My colleagues often mock me for my faith in human sciences, "Studying cultures and people in the age of algorithms, you just need data and machine learning model" they said. I've been an engineer by profession for a decade, but to develop new perspectives on the characteristics of our product, I interpret data points and anthropological input across information systems.

I try not to see people as numbers, but as individuals. Choosing a group and then understanding them should be the goal. All great companies are based on an insight into how to serve a group of people in a meaningful way. Looking at and listening to people is better than asking them.

My job is to listen, not just to people but also to the world they make for themselves, their tone, intentions, and moments of silence. Listening requires more than just registering sounds. It means decoding them, from the most distinct, to the quiet and the obscure. I stay in the unfamiliar for as long as something is meaningful. I’ve found that when exploring the unknown, learning to decode its sounds pays off.

Like learning a new language it involves building sensitivity towards the words, interactions, and the gestures which give rhythm to a culture. In my work listening is not passive. It takes risks to get real answers. The key for me is learning how to maintain an insider-outsider role - keeping up the rhythm of a conversion while at the same time being able to take a step back to interpret, probe, and hear what’s not being said.

The most exciting part of the work is its dynamic element. I’ve travelled around the world and seen my hypotheses turned on their heads. The critical factor - the linking piece often ends up being a surprise - is something I identify only when running back through the data and seeing it in the context of the whole. This whole is the entire experience of the culture, the people I’ve met, and the odd things that were mentioned in passing. Sometimes, one phrase can make everything else make sense.

Studying people is not a linear process. It is often blurry, ambiguous, and yields responses slowly. But I always leave the field feeling like I’ve seen something true. The reality of the place or a phenomenon dons on me later; not in a flash of insight but through the rigorous process of creating meaning.

New to LessWrong?

New Comment
5 comments, sorted by Click to highlight new comments since: Today at 8:26 AM

[mod hat] – This felt an edge-case for "should it go on frontpage?", and I might be convinced that it's a good fit for frontpage as-is. It felt a bit like it was in the genre of "personal blogpost" (this does not intrinsically mean it shouldn't go on frontpage, but suggests erring on the side of leaving it on the personal blog), and I got a vague twinge of "author is trying to persuade me of a thing."

Insightful feedback. I just started writing and it's a lot harder than I thought it would be. Thank you!

This was interesting, although it took me a full read through to get a sense of what sort of post it was. I'd be interested in a followup post that digs into specific lessons you've learned about how to listen.

(I recently read Design of Everyday Things which talks a bit about this, but was similarly light on "how to actually go about building a model on what people want/need based on observation.")

I'm half-way through Design of Everyday Things (with Thinking, fast and slow in backlog). I will definitely write a follow up.

[-]Elo5y20

You may like the keywords "warm data" on the topic of data that includes people.