The myth of 10x faster
How Product Managers can be more effective with AI
Everywhere you look, there’s noise claiming AI is making product managers "10x faster". It’s not. That narrative is lazy and misleading. What AI is really doing is removing excuses. It’s exposing gaps in thinking, decision-making and prioritisation. The best PMs are using AI to amplify their impact. The rest are just adding more noise to the system.
AI doesn’t replace thinking. It accelerates output. If your thinking is shallow, AI will only get you to the wrong answer faster.
The problem: hype vs reality in AI-driven product management
AI isn’t a cheat code. It doesn’t replace the fundamentals: clarity of thinking, stakeholder alignment and product judgment. Yet countless teams are treating it like a magic lever—expecting exponential velocity while skipping the messy, human parts of product work.
Let’s be clear: generative AI is a real productivity tool. But the gains are nowhere near 10x, and they certainly don’t absolve you from being a rigorous, strategic PM. In fact, if you’re not careful, AI will just make your bad decisions look more polished.
What the research actually says
Here’s what the data shows:
McKinsey (2024): AI improved PM productivity by ~40% and reduced time-to-market by 5%. Impressive, but nowhere near "10x".
arXiv Study (2023): PMs saved ~13–15% of time on documentation. No evidence of exponential speed gains.
Lenny Rachitsky’s field research: AI helps PMs go deeper on strategy and planning but doesn’t replace product sense, storytelling or critical thinking.
The delta is tactical not transformational.
AI is a force multiplier. Not a substitute.
AI works best when paired with strong fundamentals. It gives leverage to experienced PMs who know how to ask sharp questions, make tough trade-offs and simplify complexity. Used well, it can accelerate:
Competitive analysis: tools like Perplexity and Claude can summarise markets in minutes
Customer research: ChatGPT and custom GPTs can sift through call notes or surveys fast
Trend mapping: you can build live AI-driven radar for emerging patterns in your space
But none of this is plug-and-play. AI outputs are often incomplete, overly confident or just wrong. You still need to validate, interpret and connect the dots. If you don’t know what good looks like, AI won’t get you there.
AI for documentation: useful but shallow
Tools like Notion AI and ChatPRD are gaining traction for:
Drafting PRDs
Summarising meetings
Generating user stories
This helps, especially for junior PMs. But here’s where they break:
Complex trade-offs: AI can’t reason through edge cases, sequencing or political context
Org-specific nuance: every company has its own rhythms, language and internal logic
Product vision: AI doesn’t understand your strategy—it regurgitates structure, not insight
Think of AI as scaffolding for writing not a replacement for your judgment.
AI for user insights and execution: assistive not autonomous
Let’s break this down:
User feedback analysis
Tools like Zeda.io and Dovetail AI cluster and tag user feedback at scale
They surface themes, but often miss context or intent
Human review is still required to spot what matters vs what’s noise
Execution tools
Gamma AI makes slide creation fast, but struggles with enterprise formatting
Coda AI can automate internal workflows, but has a steep learning curve
None of these tools handle stakeholder alignment, trade-offs or internal politics
These tools reduce manual effort. But they don’t touch the real work of product management: influence, prioritisation and narrative.
The risk: AI can make average PMs worse
The danger isn’t that PMs don’t use AI—it’s that they overuse it, blindly. Especially junior PMs who haven’t built their product muscle yet. Here’s what happens:
AI-generated plans get shared without critical review
Shallow PRDs get approved because they "look complete"
PMs lose context by outsourcing too much synthesis
The result? Teams ship faster in the wrong direction.
Senior PMs know better. They use AI to augment their thinking, not replace it. They treat AI outputs as first drafts—not final answers.
Where AI helps most (and least)
Let’s get tactical. Here’s where AI actually delivers impact:
What to do instead: a practical AI playbook for PMs
Here’s how smart PMs are using AI today:
Start with clarity. Before prompting a model, write down what you actually want to learn or decide. AI can’t think for you—it can only amplify your intent.
Use AI for velocity, not vision. Let it draft, summarise or cluster—but you do the hard work of insight, prioritisation and communication.
Validate everything. Never ship AI outputs without review. Cross-check facts, challenge assumptions and rewrite in your own voice.
Build AI muscle slowly. Start with one workflow (e.g. meeting notes summarisation or feedback clustering), prove ROI and then scale.
Stay in the loop. Don’t automate your way into ignorance. Use AI to dig deeper, not to disengage.
Train your team. Create AI usage guardrails, run internal workshops and set expectations around validation and quality.
Bottom line: AI won’t make you faster, it’ll make your gaps obvious
AI is a force multiplier. But if you’re unclear, indecisive or politically naive, AI won’t help. It’ll just make your flaws scale faster.
The future of product management isn’t about who types fastest or prompts best. It’s about who can think clearly, validate quickly and drive outcomes with precision. AI can help you get there but only if you’re already strong at the fundamentals.
If you’re a PM trying to "go faster" with AI, you’re asking the wrong question.
Ask: how can AI help me make better decisions, faster?
That’s the competitive edge.