak it too.
• Complex integration across apps – currently, we really only have import/export capabilities in our world of apps. With the world of MCPs taking off, I can see apps being able to interact with each other, orchestrated by some higher-order LLM. This will enable powerful integrations, and enable the user to do more without learning how to master something like Excel, PowerAutomate, or other productivity enhancers.
• Cost of hardware – Phones are pretty expensive, even to manufacture. If an interface could be simplified to something like a mic + speaker, perhaps hardware could get cheaper over time. This one is probably a stretch goal given that companies like Apple are still making things like the Vision Pro (not quite on that track retailing at 3400), but it might happen eventually.

And, many more. The part I’m the most excited about is improving the user’s ability to interface with technology. I’ve always been obsessed with homescreens. Ever since iOS 7, I’ve jailbroken/modded my way to the optimal homescreen setup, the intersection between perfect aesthetics, usability, and comfortability. The best setups that I’ve made for my homescreens were oftentimes the simplest. I believe AI can truly pave way to this simplistic future, by taking complex things and boiling them down to conversational UI.

I think with this, there’s also improvement in how AI looks. I think the conventional chat model that ChatGPT introduced us with may be outdated. How to improve on this? Not sure, but for me personally, I don’t like reading a bunch of stuff. I’m a very visual learner, I don’t do well with walls of text, which is essentially all ChatGPT is. So, maybe there’s a way other than an audio transcription that is better for this (I don’t like these either because I lose focus!)

Excited to see what Apple/OpenAI announce. Ball is in their court. The world is definitely on the edge of pivotal technological change, and I’m super excited to see what’s next.