.svg)
.svg)
We recently attended OpenAI DevDay along with 1500 other developers to hear the latest and greatest when it comes to new products and releases from OpenAI. There was a lot!
Obviously the internet broke with the release of Sora 2, but there was a lot more that was covered including agent builders and an interesting conversation between Sam Altman & Jony Ive (best known for his sleek design work at Apple) about the future of devices in the AI era.
This blog post focuses on what matters most to creators who use and monetise custom chatbots in the OpenAI ChatGPT ecosystem.
No mention of GPTs or the GPT Store
What got our attention the most was what wasn’t said. Specifically, there was no mention of GPTs or the GPT Store at all.
Do OpenAI consider it to be a stable product that doesn't require any further improvements other than maintenance? Are they just putting it on the back burner temporarily? Are they considering to pull GPTs, the way they pulled plugins? Guess we will have to wait and see.
It is a lot of uncertainty for individuals and businesses who have invested in the GPT Store to navigate.
Our approach is to diversify platform risk.
We continue to support and grow GPT Tools which allows creators to monetise chatbots in the OpenAI GPT Store.
We are simultaneously working on our own platform that operates independently of OpenAI so that creators have an alternative option (currently in private beta for our existing user base, but looking forward to general release soon!).
What about apps for ChatGPT?
One of the first announcements on the day was apps for ChatGPT. These are apps developers can build using the Apps SDK, and then publish for ChatGPT to use within a chat with a user.
We love the vision of ChatGPT becoming more useful with the addition of apps, and opening up the chat interface to include more visual and interactive user experience.
However, there are three things that give us pause.
1. You need to code
You need to code to build an app for ChatGPT, representing a swing back to developers, rather than non technical builders.
If you're in the vast majority of GPT creators who built GPTs because you have interesting knowledge or consumer insights but don't code, then apps for ChatGPT is not for you.
2. Unclear discoverability
We loved the vision of ChatGPT natively recommending and using apps.
We’ve long advocated for ChatGPT to recommend relevant GPTs to solve the discoverability problem and deliver more relevant answers to the user, so we’re thrilled to see OpenAI moving in this direction (maybe it will be extended to GPTs in the future?).
That said, we’d love more clarity on the criteria ChatGPT will use to recommend or invoke apps. During DevDay, the demo examples showcased apps by major brands like Canva and Zillow. We hope this doesn’t signal a repeat of the GPT Store dynamic, where big brands were prioritised — sometimes at the expense of better independent apps.
3. No monetisation
Sam Altman kept it incredibly vague saying they would be some sort of monetisation option in the future.
Given the backlash OpenAI got for reneging on usage based revenue share for the OpenAI GPT Store, this is perhaps a smart move!
But the tl/dr is that you can't money from building apps for ChatGPT for now.
Given that so many other AI-powered products are way quicker to monetise, is building apps for ChatGPT the best use of our time...?
Summary
This is the third time OpenAI have asked third parties to build for their ecosystem. First it was plugins, which was killed. Then it was custom GPTs which are in some kind of product limbo. And now we're being asked to build apps for ChatGPT....
It's tempting to roll our eyes, but even the slimmest hope that this time around, OpenAI will be able to offer app builders a way to be discovered by ChatGPT's 800M weekly users is hard to resist.
Our take is that apps for ChatGPT is definitely worth experimenting with (why not?!) but if you're looking to build a business and earn revenue, look elsewhere until OpenAI maps out the rules of play for discoverability and monetisation of apps for ChatGPT.
We'd love to hear your thoughts. Do you think third time will be the charm for independent builders and developers? Or should we all be focusing on using the API or on-device AI to build our own AI-powered apps that exist outside of the ChatGPT ecosystem?
