I'm really excited about the new generative AI products because they mark the return of beta software. Google popularized the concept of [long-lived] beta software with products like Gmail [2004-2007], Google Docs [2006-2009], Google Finance [2006-2009], and Google News [2002-2006]. Now, we're seeing a lot of betas, such as ChatGPT, GitHub Co-Pilot, and Notion AI.

My teams follow this philosophy - it’s helped us create truly agile products. Build something "good enough" and get it in the hands of users, then iterate quickly. The success of the product will depend on whether people like it, use it, and invest in it.

For example, ChatGPT gained a million users in just one week, and the demand was so high that it crashed the service!

Google’s success in making beta software widely accepted emphasized 1) user feedback [have a way to capture feedback, adjust your feature roadmap], 2) testing [test with your users], and 3) transparency [it’s got bugs, it’s not full-featured yet, etc.], resulting in better products and a loyal community of users invested in their success. The Google betas generated buzz by word-of-mouth, not marketing. When users promote your product for you, you’re on to something - that’s undeniable NPS!

I used Notion AI (Alpha, not even Beta yet!) to help me write this post. AI-assisted human endeavors are about to take off!