In June 2009, the veteran lead pilot of Air France 447 was awoken from his scheduled nap to find his junior crew in trouble. Wiping the sleep from his eyes, he had three minutes to determine which of the conflicting alerts from the plane’s automated system to respond to, and what were false alarms.
He failed. Nearly 250 crew and passengers died in the Atlantic Ocean.
That’s from the easily most gripping chapter in economics journalist Tim Harford’s 2016 book Messy: The Power of Disorder to Transform Our Lives. That tragic story is a cautionary tale for our age, demonstrating the paradox of automation: “The better the automatic systems, the more out of practice human operators will be, and the more unusual will be the situations they face.”
How quickly could you look up from your smartphone if your autonomous vehicle alerted you that it had disengaged? Software, like many managers and orderly obsessives, wants a tidy world but the world is actually quite messy. That’s Harford’s message: Embrace the mess.
Below are my notes from the book for my future reference.
Notes below:
- Daniel Oppenheimer et al showed high school teachers who used unusual fonts had better performing students
- Bernice Eiduson followed scientists over careers: the top performers switched topics more than the others
- Gruber and Sara Davis argue that creative people work on multiple projects so often that it should be considered standard practice: “network of enterprises” brings new ideas from frequent breaks
- Soren Kieekegaard called this “crop rotation”
- Oblique strategies
- Mark Granovetter’s oft-cited 1973 paper on the strength of weak ties for job searches
- Paul Erdos gave an academic example of weak ties when he said “another roof, anther proof” (he’s also the one who famously said a mathematician is “a machine for turning coffee in the theorems.”
- Bonding social capital and bridging: sociologists like bonding and bridging
- De vaan, David Stark and Vedres mapped the social connections of game developers to see what paths led to the most successful outcomes: networks of teams
- Muzafer Sharif: famous social science study of two groups of boys pitted against each other as competitors after forming strong bonds as strangers
- Cass Sunstein’s work on progressive Boulder and conservative Colorado Springs: a political divide grew more when people were put with like minded people
- Irving Janis: group think
- Scott Page: “diversity trumps ability”
- Brooke Harrington: friend groups had more fun but performed worse on tasks as they sought to make everyone feel ok; diverse groups are more productive but had conflict
- Ingraham and Morris: social event tracking found executives used networking events to reinforce existing friendships, despite claiming they sought new connections
- Aldrich and Martinez-Firestone: most entrepreneurs aren’t actually creative, they mix with same people (and follow existing patterns)
- Bahns Crickett Crandall: small colleges force more diverse friendships, where large colleges allow us to find ideological twins
- Eli Pariser’s 2011 book Filter Bubble argues how the web allows us to further isolate politically
- Emma Pierson showed tweets about protests in Ferguson Missouri did not interact
- We need people who bridge and bond across social groups
- You don’t want “team harmony” you want “goal harmony”
- La Corbusier and Steve Jobs as design icons
- Author hates 5s efficiency model in workplaces (keep things clean and orderly)
- Haslam and Craig knight: office layout study in 2010
- The advertising firm Chiat/Day called their offices a campus before tech companies like Google and Microsoft (fittingly Google now occupies the former Los Angeles HQ of Chiat/Day)
- Leadership at Kyocera and other companies with lean-office culture hated employee knickknacks — but that’s how employees feel at home in otherwise sterile places
- MIT Building 20 was a famous collision of specialities in part because it had a baffling room numbering system and was rushed to construction. Influential cross-disciplinary work took place there, like cognitive scientist Jerry Lettvin’s seminal 1959 paper “What the frog’s eye tells the frog’s brain” (on what the eye processes) and Noam Chomsky collaborate with Morris Halle
- Offices should look like they way employees use them, as a “tool shop,” not the tidy way a boss dreams of them. Famous story of Steve Jobs wanting just two bathrooms for Pixar HQ to encourage serendipity but a group of pregnant employees got Jobs to backdown because they didn’t want to walk so damn far to the bathroom.
- Neuroscientist Charles Limb on improvisation: “it’s magical but it’s not magic”
- Tips from improv: habit of yes, listen closely and take risks
- In 1995, Faurie & Raymond published their left-handed fighting hypothesis, in which left handed people survived in society because fighting them is difficult
- Patterns of Conflict was a 1976 presentation by Colonel John Boyd outlining his theories on modern combat and how the key to success was to upset the enemy’s “observation-orientation-decision-action time cycle or loop”, or OODA loop
- Steven Kerr’s 1975 paper: “On the folly of rewarding A, while hoping for B “ (unintended consequences of metric chasing)
- In 1995, economist Peter Smith published a paper demonstrating the unintended consequences of public performance data, and tracked other unintended consequences from metrics
- Jeremy Bentham: “impossibilizing the knowledge” is when you track several metrics and can’t solve for one and so judgement plays a role. You don’t want one tidy number which can lead to “overfitting” (or over-correcting for that one number)
- Paradox of automation: “The better the automatic systems, the more out of practice human operators will be, and the more unusual will be the situations they face.”
- In three ways: hides incompetence; erodes skills and automaton fails at most challenging times
- In 1980, aviation expert Earl Wiener began questioning the limits of automation, leading to Wiener’s Laws, including: “Digital devices tune out small errors while creating opportunities for large errors.”
- Rebecca Pilske: Veteran meteorologists made their predictions and then used software to spot gaps; Newer meteorologists relied entirely on software, missing the chance at learning to spot anomalies.
- Rather than humans monitoring machines why not the opposite? Or at least machines must engage their human supervisors more often to keep them engaged and attentive.
- Micro biome : fecal transplant
- Jane Jacobs: diverse industry cities will last longer than one-industry like Detroit.
- “Specialized cities are fragile”
- Academic AnnaLee Saxenian, with a speciality in technology clusters, demonstrated that in 1994, Boston’s 128 Corridor was specialized in siloed companies like Wang, Raytheon and Sun whereas Silicon Valley was full of gossiping, catty, culturally diverse intermixing. The mindset was one of an “ecosystem” not one industry.
- Cesar Hidalgo’s 2009 paper on economic complexity: relationship between a diversified economy, a complex economy, and a rich economy. “Variety and sophistication go hand-in-hand.” 216
- “The economies that do lots of things tend to do most of those things very well.”
- The Big Sort book
- Planners like the simplicity of zoning but like an orderly forest it is fragile and prone to lack of dynamism (and disease). He compares to a German forest example
- Research Diedrick Stapel faked 55 lab results because everyone wants neat and simple results
- The Broken Windows Theory published by The Atlantic in 1982 was also an oversimplified tidy
- Steven Levitt rejects broken windows. Instead, he argued the decline in crime in the 1990s was attributed to five main factors (initially four and added a fifth): more police, more prison population, the decline of the crack epidemic and the legalization of abortion in the 1970s – taking lead out of gasoline in the late 1970s was added
- Fabian Waldinger: Nazi Germany chasing away brilliant Jewish and supportive scientists (Jon Von Neman, Einstein) cost them economic output
- Of Benjamin Franklin’s 13 virtues, it was ORDER that gave him the most difficulty
- David Kirsh: neaties and scruffies
- A messy desk gives you signposts of what you were most recently doing. It is a kind of physical to do list. It is intolerable to work in someone else’s mess because it’s like you’re sitting in someone else’s language (“sign posts for someone else’s journey”)
- Filers and pilers
- Noguchi system and simpler versions: keep moving what you use the most so you know what to leave behind
- Daniel Kirschenbaum: monthly planning was more effective than daily planning
- How Mathematician Chris McKinlay used OK Cupid
- The problem with tidy checklists for dating or jobs is that we often don’t actually know what we want
- Marcel Proust’s questionnaire
- Facebook expanding the “like” button wasn’t for users but for advertisers
- Brian Christian won the Turing Prize as human with three tactics: went off book, got personal and interrupted
- Peter Gray at Boston College: informal games for kids boost adult creativity
- Emdrup Junk Playground: no more dangerous than standardized playground