Hà Phan

Yeah I have this voice in the back of my head saying, how many more years can we be employed?

I feel sad over all for all the people recently impacted by layoffs. There is a physics here that we can’t overcome.

It teaches you to pay attention to nuanced behaviors as cues to different mental models. The other important piece is learning to design the experiment to get users to reveal what’s in their head vs validate what is there. That’s a hard thing to explain but it’s very powerful.

My boss likes to quote me specifically about the way I approach research: “Design the research to prove yourself wrong.” This means to provide a version to falsify your assertion. The trade off has to be clear. The comparison will clarify the nuances of the motivation, and the decisions that follow.

Not sure. There are some people who instinctively already think outside of the box and coaching them on research thinking sharpens that edge. For me, that instinct was cultivated by conducting a lot of first hand behavioral research studies in highly nuanced problems.

I’ve often been asked, “How did you know to ask that question?” I can’t answer it. I think I instinctively see the nuanced behavioral hypotheses. For instance it’s different to design for exploration of a 3D space vs designing for efficiency of navigation. I actually think that’s creativity.

An instinct you develop over time is to get a sense of the political risk in the various stages of an initiative. The political context always informs the strategy. This is why conviction is important but more importantly it informs situations you want to sidestep, problems you don’t bet on.

Something that is unique to working on Rich Media emergent tech teams at Zillow is that you get to see inside your colleagues’ homes because many of us at one point or another has captured our own home for experiments.

2 key ideas I had that led to multiple inventions was not something AI came up with. One of them came from me thinking about it deeply then building a quick demo to show engineers. The 2nd came from an insight where I realized the time users actually spent on an experience & film making principles.

Provocation is not the product of alignment.
Not every problem demands provocation or disruption.
But every invention begins with it.

I often tell people that I have massive ambition when it comes to bold ideas, but not when it comes to titles. Sometimes the bold ideas work out and result in career growth but I’m not motivated by titles. If it works out, it works out. I believe that the work should speak for itself.

Yeah, at times it feels exhilarating to still be part of the race, and other times you recognize that the goal for every business is to try make your job obsolete.

There’s a part of me that marvel at all that AI can do and another part of me that grieved for the society that we are creating.

It’s unlikely they will make it available for the public. The work is protected by patents.

I presented our work on stage this week & I told my team mates afterwards that if I accomplished nothing else in my career - even if it never launched, I’d still be happy. I am so proud of the work but I am prouder of our journey, and since it is an invention, patents outlive digital products.

Sometimes I think that my ability to think outside the box is entirely due to the diverse experiences I’ve had in my life. Some of them could be seen as disadvantages, bc I didn’t follow the prescribed path. But when you have nothing to lose you don’t care.

Sitting next to my Eng Mngr who has a PhD from Harvard and a table full of well accredited people on our innovation team makes me feel a bit unworthy. They’re amazing but I am somewhat of an outlier in a sense that I went to a state school and am entirely self taught.

At an offsite retreat this week, and it’s great to see engineering and science adopt the word UX in the context of AI and automation. IMO, it is perhaps the most powerful framing of UX.

The smartest people I know are the ones who can confidently articulate the concise unknowns to be proven. They don’t assume to know better. That way of working also enable others to join as thought partners.

My role changed recently. Before when I led teams that worked on Search & Recommendations, the utilities & capabilities were highly specific. This team can have broad reach. So at this altitude, I need to quickly learn the larger org. Everything is a research problem at the beginning. TBC

There are lots of rituals, operational tools & processes you build to enables R&D and experimentation. Things like, configurable prototypes to communicate what’s under the hood, easy research recruiting process, a common language around testable hypotheses and prototyping, and a show vs tell culture

Me to team mate: What you’re working on is not the UI. You’re working on automation should do, which is largely invisible. It’s like the cruise control for the UI. You think you’re driving but it’s doing much of the steering for you.

You bound the goals from a hypothesis POV, whether it’s optimization or new value. You’re always defining a concise behavior within a context to yield a certain outcome that drives specific goal. The goals can scaffold. You identify the trade off in the hypothesis.

What you should align on is goals—testable goals. Those goals define trade-offs. Decision-making is about being explicit about both, bc there are always trade-offs. If you’re vibing through decisions without this, even perfect vibe alignment won’t help. Every decision after that will be a battle.

I’m changing roles and someone asked me to write down all my anticipated tasks for an emergent tech project I’ve been on. It’s a loose list of UX proofs and you don’t know what you don’t know til you get there, and those proofs require a background of the scaffolding of the other proofs.

I have done this with AI but a skilled human is more fun to collab with. There are emotional high points, relationship building and landing on a common religion so to speak that you don’t get with AI.

Aside from building prototypes w/scientists and engineers, last couple of weeks I spent time co-designing and prototyping with another designer. We riffed on each other’s ideas, and we discussed trade offs, why this decision over that. Having a partner to push against moves things forward quickly.

I heard a guy on TikTok say that he told his partner that should something happen, he’s willing to die to protect innocent people. It was casual & matter of fact how he said it, like he was going to stop by and pick up dinner. Choices in existential times.

I told a leader that they don’t need to learn everything. They need to understand the bets in their domain so they can influence the strategy and outcome, then rely the senior ICs to go deep.

Process scales behavior. Culture scales trust and conviction. One enforces rules; the other creates instinct. Every enduring product carries the fingerprints of the culture that made it.