Product Management .. evolved with AI

For the past 3 years, I have been tracking major tech companies and their communication around AI.? Almost all of it is, either buzzword bingo or designed around poorly developed ideas.?The Apple WWDC 2024 Keynote was a masterclass in how to take a technical concept and translate it into user centred design. Ofcourse, 90% of the articles written afterwards didn’t seem to understand the difference between Apple Intelligence and ChatGPT. Most articles failed to even emphasise that ChatGPT was not a core element of the Apple's strategy. Effectively what Apple has done is to open up to the possibility that OpenAI will become one of many ‘Apps’ in the AI App ecosystem that Apple seems to be building. That may not really help the case of AI companies because this will create a competitive marketplace for Diffusion models, LLMs and VLMs.

The most impressive AI on the keynote was not ChatGPT but the MathNotes feature and the way Siri has been softly evolved into ‘voice + text’ assistant embedded into the platform experience.? So what does this tell us about the role AI will play in refining the art of product management ?

Increase the relevance of understanding human interaction patterns & choice architecture.

Product management has progressively become more deeply connected into the realtime flow of human interactions. When AI gets used as a force multiplier, the process of generating options for human interactions within a flow will lead to a multitude of ideas and options.? For a product manager and a product leader, the essential skill will be to understand when to break a pattern and when to reinforce a pattern.

Make writing and editing an essential skill for product management.

Many of the AI tools rely either on written or visual cues. Prompt engineering is a technical term for the idea of? crafting a story and a user journey in the interaction with an AI model. Good product management is often an editorial skill - Priorities are set inside a narrative of what we choose to prefer or a path we choose to take. Abundance of AI tools will narrow the distribution of product managers around the median as it will be easier to write an average quality PRD without doing the required thinking or synthesis.

Increase the relevance of the skill of hypothesis development and testing

In any scenario with multiple options, the primary product management skill tends to be the process of discovery through hypotheses building and testing. AI tooling will increase the speed of prototyping and hypotheses testing. It will be the product manager’s imperative to define a testable hypothesis as a precondition to prioritisation. Because of the open ended branched nature of most AI outcomes, we will see a stronger shift towards ‘ May the best hypothesis win’.

Increase the relevance of synthesis and insight

An abundance of ideas and speed at which options get created will generally slow down the product management growth path. It will take longer to build the insight required to accelerate or ignore the options generative AI creates. Personally I do not believe that there is such a thing as AI product management. AI in its various branches, is a tool, a very valuable tool indeed. Product Management is a professional skill at the intersection of many of the concepts that feed AI - Concepts like diffusion, synthesis, pattern recognition etc. In general terms, AI (especially the generative kind) will narrow the distribution curve within the profession and hence reduce the number of pure product management roles. Product managers will be able to synthesis current knowledge in faster ways but they will need to go from analysis and synthesis to decision making across more wider array of options.? PMs will have to choose between deep specialization of areas of overlap between customer-business-technology or moving into general management charter.? This is how synthesis and insight become more relevant for product management roles.

Read more: Shift the median or narrow the distribution... A ChatGPT tale.