AI + UX: Product Design for Intelligent Experiences

The speaker is a product designer who started their interest in AI while working at UiPath on a product called Clipboard AI, which won Time Magazine's Best Invention of 2023 award. They are currently designing for the collaboration space at Miro.

The speaker believes AI is losing hype because there aren't as many valuable use cases as expected, companies are losing trust due to lack of return on investments, and the industry is entering the 'trough of disillusionment' according to Gartner's hype cycle.

Challenges include working with bias in AI systems, ensuring safety and security, dealing with the unpredictability of AI outputs, building trust and transparency, and navigating issues of ownership and intellectual property.

Designers should start with people, fight bias, ensure safety and control, make technologies accessible to everyone, be accountable, promote good, help people grow, build trust, and design systems that age well.

AI can enhance user experiences by making products more relevant through personalization and context awareness, decluttering apps, making products more accessible, and being helpful by explaining itself better to users.

Potential pitfalls include overpromising what the AI can do, disappointing users with less-than-expected results, and losing user engagement if the AI-generated content is not up to par.

The speaker believes AI will not replace human designers because AI cannot yet figure out the problems to solve, requires multidisciplinary efforts to deliver solutions, needs a lot of guidance, and lacks empathy and a true understanding of human psychology.

The speaker believes AI will become an implicit part of product solutions rather than a special feature, helping to remove friction from workflows and enabling people to express their creativity even if they lack traditional skills.

Companies should experiment with AI technologies, look for meaningful opportunities to use them, and set realistic expectations internally and externally about what AI can and cannot do.

The 'intent-based outcome specification' paradigm involves giving the computer a goal, which it then understands, deconstructs, and makes decisions to create the desired output, reversing the traditional locus of control in design.