Copy
View this email in your browser

Dark Matter: 25 July 2018

I'm writing from a hotel room in Skowhegan, Maine — where, tomorrow, I'll make my own pair of shoes by hand. 

The most-clicked links in last week's Dark Matter:

  1. Fathom's visualization of Russian paid social targeting, Fakebook
  2. Rodney Brooks' brilliant history of AI and the rise of superintelligence
  3. TensorFlow's browser-based machine learning examples, called Seedbank

This week in something called 'quiver doodles':

This week, begin with Leisa Reichelt on why she — and the teams she leads — still begin their work on walls rather than on screens:

Digital things look ‘finished’ too soon. when something is a work in progress on a wall, it looks unfinished, so you keep working on it. moving things around, reshaping things, connecting things, erasing things, and making them again. Walls make it easier to iterate. Iteration, in my opinion, is massively correlated with quality.

I have strong feelings about walls, and their role in iteration (mostly of my own perspectives, I'm afraid), and moving from wall-cultures to an environment in which relatively few permanent vertical surfaces exist has been a challenge for me. For now, I'm using windows.

* * *

Venkatesh Rao is playing with better cards than the rest of us: his latest post, on modeling 'entropic ruin' ("predictably dying faster than you need to") to strategic outcomes is fascinating, though you'll likely need a few read-throughs to comprehend it in full. Be sure to read the whole thing, friends.

* * *

"No one really wants to understand your service" : the slides from Rachel Coldicutt's talk, 'From Usable to Understandable to Responsible', are absolutely wonderful. Spend a few minutes, and absolutely give her a Twitter follow, kind reader. Hers is one of the most compelling voices in UX right now.

This week in 'look what I built with scikit learn':

Almost certainly my first link to a piece in Confectionary News: reader and dear old friend Brianna points to a piece outlining Mars Co.'s investment in satellite technology designed to help small, independent cocoa farmers better predict weather patterns and allocate limited resources. It's a rather good example of the 'brand actions' we hear so much about and see so scarcely. 

This author acknowledges that Mars Co. almost certainly does all manner of horrible things, and that you need not flood his inbox with examples. 

* * *

Much of the language around the workplaces of the future has focused on the promise of symbiotic relationships between machines and humans — the idea of 'man plus machine' has become its own tired trope in some circles. Then, this week, in IEEE Spectrum

Many robots lack a humanoid appearance and thus may not be able to make use of such cues, yet it is still critical that such appearance-constrained robots communicate their intentions to users. I think automobiles are a good example of this where we have a technology that we commonly use to communicate to other people, but in a different manner than traditional social communication.

* * *

Irving Wladawsky-Berger writes for the grand old web: Typepad, rich with links, still uses canonical categories. He's also a joy to read. This week's post is especially good, citing a paper by professors Erik Brynjolfsson and Tom Mitchell, identifying eight criteria for tasks that are especially well-suited to machine learning applications:

  1. Learning a function that maps well-defined inputs to well-defined outputs 
  2. Large (digital) data sets exist or can be created containing input-output pairs
  3. The task provides clear feedback with clearly definable goals and metrics
  4. No long chains of logic or reasoning that depend on diverse background knowledge or common sense.
  5. No need for detailed explanation of how the decision was made
  6. A tolerance for error and no need for provably correct or optimal solutions
  7. The phenomenon or function being learned should not change rapidly over time
  8. No specialized dexterity, physical skills, or mobility required

Strong recommendation that you spend 6-7 minutes with the entire piece. 

* * *

Have a look at Abhishek Singh's work, applying TensorFlow to training an Alexa to respond to sign language. Too often we allow the superintelligence narrative to overshadow the mundane brilliance of small, perfect ideas like this. 

This week in using python to unload on Boston:

Geoff Boeing, urban planning rock star, at it again: deriving the complexity of navigating major metro areas by evaluating the distribution of city street orientations (yes, I promise this is good):

Most cities’ polar histograms similarly tend to cluster in at least a rough, approximate way. But then there are Boston and Charlotte. Unlike most American cities that have one or two primary street grids organizing city circulation, their streets are more evenly distributed in every direction.

* * *

From Racked, news that 'cam girls' are using Amazon wishlists to enhance their earnings from dedicated fans:

As far as the internet goes, she says she makes better (and faster) money stripping, but Amazon wish lists offer a convenience the IRL world can’t. Namely, a surprise Dyson vacuum, expensive skin care, and, in one case, catnip bubbles.

* * *

DevelopmentAesthetics is a tumblr chronicling the gentrification of London's once-affordable neighborhoods through the aesthetics of banal 'coming soon' condo signage — and it would be a touch brilliant were it not so overwhelmingly sad.

* * *

Finally, Radiolab has been releasing an extraordinary series of episodes on the nature of gender, Gonads, that are absolutely worth your time. The episode titled 'X&Y' is especially good.

Until next week.

Copyright © 2018 Dark Matter, All rights reserved.


Want to change how you receive these emails?
You can update your preferences or unsubscribe from this list.

Email Marketing Powered by Mailchimp