Major AI products are making strategic pivots around trust and creative workflows, from Figma's code-to-canvas integration to Perplexity abandoning ads to preserve user confidence.
Today in AI Products
 Figma |
Feb 17 |
Figma launches Claude Code to Figma integration
Figma announced that the future of design is "code and canvas," introducing Claude Code to Figma integration to give builders more choice in how they create. The feature allows developers and designers to move seamlessly between coding and visual design workflows. Source →
Designer's Takeaway: Consider how your design process could benefit from tighter code-design integration, and explore tools that bridge the gap between technical implementation and visual creativity.
Pattern: Augmented Creation
 Perplexity |
Feb 18 |
Perplexity abandons advertising model to preserve user trust
Perplexity AI has completely dropped its advertising plans that were introduced in 2024, citing concerns that ads could undermine trust in AI systems. The company is now focusing entirely on subscription-based revenue, making a significant bet on maintaining user confidence over ad revenue. Source →
Designer's Takeaway: Notice how monetization strategy directly impacts user trust in AI products, and consider designing clear value propositions that align business models with user interests.
Pattern: Responsible AI Design
 Gemini |
Feb 18 |
Google adds music creation with Lyria 3 model
Google launched Lyria 3 in the Gemini app, allowing users to create custom, high-quality 30-second music tracks from text prompts and images. This represents Google's expansion into creative AI beyond text and visual content. Source →
Designer's Takeaway: Apply multimodal input patterns when designing creative AI tools, and consider how different input types (text, image, audio) can work together to enhance creative expression.
Pattern: Multimodal Interaction
 Vercel |
Feb 18 |
Vercel redesigns runtime logs with visual filtering
Vercel launched a redesigned search and filtering system for runtime logs that converts text filters into visual pills. Complex queries with multiple filters become scannable and editable without retyping, making log exploration faster and more intuitive for developers. Source →
Designer's Takeaway: Consider transforming complex text-based inputs into visual, scannable elements that users can easily understand and modify at a glance.
Pattern: Progressive Disclosure
 Anthropic & Infosys |
Feb 17 |
Infosys partners with Anthropic to build enterprise AI agents
Infosys announced a partnership with Anthropic to develop enterprise-grade AI agents using Claude models, specifically targeting telecommunications and other regulated industries. The collaboration aims to bring autonomous AI capabilities to complex enterprise workflows. Source →
Designer's Takeaway: Notice how enterprise AI agents require different UX considerations than consumer products, particularly around compliance, auditability, and integration with existing business processes.
Pattern: Collaborative AI
Today's Takeaway
Trust becomes a competitive differentiator in AI products
Perplexity's decision to abandon advertising shows how trust is becoming more valuable than traditional revenue models in AI. As AI capabilities advance, users are increasingly sensitive to potential conflicts of interest between business models and product reliability. Design decisions that prioritize transparency and user confidence over short-term monetization may prove more sustainable in the long run.
Want to learn more about the patterns mentioned today?
Explore All 28 Patterns →