Major platforms add video generation capabilities while Perplexity abandons ads to preserve user trust, highlighting key UX decisions in AI product evolution.
Today in AI Products
 Vercel AI Gateway |
Feb 19 |
Video generation models now available through unified API
Vercel AI Gateway added support for multiple video generation models including Google's Veo, Kling, and Grok Imagine Video. Developers can now generate cinematic videos with synchronized audio through a single API interface, currently available for Pro and Enterprise plans. Source →
Designer's Takeaway: Consider how video generation capabilities change content creation workflows - users will expect seamless integration between text, image, and video outputs rather than separate tools for each media type.
Pattern: Multimodal Interaction
 Perplexity |
Feb 18 |
Drops advertising model citing AI trust concerns
Perplexity announced it's phasing out sponsored answers and ads from its AI search platform, with executives stating that advertising could undermine user trust in AI responses. The company is shifting to focus entirely on subscription and enterprise revenue models. Source →
Designer's Takeaway: Notice how trust becomes a primary design consideration in AI products - traditional monetization patterns like ads may conflict with user expectations of unbiased AI assistance.
Pattern: Trust Calibration
 GitHub Copilot |
Feb 17 |
Claude Sonnet 4.6 now available for code generation
GitHub Copilot users can now access Anthropic's Claude Sonnet 4.6 model for code completion and generation. This gives developers choice between different AI models within their existing workflow, allowing them to select the best model for specific coding tasks. Source →
Designer's Takeaway: Apply this multi-model approach to design tools - give users choice between different AI capabilities rather than forcing them into a single model's strengths and limitations.
Pattern: Adaptive Interfaces
| Vercel |
Feb 17 |
Case study shows AI testing agents ship in hours, not weeks
Stably's 6-person team demonstrates how modern AI development platforms can dramatically reduce deployment time for AI testing agents. The case study highlights how removing infrastructure complexity allows teams to focus on core AI functionality and rapid iteration. Source →
Designer's Takeaway: Consider how reducing technical friction in your design tools can unlock similar productivity gains - the easier it is to prototype and test AI features, the more designers can focus on user experience refinement.
Pattern: Progressive Enhancement
 Anthropic & Infosys |
Feb 17 |
Partnership targets AI agents for regulated industries
Anthropic and Infosys announced collaboration to build AI agents specifically for telecommunications and other regulated industries. The partnership focuses on creating compliant AI solutions that can operate within strict regulatory frameworks while maintaining effectiveness. Source →
Designer's Takeaway: Apply regulatory thinking to AI UX design early - consider how compliance requirements, audit trails, and safety measures can be built into the user experience rather than added as afterthoughts.
Pattern: Responsible AI Design
Today's Takeaway
Trust-First Design in AI Products
Perplexity's decision to abandon ads reveals how trust considerations are reshaping AI product strategy. As AI becomes more integrated into user workflows, maintaining credibility and user confidence takes precedence over traditional monetization approaches. Design decisions that prioritize transparency and user benefit over revenue optimization may become competitive advantages.
Want to learn more about the patterns mentioned today?
Explore All 28 Patterns →