On the Future of Products

after finishing the last piece on the future of (software) apps, i went for a walk and continued piecing together my thoughts on products as a whole (not solely digital, but also physical). i came to the following belief:

the most meaningful & innovative UX challenges in tech are moving from screens to physical products

what does this mean?

1. user preferences will shift to the superior hardware option, making the hardware more valuable in purchase decisions
2. hardware will replace software as the differentiator
3. an acceleration in products defined by being software + hardware products, as opposed to those that stand on only one of the legs

another way to say this: i am bullish on physical products in tech

and i want to work on them

congrats: you’re early to this post. i haven’t fully fleshed out the reasoning and ideas (including the above three points), but publishing early gives me something to iterate on. the above three points and the following are a slightly-consolidated chain of thoughts that will likely be revisted soon:

why do i think this?

• AI makes software invisible: so much work is done in software to make products that “just work”. a common goal is to eliminate steps/clicks, things that the user needs to configure or worry about. AI is this ultimate version of this because it is capable of automating and generating. As the models and their capabilities both expand and become more controllable (by both the end-user and the developer), there’s less the user will need to do.
• what’s scarce becomes valuable: AI is making software abundant. cheap and easy to make by anyone. therefore what’s harder to make, the hardware, becomes more vaulable. what’s “premium” will more likely come down to a decision between hardware options, not OS or UI.
• the ideal interface is no interface: the application layer is already being compressed to an agent. the rigid applications that used to sit between you and the data/information you needed is being replaced by an agent that gets all the information for you (currently through natural language). what does the ideal version of this agent-interface look like? not sure. my previous post discussed a few points.

obvious objections:

• what if someone builds a differentiated AI-first interface? AI can easily code the copy.

on point (3) about an acceleration in software + hardware products: we’ve already seen how we’ve turned hardware products into software and hardware products.

hardware -> software + hardware: fridges, glasses, cars. software -> hardware + software: ?? what things do we do with primarily software products today will become things we do with, or primarily through, hardware products ??

key to either scenario: the right timing and solving real problems. either of these can make a product flop: google glasses being too early before the software was good enough or juicero using wifi to squeeze a juice packet.

contrarian? sorta. it goes against the popular (although now dated) belief that “software is eating the world” and the general digital solution supremecy in tech and business circles has mainly been due to their scalability, margins, and ease of distribution. but the pendulum is already swinging (as discussed in the previous article with ambient computing and AI-native devices). as you will see, i don’t even think the product needs to be AI-native, for this conclusion to hold true.

through my own experience building products, i’ve seen how hardware + software products bring a ‘magic’ that software products could never deliver on their own:

• a record player that brings your memories to life with music
• a floating bed frame that feels like sleeping on a cloud

anything that reminds us we are human and grounds us (let me touch, feel, and be a part of it) will be more meaningful as we try to validate our existence amongst AI-counterparts whose numbers and intelligence surpass our own.

and any new trend that reduces our screen time will forever be a friend of mine.

now this conclusion didn’t all just come together today. it was part of a series of thoughts over the past couple years and since accelerated through my experience in the past year where I’ve gotten more familiar with AI, software, and glimpses of what the future will look like in that world. and it’s based on what i’ve always wanted and cared about: building things that genuinely help people.