From: aidotengineer
Munir, CEO and co-founder of Porsche AI, shares insights from his journey as a recent AI startup founder, transitioning from product roles at large technology companies like Stripe, Google, and Amazon [00:00:09]. This presentation aims to be particularly useful for those considering starting their own company or those in a similar stage in the AI space [00:00:25].
User Problem Discovery in AI
Munir notes a significant difference in user problem discovery and defining user problems in AI startups compared to his previous experiences in big tech [00:02:09].
Uncharted Territory
Traditionally, user problems often have a discernible shape and adjacency to familiar issues [00:02:31]. For example, at Stripe, a user’s need for card issuance related to their Stripe balance had a clear anchor [00:02:41]. However, in an AI startup, it feels akin to playing a “Soulslike” game for the first time, where understanding how to combine elements (like gear in a game) is initially unclear, even if the general goal (like increasing stats) is known [00:02:58].
Emergent Problems
Users in the AI space often don’t fully articulate their problems [00:03:25]. They recognize a desire to use AI and see potential opportunities, but the actual user problems are often an emergent property of the space, yet to be defined [00:03:28]. This is complicated by the rapid evolution of technology and what’s possible with AI [00:03:39].
Dynamic Roadmaps and Iteration
Unlike conventional three-to-six-month roadmaps, AI development necessitates making directional bets and iterating very rapidly by launching products and observing user reactions [00:03:44]. The team’s focus and anchor become hypothesis-driven, deliberate iteration, rather than traditional roadmaps [00:04:00].
Building New Narratives
Established practices like user storyboards and critical user journeys often become less relevant [00:04:24]. Munir likens this to playing No Man’s Sky, where players initially lack a clear story arc or mission and must create their own narrative [00:04:39].
Instead, AI startups must:
- Help users overcome the initial inertia of adopting AI for specific use cases [00:04:58].
- Collaborate with users to establish these narratives and anchor on a particular use case [00:05:12].
- Once a use case is identified, the process moves towards more familiar territory, where the product can be refined against that specific use case and its associated challenges [00:05:17]. This is like “colonizing a planet” in No Man’s Sky, leading to more tangible and scoped problem spaces [00:05:33].
This dual aspect presents both an opportunity and a challenge [00:05:48].