Aakash GuptaGoogle AI PM Reveals the Tools 99% of Product Managers Don’t Use
Aakash Gupta and Marily Nika on google AI PM shares overlooked tools and AI-first workflows for PMs.
In this episode of Aakash Gupta, featuring Marily Nika and Aakash Gupta, Google AI PM Reveals the Tools 99% of Product Managers Don’t Use explores google AI PM shares overlooked tools and AI-first workflows for PMs Marily Nika outlines a six-tool daily stack—AI Studio, Opal, NotebookLM, Perplexity, custom ChatGPT/Gemini GPTs, and Fireflies—and ties each tool to a specific PM job-to-be-done.
At a glance
WHAT IT’S REALLY ABOUT
Google AI PM shares overlooked tools and AI-first workflows for PMs
- Marily Nika outlines a six-tool daily stack—AI Studio, Opal, NotebookLM, Perplexity, custom ChatGPT/Gemini GPTs, and Fireflies—and ties each tool to a specific PM job-to-be-done.
- She argues the modern workflow is “idea → prototype → pull in engineers/scientists,” reducing reliance on PRDs early and using documentation mainly for complex, cross-functional work.
- She demos AI prototyping to rapidly turn natural language into functioning mini-apps, compares AI Studio vs Opal, and highlights practical constraints like quotas and iteration needs.
- She shows NotebookLM as a “closed-context” research assistant for domain upskilling, interview preparation, and large-scale synthesis of qualitative research, with standout audio-overview experiences.
- She frames AI adoption as a career moat—PMs won’t be replaced by AI, but by PMs who use AI—and provides an 18-month roadmap plus interview red flags for aspiring AI PMs.
IDEAS WORTH REMEMBERING
7 ideasPrototype before you socialize the idea.
Nika replaces the old “idea → PRD → comments” flow with “idea → AI prototype → debate the prototype,” pulling in engineers and scientists after the behavior is visible to accelerate alignment.
Use AI Studio when you need higher-fidelity iteration and transferable output.
AI Studio is positioned as better for visual/design tweaks via conversation and for mimicking existing UI by uploading screenshots; it can still require quotas and iteration but speeds early exploration.
Use Opal when you want a fast, disposable workflow or mini-app.
Opal turns one sentence into a multi-step workflow (Zap-like blocks) and a runnable mini app; it’s great for quick utilities (e.g., empathy maps) but may be “wonky” and is less code-portable.
Treat NotebookLM as your “domain ramp-up” and synthesis engine.
Because it only uses provided sources, it’s effective for learning new domains, compressing long videos into interview-ready bullets, and distilling large volumes of UXR without web noise.
Leverage Perplexity’s discussion filter to capture the ‘voice of users.’
Turning on “discussions and opinions” and focusing on Reddit yields rapid, source-linked insight; she then prompts it to convert findings into MVP feature lists aimed at product-market fit.
Invest in one or two custom GPTs that match your voice.
Her PRD generator probes for missing inputs, then outputs personas, features, metrics, and even mock/UX ideas; she argues even a mediocre AI draft beats a blank page and saves compounding time.
Normalize AI use—hiding it is the real mistake.
She recommends explicitly stating you used AI and even sharing the generator link; she believes the competitive threat is not AI itself, but peers who use these tools while you don’t.
WORDS WORTH SAVING
5 quotesGoogle launched AI prototyping on AI Studio. It's amazing. Under 10 minutes, you have a fully functional app.
— Marily Nika
Whereas now, the workflow is you have an idea… you prototype and then bring my scientist and engineer in.
— Marily Nika
It's that it only uses the knowledge you provide.
— Marily Nika
If you want one takeaway from this podcast is don't be embarrassed to admit that you use AI.
— Marily Nika
Who is gonna replace PMs is people using these tools when you're not.
— Marily Nika
QUESTIONS ANSWERED IN THIS EPISODE
5 questionsAI Studio vs Opal: what decision rules do you use (fidelity, speed, portability, audience) to pick one over the other?
Marily Nika outlines a six-tool daily stack—AI Studio, Opal, NotebookLM, Perplexity, custom ChatGPT/Gemini GPTs, and Fireflies—and ties each tool to a specific PM job-to-be-done.
You mentioned uploading screenshots to match a design system—what prompt pattern and assets (tokens, components, CSS) make that reliably consistent?
She argues the modern workflow is “idea → prototype → pull in engineers/scientists,” reducing reliance on PRDs early and using documentation mainly for complex, cross-functional work.
In your “prototype-first” workflow, what specific situations still demand a traditional PRD early (regulatory, platform changes, multi-team dependencies)?
She demos AI prototyping to rapidly turn natural language into functioning mini-apps, compares AI Studio vs Opal, and highlights practical constraints like quotas and iteration needs.
NotebookLM: how do you structure sources and prompts to avoid biased synthesis when your inputs are incomplete or skewed?
She shows NotebookLM as a “closed-context” research assistant for domain upskilling, interview preparation, and large-scale synthesis of qualitative research, with standout audio-overview experiences.
Perplexity Reddit research: how do you validate that vocal Reddit users aren’t misleading your prioritization versus your actual target segment?
She frames AI adoption as a career moat—PMs won’t be replaced by AI, but by PMs who use AI—and provides an 18-month roadmap plus interview red flags for aspiring AI PMs.
EVERY SPOKEN WORD
Install uListen for AI-powered chat & search across the full episode — Get Full Transcript
Get more out of YouTube videos.
High quality summaries for YouTube videos. Accurate transcripts to search & find moments. Powered by ChatGPT & Claude AI.
Add to Chrome