Brown says AI was part of every task in his workday. He uses a prompt to review product documents for customer evidence, logic, assumptions, risks, and tradeoffs, and another to anticipate questions from regular stakeholders using their past writing and feedback.
Best AI Workflows to Take Back Time at Work from Dropbox VP
Speaker
Morgan Brown, VP of product for AI at Dropbox, talks with Peter Yang on Peter Yang’s channel about practical AI workflows, building Dash, and how AI may reshape product management.
Brown describes Dash as AI-powered search and organization across work apps. In a demo, he searches for a music video project and finds related Slack messages, documents, images, and video clips. He says knowledge workers spend about 160 hours a year looking for information and switching contexts.
Brown says Dash depends on relevant, properly permissioned source data, which Dropbox’s Protect and Control tools help administrators manage. The team evaluates search and answers with LLM judges, human experts, product managers, and golden sets of prompts and results, while providing citations to source documents.
Brown defines moving fast as increasing the rate of learning, not simply producing more output. He recommends testing hypotheses with prototypes, customer conversations, research, or sales-call analysis, then using results to guide the next step. Clear decision principles can help teams resolve whether a disagreement concerns goals, principles, or options.
Brown looks for PMs who can structure ambiguous problems, execute quickly, and bring product sense and a point of view. He expects AI to handle more busy work and coordination, giving PMs more time with customers and teams and for creative problem solving.
Takeaways
- Turn recurring document feedback into a reusable prompt so reviews focus on gaps and risks.
- Prepare for stakeholder meetings by grounding likely questions in those stakeholders’ actual writing and feedback.
- Measure speed by how quickly a team tests a hypothesis and learns, not by how much it ships.
- State the principles behind a decision so teams can identify what they actually disagree about.
- Use AI to reduce coordination work and protect time for customer conversations and creative problem solving.
Mentioned in this talk
Search inside this talk
What was claimed 14
- role atMorgan Brown is VP of Product for AI at
Dropbox
“Morgan is the VP of product for AI at Dropbox.”
YouTube Unverified1 source· Talk
- built
“building Dash, Dropbox's new AI agent for work”
YouTube Verified2 sources· Talk, Job posting
- usesMorgan Brown uses
ChatGPT
Runs a daily script to summarize the top three AI papers published on arXiv“I have a ChatGPT Operator script that runs every day.”
YouTube Unverified2 sources· Talk
- feature
Dash: AI-powered search and organization across connected work apps
“Dash is AI-powered search and organization that connects all your work apps in one place.”
YouTube Unverified1 source· Talk
- feature
Dash: Searches across text and multimedia content including images, video, slides, and PDFs
“Dash is focused on working across different media types.”
YouTube Unverified1 source· Talk
- metric
Dropbox: A study found knowledge workers spend about 160 hours a year looking for things and switching between contexts
“knowledge workers spend about 160 hours a year just looking for things and switching between contexts”
YouTube Unverified1 source· Talk
- feature
Dash: Creates documents from reusable custom templates and company tone settings
“One of my favorite features is the company tone”
YouTube Unverified1 source· Talk
- featureDropbox Protect and Control: Lets administrators manage data flowing into Dash and document visibility
“allows administrators to manage the data flowing into Dash and manage the visibility of documents”
YouTube Unverified1 source· Talk
- usesLLMs judge relevance and ranking; golden sets and human experts complement evaluation
“We use LLMs as judges for relevance and ranking”
YouTube Unverified1 source· Talk
- previously atMorgan Brown previously at
Instagram
“When I managed a team of PMs at Instagram”
YouTube Unverified1 source· Talk
- previously atMorgan Brown previously at
Meta
“When I got to Facebook”
YouTube Unverified1 source· Talk
- previously atMorgan Brown previously at
Shopify
“You worked at Meta, then with Tobi at Shopify”
YouTube Unverified1 source· Talk
- usesMorgan Brown uses
GrokUses it as part of a council of LLMs“Right now I use Dash, Claude, ChatGPT, and Grok.”
YouTube Unverified1 source· Talk
- usesMorgan Brown uses
Claude
Uses it as part of a council of LLMs“Right now I use Dash, Claude, ChatGPT, and Grok.”
YouTube Unverified1 source· Talk









