Month: April 2017

Neuralink

wizard

Interrupting the normal schedule, here’s a thing that I found interesting (long post warning): http://waitbutwhy.com/2017/04/neuralink.html

 

Advertisements

Games for Testers: Sherlock Holmes Consulting Detective

SHCD
Elementary

So I’m into board games (no – not monopoly, real boardgames) and this is my current favourite.  Actually anyone with a sciencey investigative mindset will love this game.  From now on, anyone who asks what could be fun or interesting about testing is likely to get a copy of this thrust in their face as this really embodies the fun of poking around and investigating a mystery until you understand it.

In short, you and your friends take on the role of the Baker Street Irregulars.  Holmes introduces you to a new case with some initial thoughts and leads, and then you’re on your own to explore London and track down information from a huge massive choose-your-own-adventure-style book.  Except there’s no “turn to page X or Y”.  You can go anywhere.  Where do you want to go?  What’s useful?  That’s for you to decide.  And you only stop when you want to.  At which point you go back to Holmes Study, answer a quiz (which you don’t get to see before you declare that you’re done) on what’s been going on  and then Holmes grandstands and explains the actual answer and how he worked it out (inevitably far faster than you – but the information was all there for you to see).

You win and lose as a group, and while you can play it alone, a group of 4-5 is great, because the game rewards knowledge, thinking, understanding and the sum of investigators thinking and riffing off each other is vastly better than the individuals.

For a longer and better (and slightly surreal in places) review of the guts of the game, try this.  It’s worth a watch.  Go on.  I’ll wait.

But this isn’t just a great game.  The things that make it fun are exactly the things at the core of testing.

  • At any point, you have control over what you do, where you go, what you decide to investigate, and you make the judgement of when you’re done understanding the world.  More (correct) understanding means more points at the end of the game.  But each extra test costs you a few points.  So the key mechanic is the judgement of when to stop, when to search wide, when to search deep.
  • It rewards diverse approaches, thinking laterally, and the ability to quickly create and abandon many hypotheses.
  • You’ll inevitably make a model of how the game works and make assumptions based on that – and then discover your model was broken.
  • Sometimes the problem is that you don’t know what you don’t know.  Sometimes you know exactly what you don’t know and the cleverness required is in constructing a way to find it.
  • While there’s one main mystery to be solved in each case, you can also uncover a whole load of other things (which get you more points).  There’s an unknown number of things to find – and some of them aren’t worth the time and effort required to find them.

 

If you buy this (hint: you should), here’s one steer that would have made our first case much better:  Just like in testing, going to too few places and getting the wrong answer is much worse than doing more digging and getting a right-er one.  Typically, investigating finds other interesting things (with bonus points) and the game has a lot of detail to enjoy, so its definitely worth going to more rather than fewer places.

Now go and have some fun.

Books: Lean In and Lean Out

 

 

I’ve just finished a couple of books and accompanying criticism.  Given the context of a recent events, it seems as good a point as any to write about them.  “Lean In” and “Lean Out” (actually Lean Out is a collection of essays).  Both are about dealing with sexism in tech.

Lean In, I found interesting.  A bunch of thoughts and examples where women miss out not through someone being explicitly openly sexist, but through a load of continual low-grade things that might bunch go under your radar where “things people do that we recognise as good development” tends to align much better with “things men do” than “things women do”.  A load of examples added to my mental list of  things to actively look for to help people (and also incidentally a bunch of things to look out for, to help anyone who has lower confidence).  Lean In isn’t a magic fix, but is definitely a useful read for anyone trying to work on the problem “The tech company where I work seems basically OK on gender, but I know about subconscious bias; I wonder what we can and should be doing better?”

Lean Out, was much more sporadically useful – which is reasonable to expect from a collection of essays.  The biggest thing I took away from it was that my prior assumption that most of tech is like where I work is apparently flawed…

I was reading these books in the context of worrying about improving a culture at work where we’re rationalist, try to treat everyone the same, but social norms and subconscious bias work against that.  “Lean in” provides good thoughts that help on that, working against the biases and over time creating a culture where those biases are fewer or don’t exist.

It seems that that’s not “normal” in at least a significant fraction of tech companies (from  what I’ve read it seems that this is especially true start-ups or companies that have grown fast from start-ups and not had time/effort put in to develop their culture).  In a company where the culture is be a bunch of total arseholes as described in several of the essays in Lean Out and in wider news recently, I can see and feel why it seems reasonable to be angry and shouty and go down the “don’t fix it – bail and start again” suggested by various of the Lean Out essays.  For the record, it shouldn’t be for “feminists” to hold arseholes to account.  They need to be held to account by any humanist.  It’s not (just) about not being sexist.  It’s about not being a dick.

That’s not to say, I’ve stopped worrying about how to improve our company culture (Just because we’re miles over the “Not being full of arseholes” bar, doesn’t mean we should get complacent!)  More, that Lean Out in particular helped me see and understand where some of the anger is coming from.