Hacker News new | ask | show | jobs
by zug_zug 3098 days ago
tl; dr:

1. Author observes that he's witnessed a lot of emotional backlash to the suggestion that soft skills may outweigh hard skills in tech; suggests the emotionality of the reaction is indicative that engineers worry it just may be true.

2. Justifies the importance of soft skills by citing personal experience of teams of 100s+ people needing to be coordinated

3. Proposes that a [software] system's interaction with the outside world is a larger problem than the system itself [i.e. UX tougher than code]

4. Offers a metric to assess management by: if people dislike managers then those managers must have very low soft-skills. Proposes the traditional approach to management is a large failure in tech.

5. Describes soft-skills but gives no hard examples

6. Proposes creating vocabulary to represent soft skills such as requirement-gathering so that people can care more about these skills

---

My aside: I think 1/2 largely depend on the product, some products are almost entirely technical (alpha go, bitcoin), we'll only ever reach agreement by not oversimplifying the question.

I agree on 4, but I really think 5 is crucial. I think everybody has a very different idea of what "Soft skills" are and how well other people think they do at those skills. For example, I think somebody who creates laughter at work is a huge asset, others may not.

2 comments

On (2) -- I picked a personal anecdote, but I've seen this be very true on much smaller teams, as well. (Including, as it happens, AlphaGo! Think about what went in to creating the circumstances where it could be tested... they did some heroic labor)

On (3) -- not just UX! Interaction with the outside world is a feature of every aspect of how people will use (or misuse) the system.

(5) Totally agree that this is crucial. I did give a few examples (later in the article) but I think that coming up with a good taxonomy of these skills, and patterns for learning and teaching them, is a huge challenge going forward.

Many intuitions are useful: for example, you note that some people want more laughter at work, others don't. A good leader should be able to make that meta-analysis, figure out what the team needs, and create that. But needs to do what? How do we describe the (positive) effects of laughter that you're trying to capture?

Thanks for the tldr!

> 3. Proposes that a [software] system's interaction with the outside world is a larger problem than the system itself [i.e. UX tougher than code]

I hate this mindset that tries to divorce UX and code. If you're making a CRUD app, don't think of it as a tiny pure nugget of reusable API goodness surrounded by an unfortunate layer of messy UI that has to deal with the real world. The app IS that messy UI code. The rest is just a database schema. Don't try to abstract away your job.

Of course UX and code are different, but I think they're inseparable and anyone doing either should have a firm grasp of the other.