Day 3 Session Notes–Use AI to Drive Outcomes that Go Beyond the Design Sprint

— Thank you and get things started by thanking rest of curation team as today has been stellar and really compelling

• Dove-tail on final thoughts and being provided agency to experiment is invitation to innovate

— Focus on innovation in action

• Take through framework that operates likes recipe rather than formal equation

— Did great job introducing us

• Joe started business called Gyroscope and worked for many verticals and brands
• Pooria worked in AI for combined 20 years

— Pooria did Generative AI at Autodesk and showing off AI avatars and working to get hair to that level

— Why are we here? Kick off story-time with an origin story of guided discovery, walk through framework with drivers and outcomes, going through case study and project through completion and clearance to share

• Wrap up with pragmatic tips

— Why are we here? Kick off story-time with an origin story of guided discovery, walk through framework with drivers and outcomes, going through case study and project through completion and clearance to share

• Wrap up with pragmatic tips

 

— We will start with a bit of negativity and tragic origin story

 

— Few years ago working for a multinational CPG, and contracted in under external design work, and nothing usual of working on client projects with challenges

— Had perfect storm of bad outcomes

• Narrow scope, strong leader, while also having us look at systemic process issues
  • All to be done within a single week
• Throughout this, we were asked to color inside the lines

— Delivered design that was prototype and tested, but…

• Design sprint was wrong tool to use, and client was not aware of other methods working on
• Lot of frustration caused friction, and no one knew what to do next once project solved

 

— It even might have caused long-term harm

• Design sprints became scapegoat for anything that went wrong
• Old methods design was meant to change were hardened and put back into place

— Design thinking delivers, but is solely outcomes focused and has good elements and bad elements

• But if no value, why doing it?

— Sprint playbook needed improvement

— Wouldn’t be here if design sprints weren’t successful at tacking certain problems

• Things it can’t solve for

— Design sprints focus on high volume thinking, which is not always right solution, and even more so on tight timetable

 

— If value proposition of design thinking is empathy, we need data to do that

• Led to guided discovery process

 

— Focus on positive and what guided discovery is

— Recognize that we are taking individuals from multiple disciplines on a journey

• Giving tools and methods for meaningful outcomes on other end and important learnings established along the way

 

— Framing mechanism around this is AI

• Likely don’t recognize first interaction with AI was Google auto-correct, and it’s targeted ads

— AI now viewed as table stakes for every roadmap, whether as feature enhancement or capability improvement or M&A acquired or acqui-hired by company

• As discipline under the design umbrellas, we need to leverage tools to competitive advantage

 

— Overview of guided discovery framework and four modules of thinking and doing while a fifth module is dedicated to packaging

— Collection Module

• Shopping for data that’s available, and marketing has data on purchase intent, and buyer data as well
• Customer Experience to give feedback on reliability and utility of items in data
• Development has tons of instrumentation data and higher-level of utility

 

— Ability to align with partners and make educated hypotheses and 360 degree picture for virtual personas

• First two modules of guided discovery require iteration on virtual persona

— Following module is concept generation and then narrowing in and zooming out

• AI tools to manifest ideas at same level and then diverge or converge

— Accelerated timetable and taking prototype and sharing with virtual persona for directional feedback

• Virtual persona won’t supplant talking to people— but very good for directional feedback

— Refinement and running at very fast rate

— After evaluation high-level roadmapping exercise and scaling and scoping of outcomes and crawl, walk, run of delivery to outline next steps to take

• Dedicated module allowed for playing back understanding

— Take time to align stakeholders and clarify process, prior to packaging

— Creating operating model to understand and grasp subject matter expertise

— How is this unique?

• Virtual persona aggregates data inside org to provider better picture of what persona needs, behaviors, and goals are
  • Personas always need enhancement, and driving guidance of discovery
  • Data comes from different areas of organization and want to learn more
• Idea generation and using AI tools to diverge wider and level playing field from designers to non-designers
• Teaching partners to use different tolls and methods and documentation of what worked and what didn’t

— Landscape changing all the time, and not one way to deliver this, and encourage you to mix and match

— Reminder of five discrete stages in framework

— Had active project for live music application and niche customers in mind, i.e. sound engineers

• Don’t need to be sound engineers to think like them

— Have AI tools to generate concepts

 

— And here is overview of how data fed into AI, and data at rest

 

— Understand context and design sprint, and element that doesn’t get created and take from stakeholder users and buyers and valuable info to items made and feed context

• Share with client and iterating straight from get go

 

— Concept generation stage, but with Gen AI to dictate type of visual form for product to take and mood-boards from intention

• TLDR of using AI to direct images themselves

 

— Create a POC, and having AI to interpret sketcher and make it much easier for low fidelity mockups

• AI getting it wrong can lead to new elements for ideation

 

— Then make sure assumptions get made and correct. Take everything created and interacting with virtual persona

• Evaluation a lot faster and can talk to humans later in process
• Not replacement but augmentation to process

 

— Creating renders to have curated version of product intention and best elements of this to increase fidelity without effort, and better outcome then print method

 

 

— Actually focus on design, and use AI tools throughout design process

 

— Before final section of pragmatic tips. Can leverage AI tools for creative processes inside Figma and other tools, based on configurations

• More concrete than what client could share

 

— Here are tips you can use

• Newsletters snd substacks for reference
• Research blogs and non sci-fi sides

— Product releases and research papers and signal to noise in AI space, so take a screen cap

 

— Can find this without looking hard

• Check out billion dollar teams presentation and same philosophy here

 

— Always self-paced learning and one way to not let down your instructor and lots of free learning from finest universities in planet

— Most important things for AI tech is to learn to speak AI prompts, and don’t be intimidated by learning curve

— Here are tools we used in process—but this list has a lot of change and changed many times already and first two rows sticking around

— Since technologies always keep evolving, avoid dogma and keep iterating and try out new tools, once you have methodology scale it up

— Thank you for having us, and eager to wrap up