Google indexes structured web data. Just being able to search it makes it incredibly powerful. This goes for anything where there is a vast amount of data.

Increasingly our conversations occur via digital means and with AI tooling it will become possible to create structured data at scale from these conversations. I imagine a future, along with so many others, where we will invite our bots to our conference calls or who would reside in our messaging conversations. It will start with specialised bots that perform one function extremely well, such as Eva (acquired by Cisco) for note taking or x.ai for scheduling meetings. Over time we will see singular bots performing more actions, consolidating the marketplace, until we just have a handful who become our day to day assistants. I don't know how long this will take.

Human conversations are nothing short of complex. Not just a communication device (otherwise we would all speak the same language) nor an exchange of data. Relationships form from our language. In and out groups form from speaking the same language. Slang evolves. There is no strict process we humans follow and the conversational flow can double back, make subconscious leaps and at times become an electrifying experience.

So conversations are really complex but we have to start somewhere and I'm initialing focusing on the problems faced by Product Managers in collecting feature requests. In our conversations as product managers we are performing multiple actions such as collection of data (primarily product feedback and requests), validation of that data (sorting and de-duplicating), aggregating this information based on current context and understanding and prioritising all with in a few seconds.

These actions of collection, aggregation and decision making are not unique to product managers and will apply to a variety of job functions.