10,000 feet

Can I say something catty?  We need the first big GenAI flop.

Give me the Quibi or the Google+ of the GenAI era—the classic example of how a big team ignored user needs, didn’t really solve a customer problem, and had a big oopsie because of it. 

It’s not that I want to see a product faceplant; it’s that I truly think we need this moment, as an industry, to bring us back to our senses.

Because GenAI has a strange ability to make human-centeredness feel irrelevant now.  

I think this is because ChatGPT gooped the universe into thinking that AI is all-powerful. Or that hyperintelligence is here. It has started to feel like all you have to do is apply GenAI to a business problem to create immediate value, no design or human-centeredness needed.

Perhaps I’m projecting because, even after 15 years in this field, our discipline started feeling irrelevant to me. I even contemplated changing the name of my company, Team Human, because the human didn’t seem to matter anymore.  🤦‍♂️

It wasn’t until I started building with GenAI that I realized how silly this perceived irrelevance is. The technology is not all-powerful; LLMs really are ‘stochastic parrots’ that need to be carefully coaxed, refined, DESIGNED, to solve people’s needs/create unique value. This is what we all do best. 

So I guess, in summary, even in the age of AI, we are Team Human.

7,000 feet

I want to focus on the minimum viable human-centeredness, required in the AI era: 

1. You still have to find a real customer need to solve.

2. GenAI has to solve the need in a meaningfully better way than traditional tech. 

These two simple commandments have become our product development gospel; however, it’s not what companies are practicing today. 

There is a palatable fear of being “blockbustered.”

Boards are pressuring executives to figure out GenAI fast; consultants are preying on AI insecurity and pitching random use cases; companies are running panicked proof of concepts and shipping silly little features that no one will use.

We know how this ends. Without minimum viable human-centeredness, GenAI will not be valuable…for customers or the business. It’s a recipe for wasting investment, organizational energy, and excitement. 

Call me old fashioned, but do you know what I think is missing? A strategy, centered around user needs. 🤸

• Companies need to carve out a uniquely valuable position to play with GenAI—rather than throwing spaghetti at the wall. 

• Companies need a clear plan—to respond to the pivotal opportunity and to galvanize the organization amidst all of this AI uncertainty.

5,000 feet

I’m going to share our high-level approach that we’re using to help companies build a GenAI strategy.  

There are only so many details I can fit into two paragraphs (without this newsletter reading like an overbearing recipe) so please reach out if you’d like more detail.   

The gist is this! Instead of picking a random use case and firing it into the world, we methodically re-examine every customer need, big or small, through the lens of GenAI. This allows us to identify the most uniquely valuable needs to solve with the technology. 

Like we’re organizing a messy closet, we go through years of past decks and studies, to create a foundational needs map. This can feel overwhelming at first, but like spring cleaning, after you start it becomes quite enticing—and eventually therapeutic. 

Once we have a full picture view of the user’s needs, we go need-by-need and attempt to ideate a GenAI solution.

• …can GenAI help with this reporting need?

• …can it help with this customer service need?

• …can it help with this onboarding need?  

Often, there is a GenAI solution that solves the need in a meaningful better way—but sometimes there is not. Again, we don’t force a solution that doesn’t leverage GenAI’s unique capabilities.

At the end of this exercise, we’re typically left with 15-20 GenAI features, little windows into how GenAI interactions would work across the product experience.  However, instead of calling these features, we call these “skills.” (For more context, I wrote about  last week.) 

Ground  

15-20 skills does not make a foundational strategy; we need to carve out a stronger position. That’s why we evaluate every skill through two lenses: value and potential risk. 

VALUE i.e. If this skill reliably works, how much value does it create for a user? 

• We mock-up each skill as an animation that illustrates the functionality through a real-life scenario.

• We run a large-sample quantitative study to test relative value.

• We calculate significance and rank each skill by its score.

RISK i.e. if a skill misperforms, how severe is the potential harm and how strong is the mitigation?  Discuss the following…

• Direct risk | When an error occurs at this moment, how much harm can it cause?   

• Indirect risk | How vulnerable is this skill to privacy risks, security issues, bias reinforcement? 

• Mitigations | How strong are the mitigations we can take to prevent risk?

All of this discussion culminates in, yep you guessed it…a 2x2 matrix. Instead of looking at a list of 20 potential skills, this analysis helps us select right needs to solve (through the right skills).

• MVP (high value, low risk) | We prioritize these skills and move them to the proof of concept stage.

• Horizon 1 (low value, low risk) | We stage these skills after the MVP launch, in order to build on the problem area and deepen value.

• Horizon 2 (high value, high risk) | We push these skills to a later date, knowing they require greater reliability or infrastructure to minimize harm.

• Kill (low value, high risk) | We sideline or rework these skills.

I have a feeling this general approach feels right in your HCD heart, and I’m hopeful this can start to calm the insecurity that human-centeredness has reached its expiration date.

I will be sharing more details about our approach in upcoming weeks: including the importance of building problem-based capabilities (not just skills), and the value in creating a partner-based experience vision.

What an adventure we’re on; you can feel a new version of the world emerging underneath our feet. This newsletter is a new endeavor for me. And I’d appreciate your help. 

If you’re an innovation leader, hungry for a GenAI strategy, schedule a call. 

• We offer a 6-week GenAI sprint to build a human-centered strategy + vision  

• We also offer a 3-hour workshop to build AI intuition and begin crafting a partner-based vision. 

If you’re a designer, researcher, or practitioner, I’d appreciate you subscribing or sending this to someone who could benefit from this content. 

Have a human week,
Alex