product · startups · Technology

The “Break Things” Era is Over: AI’s Ethical Emergency

For far too long, the gospel of “move fast and break things” has dictated the rhythm of product development, especially in Big Tech. It wasn’t just a catchy slogan; it was a fundamental, often flawed, philosophy. User experience (UX) research? Too often relegated to a rubber stamp, validating predetermined instincts rather than challenging them. Behavioral analysis? A perpetual rearview mirror, confirming what users did, rather than anticipating what they shouldn’t have to do, or what profound influence our creations might wield.

This backward approach, ironically, now faces its ultimate test: Artificial Intelligence.

AI products aren’t passive tools. Large language models, predictive algorithms, and personalized recommendation engines don’t merely respond to users; they co-create experience. They shape behavior with an intimacy and scale we’ve never before witnessed. Yet, here we are, attempting to apply the lagging UX processes of a bygone era to the most leading-edge technology humanity has ever conceived. It’s like trying to navigate a hyperspace jump with a map drawn for a horse and buggy.

The Headlights We Desperately Need

The stakes are no longer just about usability; they’re about humanity. When design fails in the age of AI, the consequences aren’t minor inconveniences. We’re talking algorithmic harm, embedded discrimination, the rampant spread of disinformation, and a deep, systemic erosion of trust.

Consider the landscape: AI systems are increasingly mediating our most fundamental human experiences. They:

  • Personalize education, finance, healthcare, and justice.
  • Predict and influence our mental health, moods, and purchasing.
  • Mediate interpersonal relationships — from dating apps to social feeds.

To continue treating human insight as an afterthought in this context isn’t just negligent; it’s dangerous. We need a fundamental shift in perspective. UX and behavioral research must become the headlights of AI product development, proactively illuminating the treacherous road ahead. We can no longer afford to learn from where we’ve already crashed.

From MVP to MAP: Orienting Ourselves in a New Reality

The traditional product playbook preaches the gospel of the Minimum Viable Product (MVP): build something simple, get it to market fast, and learn from user feedback. A noble idea, perhaps, for a simpler time.

But with AI, “learning from failure” takes on a chilling new meaning. It can translate directly into:

  • Reinforcing societal biases at scale.
  • Violating privacy with unprecedented reach.
  • Misleading users into financial or emotional distress.
  • It scales misinformation or addiction loops with devastating efficiency.

Failure here isn’t just a costly pivot; it’s a profound ethical and societal liability.

This is precisely why we must abandon the MVP mindset for something far more critical: the Minimum Aligned Product (MAP).

A MAP isn’t just “viable”; it’s oriented. It’s built with intentional alignment:

  • Aligned with user values, not just their clicks.
  • Aligned with cognitive and emotional safety – a non-negotiable baseline.
  • Aligned with social, ethical, and cultural expectations – understanding context before deployment.
  • Informed by probabilistic models of user behavior before launch – anticipating impact, not just reacting to it.

MVPs are about iteration. MAPs are about orientation. One risks incremental improvements; the other guards against catastrophic misdirection.

Introducing HAI/UX: A Compass for Human-AI Insight and Experience

To operationalize this critical shift, we propose HAI/UX – a framework for Human-AI Insight and Experience. This framework elevates the role of research and data science from a supporting act to a central, guiding force in AI-driven product development.

  1. Ethics-Centered Experimentation: A/B testing, in its current form, can be a masterclass in optimizing manipulation. HAI/UX demands ethics red-teaming be woven into the very fabric of experimentation. We must proactively ask: Who might be harmed? What cognitive biases are we unknowingly exploiting? Is consent genuinely clear, or merely a click?
  2. Continuous Behavioral Forecasting: Forget static personas. We need to leverage large-scale, longitudinal behavioral datasets to predict user adaptation, identify emerging risk patterns, and flag ethical flashpoints before they become crises. Imagine, for instance, forecasting how patients might dangerously overtrust an AI medical chatbot under duress, then designing in deliberate friction to mitigate that risk.
  3. Probabilistic Personas: The rigid personas of traditional UX are wholly insufficient for AI’s fluidity. We must embrace personas as dynamic, probability fields shaped by context, time, and interaction with AI. A “young voter,” for example, isn’t a single demographic; they’re a complex probability field of disengagement, activism, conspiracy exposure, and curiosity—each activated by different AI nudges. Designing for this variance is paramount.
  4. Agent Co-Design: As AI agents evolve into co-actors in user journeys, we must pivot from designing for users to prototyping with them. Invite users to co-create with the AI: How should it express uncertainty? When should it ask for permission? Should it reflect user values or challenge them? This isn’t just empathy; it’s essential collaboration.
  5. Embedded Insight Pipelines: UX and ethical insights cannot remain quarterly reports. They must become live signals, monitored by engineering teams alongside latency and uptime. Design becomes a continuous feedback loop, not a retrospective analysis.

The Broader Implication: Building With People, Not Just For Them

This isn’t merely about tweaking product roadmaps. It’s about fundamentally rethinking how we build systems that impact human lives. HAI/UX shifts the paradigm toward:

  • Inclusion: Not a box to tick, but a dynamic, shared governance process.
  • Accountability: Researchers as proactive watchdogs, embedded guardians, not just detached observers.
  • Trust: Built not through slick PR campaigns, but through transparency, deliberate slow thinking, and a commitment to design justice.

The Call to Action: From Sprints to Stewardship

If we fail to evolve our product strategy, AI will undoubtedly outpace our ability to humanely manage its profound impact. The time for naive optimism or blind acceleration is over.

This means a collective re-orientation:

  • Funders must recognize and invest in UX and ethical research as core infrastructure, not disposable overhead.
  • Founders must treat behavioral researchers as product architects, not just focus group facilitators.
  • Engineers must learn to incorporate friction as a deliberate feature, not merely a bug to be smoothed away.
  • Designers must shift their fundamental question from “What’s the fastest way to get here?” to “What’s the safest and most equitable way to bring everyone with us?”

We don’t need to move slower. We need to move smarter. And critically, we need to move with humans firmly at the wheel, not tied up in the trunk.

Medicine · product · Technology

Future Pharma – Prescribing Digital Interventions to Drive Outcomes

If we know apps are deliberately built to get us hooked by stimulating chemical reactions in our brain, can that be used for good? Are there ways to create digital interventions that Doctors will eventually prescribe? I wonder what the regulatory hurdles will be? What type of companies will be capable of building this type of intervention. I foresee a massive market opportunity in the next 5- 10 years barring regulatory challenges.

We’re most likely in DI (digital interventions) 1.0 – If you look at applications like Talkspace, Noom, even Fitbit, they all leverage technology to deliver some sort of medical support/ intervention. But the future is even more interesting. Imagine an app that is optimized to stimulate the creation of certain chemicals in your brain. Maybe you’ve gone to the doctor, they’ve run tests, and they see you have a deficiency of said chemical. They’d prescribe an app that stimulates the creation of the chemical and let you know how often to use it.

I should wrap up by saying I’m not in the medical field so I have no idea how this would work. But I do know, product managers, especially product managers in the consumer space, are already thinking about how their products can form mental hooks to users to make products more “sticky” . It’s time we flipped this on its head and used it for good.

product

5 Ways To Make Your Product Development Process More Inclusive

In the early part of 2017, I had the opportunity to teach aspiring product managers at General Assembly. While we were on the user personas part of the curriculum, I got an interesting question from one of my students;

”It’s theoretical, but there’s something disquieting about creating these personas based on distilling, and how it could really leave a company never creating consumer products that are designed for a broader swath of person than white, single woman with disposable income. Is there a philosophy on how to best make sure that more people are represented?”

The product development process can sometimes become a silo and miss out on opportunities for other groups of users, potential features, and revenue. Based on my experience, here are 5 ways to ensure you’re building products and practices that help capture larger groups than the standard user persona allows.

  1. Focus on what needs to be done rather than who is doing it. The jobs to be done theory is the process a consumer goes through whenever they evolve themselves through searching for, buying, and using a product. It begins when the customer becomes aware of the possibility to evolve. It continues as along as the desired progress is sought. It ends when the consumer realizes new capabilities and behaves differently, or abandons the idea of evolving. (Learn more here) The jobs to be done framework moves beyond who the user is and focuses more on the future state of the user. While demographics and other user information is important, product teams should focus on collecting the appropriate information to build that bridge from current state to future state.
  2. Have a methodology you can look back to and evaluate. Companies should keep a metric/ or a process reserved for how effective they are at capturing fringe segments, feature requests, and/or products use cases. While it may initially seem like an extra chore for an already busy product team, it helps ensure that the team is accountable to dissenting features, products, and ideas. For example, a client I worked with back in the day had a 10% rule… The 10% rule stated 10% of their feature idea had to be fringe use cases. It helped the dev, product and marketing team think about particular types of users who may not be in the initial user segment.
  3. Employ for Diversity. Inclusive product teams (of experience, culture, background, gender, race, etc) is one the most important things you can do to insure you’re bringing inclusive products to market. The internal team asks the first set of questions, plans the go to market approach, and prioritizes which features get developed.
  4. Build diversity into the product feedback process. I’ve worked with a lot of companies who’ve outsourced ui/ux feedback to third party companies in the hopes of lightening the internal load. Feedback, especially early in the product development process, is best managed by your internal team. It also gives the product team a chance to focus on bringing a diverse group of users to the table for feedback.
  5. Don’t be afraid of the niche. Companies, especially in the early days, are afraid of building a niche product/feature. Building a niche product and getting market leadership is sometimes more important than getting mass market acceptance (also depends on company and industry). Sometimes being laser focused on a specific type of customer builds the bridge you need to the mass market and/or other niches.
product · startups · Technology · Why?

Design Thinking in the Nigerian Context

It’s 2 AM and the electricity goes out. Annoyed, you walk outside to turn the generator on. In the process of turning the generator on, you realize there’s no gas in the gen. In the pitch black of night, its not difficult to see the large red canister of gas a couple of steps away. You get the canister and start to attempt to fill the gen. You realize there’s too much gas in the canister and its difficult to control the amount that goes into the gen so you start to think… You go into the kitchen and get a used plastic bottle, cut it in half and now you’ve created a funnel and cup. You head back outside, put the makeshift funnel on the gen and begin to fill the cup with gas and pour it in the funnel. You fill the tank, turn the gen on, and go back to sleep.

That story is design thinking in action. I’d argue with anyone that design thinking is not a process that Nigerians are foreign to. In fact, it’s been at the ethos at most grassroots solutions. There are so many small inventions and quick fixes that I see everyday. Why don’t you see some of these solutions in the market? I believe the challenges are threefold:

  1. How do you get people to see their solution as valuable outside themselves?
  2. How do you provide the platform for people to producttize/ commercialize their already working prototypes?
  3. How do you protect ideas and create incentives for people to continue to create?

How do you get people to see their solution as valuable outside themselves? 

This is consequence of innovating to live vs innovating to thrive. People are brilliant problem solvers in developing markets because they have to in order to survive. Design thinking concepts tend to become a framework that most people operate in without knowing it.  The challenge is being able to think beyond the problem, which is challenging for the problem solvers. I would suggest getting up to the balcony like in this story below:

Let’s say you are dancing in a big ballroom. . . . Most of your attention focuses on your dance partner, and you reserve whatever is left to make sure you don’t collide with dancers close by. . . . When someone asks you later about the dance, you exclaim, “The band played great, and the place surged with dancers.”

But, if you had gone up to the balcony and looked down on the dance floor, you might have seen a very different picture. You would have noticed all sorts of patterns. . . you might have noticed that when slow music played, only some people danced; when the tempo increased, others stepped onto the floor; and some people never seemed to dance at all. . . . the dancers all clustered at one end of the floor, as far away from the band as possible. . . . You might have reported that participation was sporadic, the band played too loud, and you only danced to fast music.

. . .The only way you can gain both a clearer view of reality and some perspective on the bigger picture is by distancing yourself from the fray. . . .

If you want to affect what is happening, you must return to the dance floor.*-Ronald Heifetz

That often the most challenging place for problem solvers to get to but is central to seeing the value in an idea or a new process.

How do you provide the platform for people to producttize/ commercialize their already working prototypes?

This a more systemic and structural problem. With a lack of manufacturing and capital in developing markets, it’s often impossible to scale a new idea. I believe the improvement of technologies like 3D printing hold a tremendous opportunity to decrease the cost and increase the accessibility of manufacturing to the masses.

How do you protect ideas and create incentives for people to continue to create?  

This is an interesting challenge that all countries face now. How do you protect people who create, while encouraging the free exchange of ideas so people can build upon them? Legal spaces like IP and copyright may not be as developed in a place like Nigeria but it presents a great opportunity to re-imagine what IP/Copyright law can look like in the information sharing age.

Politics · product · Technology · Why?

Kobayashi and the Leader of the Free World

***Disclaimer…. I’m a huge Star Trek fan. I’ve tried to simplify a little bit so you don’t have to know as much about Star Trek to understand what I’m trying to say.***

Kobayashi Maru is a star fleet training exercise that is used to evaluate a commander’s character and fortitude. The simulation in the Star Trek universe allows the cadet to command a federation star ship, and sends them to aid another Federation vessel, the Kobayashi Maru. The disabled ship is adrift in the Klingon neutral zone, and the ship commanded by the cadet entering the zone will be in violation of a treaty and liable to attack.

The cadet has to decide whether to rescue the stranded ship, creating an opportunity for an all-out war with the Klingons and jeopardizing his or her own vessel and crew-mates’ lives in the process, or leave the Kobayashi Maru to eventual destruction. If the cadet attempts to save the vessel, the simulation is programmed to guarantee that his or her own ship will be destroyed. Not only will he be unsuccessful in saving the Kobayashi Maru, but everyone else will die as well.

The object is to test the cadet’s character and presence of mind in the face of large-scale disaster and certain death. The creation of the Kobayashi Maru isn’t discussed as much in Star Trek cannon, although in the most recent reboot, it’s shown that Spock was the preliminary designer of the test. His Vulcan sense of logic proved to be very helpful in constructing the no win scenario.

When deciding on leaders, humans traditionally follow our gut and how we feel about a person. The mental models an heuristics used to make snap decisions on who to follow are legacy from our early days when we had to be very cautious about who we were hunting and gathering with. We decide leaders based on what they say but even more on how we perceive them. Don’t believe me?  Take a read about JFK VS Nixon here.

While we can never get rid of the human perspective, shouldn’t we be responsible for aiding better decisions in who should be leaders? We should have our own Kobayashi Maru that we use to vet leaders where we can objectively see their character and fortitude. To be more specific, the president of the United States should be put through more than just public opinion to become president. We have the history of the world and technology to create all possible and future scenarios to test a candidate’s decision making skills. It’s not a heavy lift at all. Here are the steps:

  1. Recognize that we are currently incapable of making the best decisions without more information.
  2. Develop a Kobayashi Maru equivalent that runs through a week of various possible scenarios (domestic disasters, economic collapse, political brinkmanship,etc).
  3. Have the potential president pick their team.
  4. Run the simulations… Evaluate the results.

The hardest step is 1. Everything else is super doable. Our armed forces train just like this. I don’t think its too much to ask a potential Commander in Chief to go through similar training and evaluation.

There’s a lot of responsibility involved with picking the next leader of the free world . Citizens should look at a Kobayashi Maru like exam as an opportunity to improve our decision making by exposing the decision making process of our future leaders in life- like situations.

Or we can just watch them play The Sims.