What practice or approach has most accelerated your team's ability to prototype and iterate on AI-driven product ideas?
Most answers emphasize rapid prototyping and iteration using AI tools or feedback loops, though some focus on specific methods like live prototyping, prompting, or adopting an engineering mindset.
7 answers
Answers

Ravi Mehta
Angel Investor · Traveljoy
Live Prototyping in Calls
Ravi Mehta describes a client whose team spends the first 15 minutes of 60-minute research calls exploring needs, builds a prototype as the conversation continues, and has the customer try it in the last 15 minutes, so a loop that took six weeks can happen in one conversation.
Ravi on Product · Ravi on Product · Read

Ben Celebicic
Chief Product and Technology Officer (CPTO) · Hinge
Rapid feedback loop
Ben Celebicic says the biggest value he sees is the feedback loop, because it lets him learn much quicker whether something is worth building by prototyping without engineering support and putting it in users' hands.
Product School · Watch on YouTube
Jonathan Evens
Product Lead · Google Deepmind
Prompting
Jonathan Evens points to prompting, saying a lot can be done by iterating on prompts and building eval sets.
Mind the Product · Watch on YouTube

Paweł Huryn
Author, Founder · The Product Compass
Prototyping with Claude
Paweł Huryn recommends teams discover, discuss, research and prototype with Claude, letting the agent run spikes to confirm what's technically possible, instead of adopting heavy software development processes.
The Product Compass · The Product Compass · Read
Sabrina Ramonov 🍄
Creator · Sabrina Ramonov
AI-assisted coding
Sabrina Ramonov says she shifted to AI-assisted coding, where she drives and AI helps with scaffolding, brainstorming and tests, because vibe coding produced spaghetti code she couldn't maintain.
Sabrina Ramonov 🍄 · Sabrina Ramonov 🍄 · Read
Todd Olson
CEO / Founder · Pendo.Io
Engineering mindset
Todd Olson says an engineering mindset is the answer: when the technology doesn't do things automatically, you iterate and improve it—maybe provide more data—to solve the problem.
DataFramed · DataFramed · Listen at 34:45

Rohan Varma
Product, Codex · OpenAI
Prototype using AI tools
Rohan Varma says prototyping with AI tools like Codex lets him build and iterate on flows and simple interfaces, letting AI generate code while he guides structure, behavior and UX quality.
The VC Corner · The VC Corner · Read
More Product Management questions
View all →- What is the most important factor in AI product success?5 answers
- Should PMs do hands-on technical work with AI tools?6 answers
- What quality matters most for effective product leadership?8 answers
- What differentiator matters most for AI product teams today?5 answers
- What practice helps PMs progress without extensive planning?7 answers










