Why the smartphone era is holding back consumer AI

This article summarizes an episode of TechCrunch’s video series featuring Goodwater Capital founder Chi-Hua Chien and Scribble Ventures founder Elizabeth Weil.

Goodwater’s founding partners : Chi-Hua Chien (right) and Eric Kim (left)/ Photo credit: Goodwater Capital

For anyone building a consumer AI product, Goodwater Capital founder Chi-Hua Chien shares a reminder. Chien compares the situation now to the mobile world before the App Store, an era full of promise but without the stability needed for a revolution.

Consumer AI is in its ‘flashlight app’ phase

Building on today’s platforms that change quickly carries a risk. Today’s popular feature can become tomorrow’s forgotten app.

Today’s top app is tomorrow’s OS feature
The first companies to create a product can see it become part of the main platform with one update. Chien compares it directly to the beginning of the smartphone age.

“When the App Store first came out in 2008, everyone rushed and created a bunch of apps,” Chien explains. “But they were apps like Flashlight… It was pretty easy for Apple to add that functionality to the OS, and it really subsumed that set of applications. The same thing is happening in AI right now.”

The big companies are already beating startups
Recent progress from big AI labs has reduced the relevance of several startups overnight. Chien has watched this trend with concern.

Chien says, “You saw a lot of these early AI apps that were super cool applications around video, around audio, around photo, but then Sora came out… and a lot of those opportunities disappeared.”

The signal for the next wave of apps

This period of unstable platforms creates challenges for founders. Building early carries uncertainty, but waiting too long can reduce opportunity. The question is: what shows that things are stable enough to build sustainability?

Waiting for things to calm down
Looking at how the mobile market grew, Chien sees a pattern that suggests the AI “platform wars” are moving into a more stable time.

He argues, “We’re right on the cusp of that equivalent in the mobile universe of the 2009-2010 era where the platforms have stabilized enough such that applications are now going to be created because it’s clear where the platforms are ending and where applications can begin.”

Competing on price is a good sign.
A clear sign of this shift is competition among other big platforms, especially when pricing becomes a key point of differentiation. 

Chien notes, “The first signal… is that you see some stabilization in the competitive dynamic between ChatGPT and Gemini, [with] market share getting split up and price competition starting to become part of how the platforms are competing. That’s the point at which it’s no longer a concern that they’re going to do everything.”

The smartphone can’t keep up

As software platforms improve, a challenge appears. The main device we use to access these new AI models is not designed for how AI works best.

Always-on, screenless devices will appear

AI will make expert services available to everyone

The biggest wins are companies built for AI

Human trust is the biggest roadblock

Changing behavior takes a generation