Did OpenAI just kill my product?

It's funny that someone would think that way. Why would OpenAI kill their biggest (API) clients?

Chat-with-PDF function is part of the GPT-4 running features. It should work with the "wrappers". Also, this is only available on the plus version ($20/month subs), so they're basically just another competitor.

The "wrappers" like any other 3rd party microSaaS existing today, are here to stay to keep on building and "hacking" the already crowded AI space.

Loose idea, you could add an integration with Azure OpenAi to win clients that value privacy and don’t want to/can’t share PDFs content with OpenAi

I think multiple products can co-exist, but it does make it harder (will need to be much stronger in both product value created on top of what OpenAI offers and have great targeted marketing/sales).

It also makes it harder to capture a mass market because that's OpenAI's so it will likely need to be niche with specialised value. But maybe that's not a bad thing.

Oh yes, it's an interesting race. We built http://olympia.chat, and today OpenAI crashed together with all the wrappers built on Ruby including ours. It's good that we were able to recover quite fast and fix things, but it was quite unpleasant. I do not think we should give up tho, friend! :)

In my opinion, while OpenAI's "Chat with PDF" feature may introduce competition, there is still potential for the AI "wrapper" apps to thrive by offering unique features, addressing specific user needs, and providing a seamless user experience.

Absolutely, the introduction of OpenAI's "Chat with PDF" feature might bring about some changes in the landscape, but it doesn't necessarily spell doom for apps like bunni ai Your perspective on validation and the importance of multiple experiences in a given solution is spot on.

The existence of multiple Twitter clients and the success of specialized wrappers for ChatGPT illustrate that users often value tailored experiences and features. Your emphasis on 3rd-party apps taking on platform risks and providing flexibility for users is a strong point. The ability for users to easily switch providers without friction is a significant advantage.

As you rightly mentioned, execution is key. By staying focused and continuing to enhance specific features that cater to your user base, bunni can not only coexist but also thrive in this evolving landscape. The nature of business indeed involves adapting to changes and finding ways to provide unique value.

Keep refining your approach, stay responsive to user needs, and bunni can maintain its relevance and distinctiveness in the "AI for PDF" space. It's about offering something more than a generic solution, and your commitment to allowing seamless transitions for users positions your app well in the market.

Regenerate

Instead of thinking about wrappers and features being killed off by OpenAI, I think it makes more sense to return to the basics.

Michael Porter's Five Forces is business school gospel, and for good reason--the framework helps firms figure out where they fit in a marketplace.

Let's look at two of those forces first: the threat of new entrants and the bargaining power of suppliers.

With apps that build on top of OpenAI's API, there is often little technical difficulty preventing new entrants from creating their version. Much of the code is freely available on GitHub. That means a shallow moat based on execution and user experience. It's a tricky spot to be in.

Then, there's supplier bargaining power. How much power does OpenAI have in this equation? There are other LLMs out there, but since their prompts and behavior can vary, a service built around ChatGPT exclusively might find it tough to shift to Claude or an open-source LLM. In essence, OpenAI, just like any other API supplier with few substitutes, can squeeze or eliminate downstream services at a whim. We saw this with Twitter and its API recently.

So, should we all abandon services built on LLM APIs? Not at all. My favorite example is Upword AI. They combine document storage with an easy-to-use interface. On the back end, they have multiple LLMs to help create summaries. They may never attract enterprise users, but then again, Notion started off as a niche product for nerds, and that hasn't held them back.

What did you use to launch your affiliate program?

I saw that and had exactly the same idea, many pdf wrapper apps will lose a lot of customers. However, if you niche down and provide additional values, there is a high chance you'll be covered.

You probably need to add more value and find your niche, like help students study for tests. If you give the right value, the money will follow.

Your product seems to violate almost all the business commandments of C.E.N.T.S (Control, Entry, Need, Time, Scale).

https://www.youtube.com/watch?v=k_KlQ_Q-7cU

So if I had to speculate, yes it will die eventually.