Hey there! We recently ran our first survey with our Stargazers audience about how professionals in product management and design currently use and feel about AI. Shout out to our early responders! 
If you haven’t seen it yet, feel free to take the test now.
Check out the thread below for our findings from this survey and a peak into how your peers are thinking about AI in different situations:
Our initial Frequency score (85%) indicates that many product leaders are already tapping into AI, just 5% away from a Very High frequency.
Participants described a wide variety of ways they integrate AI into their workflows, with themes emerging around efficiency, ideation, and analysis.
Key Uses:
Workflow acceleration: speeding up product design and research tasks.
Content support: generating copy, ideas, or hypotheses in early stages.
User insights: clustering, analyzing, and summarizing research findings.
Design enablement: kickstarting prototypes or exploring alternatives.
Participant Quotes:
“I use it to help create faster workflows for me team.”
“To speed up the product design processes, automate processes, and create AI product features.”
“Analyze and cluster user findings, brainstorm about ideas, get feedback about alternatives, write problem statements and hypotheses.”
“For kickstart and for fine-tuning..”
“Love it.”
Takeaway: AI is primarily seen as a time-saver and creative accelerator, reducing friction in repetitive or exploratory tasks.
Despite its utility, participants highlighted frustrations centered around trust, transparency, and consistency.
Key Struggles:
Team adoption & buy-in – challenges aligning stakeholders.
Transparency & explainability – not knowing how AI arrives at results.
Contextual awareness – AI struggles to retain history or adapt to nuanced needs.
Consistency of responses – outputs vary in quality and reliability.
Overhead concerns – unclear costs or efficiency tradeoffs.
Participant Quotes:
“AI is difficult to get buy-in from the whole team.”
“Transparency of processes running in the background, and to be able to quantify outcomes (with AI costs).”
“It has no contextual memory of all my work .”
“Consistency and response quality with followup questions.”
Takeaway: While enthusiasm exists, many participants remain cautious—highlighting the fragility of trust when AI lacks reliability or clarity.
Participants were asked to rate their trust in AI across common product manager/designer scenarios.
Most Trusted Scenarios:
Copy & content creation (47% Trustworthy, 27% Very Trustworthy)
Prototyping and design assets (47% Trustworthy)
Analyzing user data and insights (40% Trustworthy, 13% Very Trustworthy)
Least Trusted Scenarios:
Testing and optimization (27% Untrustworthy, only 20% Trustworthy)
Designing AI products/features (27% Untrustworthy, 0% Very Trustworthy)
Developing AI features that interact with other AI (13% Untrustworthy, 0% Very Trustworthy)
Middle Ground (Mixed Neutrality & Hesitation):
Decision support and prioritization (53% Neutral)
Personalization and recommendations (60% Neutral)
Illustrative Tensions:
High trust for creative acceleration (content, prototyping).
Low trust for mission-critical or technical functions (optimization, AI-to-AI systems).
Neutral stances suggest participants are undecided about AI’s role in personalization and prioritization—reflecting ongoing skepticism.
AI is valued as an assistant, not an authority: Participants trust AI for inspiration, drafts, and time savings, but hesitate when decisions have significant consequences.
Trust is earned through reliability: Inconsistent answers, lack of context memory, and black-box processes hold back adoption.
Creative vs. critical split: The more “creative support” tasks show high trust; the more “system-critical” or “decision-making” tasks show skepticism.
UHMMM, this test was pretty rad. 
This stood out to me the most though. I think everything we’re hearing online is that people want AI to accelerate and be a layer atop human knowledge, so it’s nice to see that’s how this data is landing, too. Great work, crew!
If only someone could help with this. This was a great ‘state of the industry’ snapshot!
This seems to be a recurring issue. Cool that the survey highlighted this!
