Lenny's PodcastAman Khan: Why top AI PMs solve problems, not chase chatbots
What separates AI PMs is shipping real prototypes, not credentials; Cursor, Replit, and v0 turn mocks into demos while curiosity beats certifications in hiring.
CHAPTERS
- 0:00 – 1:04
Why AI makes PMs higher-leverage: come to meetings with prototypes
Aman opens with a provocative idea: AI tools make it realistic for PMs to show up with working mocks and prototypes, not just PRDs. This reframes the PM’s role as a higher-resolution starting point that accelerates collaboration with design and engineering.
- •AI enables PMs to prototype quickly and concretely
- •Higher-fidelity starting points improve alignment and speed
- •Prototypes can be wrong—but still valuable for direction-setting
- •Sets the theme: AI as leverage for PM influence
- 1:04 – 6:24
Aman’s background and the episode’s focus: AI PM + long-term IC success
Lenny introduces Aman Khan (Arize AI, ex-Spotify/Apple/Cruise/Zipline) and frames the conversation around two intersecting trends: AI in product and the rise of “super IC” PMs. They set an agenda: breaking into AI PM, thriving as an AI PM, and thriving as an IC long term.
- •Aman’s career across AI-focused product roles
- •Trend toward fewer management layers and more empowered ICs
- •Roadmap for the conversation: get into AI, thrive in AI, thrive as IC
- •Why Aman is a frequently-cited standout IC PM
- 6:24 – 10:59
What an “AI PM” actually means: three distinct role flavors
Aman defines product management broadly, then breaks AI PM into three archetypes: platform PMs building for AI engineers, AI product PMs shipping AI-native experiences, and AI-powered PMs who use AI tools to do classic PM work faster and better. He predicts most PMs will increasingly fall into one of these AI-adjacent categories.
- •AI platform PMs: tools for AI engineers (observability/evals, etc.)
- •AI product PMs: AI is the core experience (e.g., ChatGPT-like products)
- •AI-powered PMs: using models/tools to enhance PM workflows
- •AI becomes as ubiquitous as databases in SaaS
- 10:59 – 13:53
Tooling for the AI-powered PM: prototyping, UI generation, and visuals
Lenny asks for concrete tools, and Aman shares a stack for rapidly building prototypes and visual artifacts. The emphasis is on using these tools to explore what’s possible and tell a story—not to ship production code immediately.
- •Cursor and Replit for fast functional prototypes
- •Vercel v0 for prompt-to-UI/landing pages and iterative design
- •Midjourney/DALL-E for logos, visual concepts, and user stories
- •AI lowers barriers across even physical/3D product exploration
- 13:53 – 18:17
How to pivot into AI PM: curiosity, problem obsession, and trying cutting-edge tech
Aman describes feeling like an outsider (mechanical engineering background, no ML degree) and stresses that AI PM still starts with loving the customer problem. He uses customer support as an example: start from an ideal experience, then evaluate which new AI capabilities (e.g., real-time voice APIs) unlock better solutions.
- •You don’t need a PhD—start with customer problems and curiosity
- •Define the ideal experience first; AI is a means, not the goal
- •Use emerging tech (e.g., voice/real-time APIs) to rethink workflows
- •Hands-on experimentation creates ‘aha’ moments to bring to teams
- 18:17 – 25:42
Skills + standing out: fundamentals, build a portfolio, and shortcut the interview loop
Aman separates “baseline skills” from “standing out.” Learn AI/ML fundamentals through accessible resources (e.g., Karpathy), then demonstrate applied understanding by building prototypes and shipping small projects as a portfolio—helping answer hiring managers’ key questions before interviews even begin.
- •Learn fundamentals (what models do, boundaries/constraints)
- •Use tools directly to internalize what’s possible
- •Build a portfolio of prototypes/projects to stand out
- •Hiring evaluates: can you do the job, are you excited, are you enjoyable to work with
- 25:42 – 29:20
Why product management isn’t dead: AI makes ‘what to build’ the bottleneck
Lenny argues AI flips the fear narrative: if building is easy, opportunity discovery and taste become the scarce skills—core PM strengths. Aman agrees and adds that AI tools amplify PM influence by improving how ideas are communicated to design, engineering, and leadership.
- •AI accelerates building; deciding what to build becomes hardest part
- •PM strengths: problem selection, articulation, and product taste
- •AI tools increase influence via better communication artifacts
- •AI PM can be a ‘highest leverage’ role in organizations right now
- 29:20 – 36:05
Thriving as an AI PM: don’t clone ChatGPT—design the right interface
Aman contrasts average vs top-tier AI PMs using the post-ChatGPT era: many teams rushed to build internal ‘ChatGPT on our data’ chatbots with unclear adoption and value. Top AI PMs resist copycat solutions, question whether chat is the right interface, and design AI experiences tailored to the real workflow bottleneck (e.g., automating analysis).
- •The ‘ChatGPT on our data’ default often isn’t the best answer
- •Different problems demand different AI interfaces (not always chat)
- •Evaluate whether a solution is truly what the business/customer needs
- •Use small builds to learn, but optimize for differentiated workflows
- 36:05 – 39:35
Idea generation that works: metrics for experimentation, hackathons, and UX teardowns
Aman offers practical ways companies can find strong AI ideas: define a metric for ‘shots taken’ (prototypes/experiments), run hackathons to make AI approachable, and study what makes AI products feel magical via teardowns. He shares an internal hackathon example where a seemingly simple AI Slack routing bot turned out to be hard due to missing context—highlighting the value of fast learning.
- •Create a metric for AI experimentation (not just revenue/KPIs)
- •Hackathons drive hands-on learning and de-mystify AI
- •Expect most attempts to fail; the goal is insight and pattern recognition
- •Do teardowns of great AI products to learn UX and interaction design
- 39:35 – 42:45
Don’t automate away the customer experience: control knobs and the ‘Ikea effect’
Aman warns against over-automation with an analogy: Betty Crocker’s ‘just add water’ cake mix underperformed until customers had to add eggs—restoring a sense of agency. Great AI products preserve user control and lower the barrier to creation rather than removing the human from the loop entirely.
- •Full automation can reduce satisfaction even if it’s technically possible
- •Users often want agency and contribution in the outcome
- •Betty Crocker example: adding a step increased adoption
- •AI should reduce friction and enable creation, not erase involvement
- 42:45 – 47:44
What separates top 5% AI PMs (continued): deliver KPIs while making room to explore
Borrowing Kevin Yin’s framing, Aman argues top AI PMs must ‘walk and chew gum’: ship real customer value and move metrics while simultaneously carving out time for prototyping, experimentation, and learning. Because the tech changes fast, successful AI PMs build a culture of iteration and accept frequent failures as part of progress.
- •Balance execution (KPIs) with exploration (prototypes, experiments)
- •Create organizational space to try tools firsthand and learn quickly
- •Accept fast change and failure as normal in AI product development
- •Continuous learning loops outperform one-off ‘big bets’
- 47:44 – 57:11
Long-term IC success: energy as a force multiplier in ambiguity
Shifting to IC career mastery, Aman introduces three themes: energy, wandering vs waiting, and amplifying signal. He explains how ‘bringing energy’ in meetings and taking personal initiative (e.g., doing outbound customer discovery via LinkedIn) reduces friction, builds trust, and helps teams move forward when direction is unclear.
- •IC PM work is hard and the bar is rising
- •Energy changes how teams collaborate and handle uncertainty
- •Initiative: do the work yourself (sales/outreach/customer development)
- •Pairing on others’ work builds empathy and stronger cross-functional execution
- 57:11 – 1:02:04
Wandering vs. waiting: the PM as the scout for the next direction
Aman describes the PM’s job as being comfortable in the ‘unknown,’ exploring while others execute known roadmaps. Many organizations prefer to wait (for the next model release, the next trend), but PMs create advantage by wandering—searching for sparse signal, iterating, and recognizing product pull when it appears.
- •Companies default to ‘wait and see’; PMs should often ‘wander’
- •The PM is responsible for finding direction amid ambiguity
- •Zero-to-one feels squishy; uncertainty is part of the process
- •Competitive advantage comes from exploring before it’s obvious
- 1:02:04 – 1:03:18
Amplifying signal with AI tools: scaling customer insight beyond your calendar
Aman explains how AI can help PMs extract patterns from large volumes of customer conversations and sales calls. By feeding transcripts into long-context LLMs (and leveraging tools like Gong), PMs can identify recurring pain points and make better decisions without attending every meeting.
- •Use AI to process transcripts and find common themes quickly
- •Gong and similar tools capture customer/prospect signal at scale
- •Long-context models enable deeper synthesis across many calls
- •AI gives PMs ‘multiple places at once’ leverage for decision-making
- 1:03:18 – 1:17:33
Just have fun + lightning round (books, shows, products, motto, SEO joke)
Aman closes his IC advice with a reframing: product can feel heavy, but staying playful and curious accelerates learning and resilience. In the lightning round, he shares favorite books (Bill Bryson; Designing Your Life), a Tour de France doc, favorite tools (Websim; plus a scented recycled-apple notebook), a Steve Jobs quote on living your own life, and a humorous aside about sharing a name with a famous cricketer.
- •‘Have fun’ as a practical mindset for sustained performance
- •Book recs: A Short History of Nearly Everything; Designing Your Life
- •Favorite products: Websim; prototyping tools; apple-scented notebook (Appeel)
- •Motto: don’t waste time living someone else’s life
- •Closing: how to reach Aman and his ‘top three content’ outreach rule
