AI products are being built faster—but clarity is still elusive

According to the Figma AI Report 2025, one in three Figma users shipped an AI-powered product in the past year, a 50% increase from 2024​. Yet while more teams are embracing AI, clarity around goals remains a major challenge.

• Only 9% of respondents said revenue growth was their top goal.
• 76% cited vague objectives like "experimenting with AI" or "improving customer experience"​.

As Andrew Hogan noted during the interview, "There’s still a lot of experimentation happening. We’re early, and a lot of teams are feeling their way through."

Without clear objectives, it’s harder for teams to measure success and iterate effectively—a theme that mirrors the broader challenges of AI in product development.

“You can just do things—but should you?” — Andrew Hogan

Design is not an afterthought—it’s a strategic advantage

If there’s one major takeaway from both the report and the conversation, it’s this: Design matters more than ever in AI.

• 95% of designers and developers surveyed said design is at least as important in AI-powered products as in traditional ones.
• Over 50% said it’s more important​.

Andrew explained it clearly: “Design isn’t just polish anymore. It’s how these products win users, drive loyalty, and stand apart.”

In a landscape where agentic AI is rapidly emerging and UX patterns are still being invented, AI-powered UX isn’t about aesthetics—it’s about functionality, transparency, and trust.

“Design is for, 95% of the respondents, just as important—and for more than 50%, actually more important—when working with generative AI.” — Andrew Hogan

Agentic AI is reshaping user expectations

The fastest-growing category in the Figma AI survey? Agentic AI, which doubled year over year​. These are tools that don’t just assist users—they act on their behalf.

But agentic AI comes with new design challenges:

• How much autonomy should an AI have?
• When should users be alerted or asked for input?
• How do we avoid users feeling a loss of control?

As Jason Giles pointed out during the episode, "The urgency of getting involved in shaping AI experiences is real."

Designers and developers must rethink traditional UX assumptions when working with agentic AI—it’s no longer enough to make interfaces intuitive; now they must also build trust and maintain user agency.

Successful teams embrace iteration and flexibility

According to the report, the most successful AI teams share common practices:

• Rapid prototyping and willingness to struggle
• Strong design and development collaboration
• Frequent customer research and feedback loops

Andrew explained it like this: “If you struggled with prototyping a generative AI-powered product, you were more likely to be successful”​.
Struggle isn’t failure—it’s a necessary part of building something new.

Interestingly, teams that acknowledged their workflow for AI projects was different from traditional projects also reported better outcomes. As the report phrases it, success comes from treating "best practices as starting points, not sacred rules"​.

“Successful teams are the ones who reflect, adapt, and evolve as the tools do.” — Andrew Hogan

Productivity gains are real—but quality still lags

AI tools are boosting speed across design and development workflows:

• 78% of respondents said AI made their work more efficient.
• Yet only 58% said AI improved the quality of their work​.

And trust remains an issue—only about one-third of builders fully trust AI outputs​.

While developers report higher satisfaction with AI tools (83%) compared to designers (69%), this gap highlights the ongoing tension between efficiency and excellence​.

AI may help teams move faster, but critical thinking and human oversight are still essential to delivering high-quality, user-centered experiences.

Building better feedback loops with AI

Throughout the interview, Andrew and Jason emphasized the importance of integrating continuous user feedback into AI development processes.

Today’s faster prototyping cycles mean teams need equally fast validation methods to ensure they’re building solutions users actually need.

Andrew noted, “There’s a looping sort of meta-discussion—we’re building new products with AI, but those products are also changing how we work”​.

By embedding customer feedback into iterative AI workflows, teams can avoid the pitfalls of "building fast but wrong" and keep user needs at the heart of innovation.

The new mindset: adapt, experiment, and stay curious

In the episode, Andrew and Jason reflected on lessons learned from past tech shifts like mobile apps and the internet’s early days.

One key takeaway? Learning through experimentation matters.

“You have to try things,” Andrew said. “You have to experiment. You have to make useless apps... because then you find the good ones”​.

The future of AI product design belongs to those who:

• Embrace ambiguity and evolving workflows
• Stay curious and flexible
• Anchor their efforts in human-centered design principles

“The next chapter won’t be written by AI—it will be written with it.” — Andrew Hogan

AI is a tool, not a replacement

The overarching message from both the Figma AI Report 2025 and the Insights Unlocked conversation is clear: AI is here to stay, but people remain at the center of great product design.

Success in the AI era won’t come from chasing every shiny new model. It will come from building with care, curiosity, and a commitment to solving real human problems.

As Andrew summed it up perfectly:

“You have to try things. You have to experiment. Because that's how you find the magic.”

Episode links:

• Connect with Andrew Hogan on LinkedIn
• Figma’s 2025 AI Report: The year of AI at work
• Figma Config 2025
• Andrew’s Double Click blog post
• Connect with Jason Giles on LinkedIn
• How to build effective AI fast: This on-demand webinar explores how AI is changing the rules for software success and how that challenges us to change the way we work and the products we create. It includes practical real-world examples and videos of customers reacting to AI experiences.
• Design for emotion, not usability: how generative AI is reshaping UX: This blog post discusses how generative AI is reshaping UX by prioritizing emotional design over just usability. It offers strategies to create AI-driven experiences that resonate with users on a deeper level.
• Effective AI—how to choose the right generative AI features and build them fast: This guide provides actionable steps to help you get your AI development right on the first try, focusing on choosing and developing the right AI features.