Does your company offer a product or just a feature? Dozens of companies have invested heavily in creating PDF readers that sit on top of OpenAI’s GPT-4 or plug in to ChatGPT Plus, including AI PDF, AskYourPDF, ScholarAI, and Webpilot. Unfortunately (for them), OpenAI’s recent update to ChatGPT Plus includes the ability to read PDFs. Oops!
This isn’t really a lesson in AI, it’s an old saw; building a third-party feature is only a good strategy if the company that owns the product decides to buy you or when your niche is so small that they’re happy you’re dealing with it. Otherwise, when the dragon breathes out (to mix my metaphors), nothing good happens.
OpenAI’s multimodal update raises several questions, but foremost is whether or not big tech is the future of AI. Do you really need immense computational power to do generative AI, or are there smaller models (requiring less resources to build and use) in our future? I’m also wondering about how “general purpose” the big models will become and how useful those features will turn out to be. Everyone can use a word processor. Can everyone use a multimodal autocomplete engine?
Author’s note: This is not a sponsored post. I am the author of this article and it expresses my own opinions. I am not, nor is my company, receiving compensation for it. This work was created with the assistance of various AI models, including but not limited to: GPT-4, Bard, Claude, Midjourney, Stable Diffusion, and others.