Technology postsSaturday May 13, 2017
IMDb Needs to Fix Its 'Known For' Algorithm
I know. Not exactly high on the list of things the world needs to fix. Even so.
First, before anything else, think about what movies Steven Spielberg is “known for.” Think about four of them. You can only choose four. It's a tough task. He's made some of the world's most popular and beloved (and lately, critically acclaimed) movies.
As you probably know, on each of its pages for any movie/TV person (star, cameraman, casting, set decoration, best boy, doesn't matter), IMDb.com lists four movies he/she is “known for.” Robert Redord, for example, is known for, in this order, “The Sting,” “Butch Cassidy & The Sundance Kid,” “All the President's Men,” and “Out of Africa.” That's not bad. Seems about right. For Audrey Hepburn, it's “Breakfast at Tiffany's,” “My Fair Lady,” “Roman Holiday” and “Sabrina.” Again: Yep.
But there's a few bugs in the system. Here's what IMDb's algorithms say Steven Spielberg is known for:
Right. Not “Jaws,” “Raiders of the Lost Ark,” “E.T.” or “Jurassic Park.” Not “Lincoln.”
I get the first two. But “Catch Me If You Can” and “A.I.”? Really?
Then you notice the role they've assigned to him for each of these movies: producer. And it all begins to make sense.
According to IMDb, its algorithm assigns a “weight” to every credit in someone's CV. These include, among others:
- “The job performed on the title (a credit as director will have more weight than a credit as production assistant).”
- “The frequency of credits for a particular job in the context of the person's filmography (writing credits may have more weight for someone who is more frequently credited as a writer than as a producer)”
It's this second that's the problem with Spielberg. He currently has 162 credits as a producer. He has 56 as a director. Apparently this means that IMDb, the most popular repository of movie information we have, sees Spielberg, the most popular movie director of all time, as primarily a producer. And since Spielberg didn't become a producer until the late '70s and early '80s, that eliminates movies such as “Jaws” and “Raiders” from “Known For” contention. (Oddly, he was a producer for “E.T.” so I'm not sure why “Catch Me” and “A.I.” trump that.)
The same is true on IMDb with other great directors who produce. In this universe, Martin Scorsese is “known for” for being 1) producer of “The Wolf of Wall Street,” 2) director of “Goodfellas,” 3) producer of “Shutter Island” and 4) director of “The Departed.” “Taxi Driver” and “Raging Bull”? Whatevs.
Actors who produce? Tom Hanks is primarily known for ... “The Da Vinci Code.” Drew Barrymore is primarily known for ... “Donnie Darko.”
Algorithms are tough, and IMDb's gets a lot right. But there are still a few bugs in the system.
Quote of the Day
“In the United States, most banks take special precautions with their Swift computers, building multiple firewalls to isolate the system from the bank's other networks and keeping the machines physically isolated in a separate locked room.
”But elsewhere, some banks take far fewer precautions. ... The central bank in Bangladesh, by some accounts, employed fewer protections against cyberattacks than many other large banks. The bank, for example, used $10 routers and no firewalls, according to news reports.“
-- from ”Hackers' $81 Million Sneak Attack on World Banking" on the front page of the New York Times.
I've been asking this question of friend and strangers for a while now: Which jobs do you think won't get digitized away in 10, 20, 30 years?
According to Barbara Ehrenreich in her review of two books, “Rise of the Robots” by Martin Ford and “Shadow Work” by Craig Lambert, the answer is: Not many. She skips over the ones we've already lost (printer, photographer) to concentrate on the ones we're beginning to lose (secretaries, travel agents, customer service in general). She quotes an expert predicting that in 10 years, “90 percent of articles will be computer generated.” She writes about college grads floundering and the longterm unemployed giving up, and why that is:
All of this has happened by choice, though not the choice of the average citizen and worker. In the wake of the recession, Ford writes, many companies decided that “ever-advancing information technology” allows them to operate successfully without rehiring the people they had laid off. And there should be no doubt that technology is advancing in the direction of full unemployment. Ford quotes the co-founder of a start-up dedicated to the automation of gourmet hamburger production: “Our device isn't meant to make employees more efficient. It's meant to completely obviate them.”
Near the end of the piece, Ehrenreich, author of “Nickel and Dimed,” gets apocalyptic:
If middle-class jobs keep disappearing as wealth piles up at the top, Martin Ford predicts, economic mobility will 'become nonexistent': 'The plutocracy would shut itself away in gated communities or in elite cities, perhaps guarded by autonomous military robots and drones.' We have seen this movie; in fact, in one form or another — from 'Elysium' to 'The Hunger Games' — we've been seeing it again and again.
As I wrote five years ago, in a review of a different movie, we're all cutters now. We just don't seem to know it. Or how bad it'll get.
Step away from the job or the girl gets it.
This was my week to begin to work from home. It's been a steady increase—working from home. Ten percent of the time. Twenty percent of the time. But office space is now at a premium, particularly in downtown Bellevue, so Tuesday and Wednesday I moved the kit and kaboodle to my home office on First Hill, where I'll be working 90 percent of the time. I know: Nice if you can get it.
It necessitated making room, of course. That was something else I was doing Tuesday and Wednesday: clearing the detritutus off the bookshelves (books I'll never read, or read again), as well as cleaning and straightening up. A clean place, well-lighted or not, helps me think. There's still work to be done, but by Wednesday evening it wasn't looking bad, and Thursday morning I actually woke up with a sense of excitement. A new day!
And around 9 a.m. the internet went out. Really? I checked the router. Yellow light. I turned it off and on. I turned the whole thing off and on. I checked the cable TV—I get both through Comcast—and that was out, too. I called Comcast, and after about 10 minutes of various “press 1 for .../ press 2 for ...” hurdles, I talked to a poor customer rep—surely the lousiest job in the world. She told me there was an outage. In my building? I asked. In the neighborhood? In the world? She only knew “outage” but said it would most likely be fixed by 4 PM.
Wait, what? 4 PM? Most likely?
After this call, I went down to the basement of our condo building, where the cable-box is located, and lo and behold, a Comcast rep was already working on the problem. Nice! Except, no, he was trying to activate service for a new resident, and wasn't getting a signal. And knew less than I did. So I filled him in while he tried to reach somebody to confirm. It took a while. Apparently even Comcast technicians are put on hold forever when calling Comcast.
Eventually he found out that a colleague of his in Southwest Seattle was running into the same issue. Eventually I found out that a friend of ours, half a mile away, was also without internet. Soon, #ComcastOutage was trending on Twitter. But it wasn't until mid-afternoon that I got the full story via The Seattle Times:
An estimated 30,000 Comcast customers in the Seattle area were affected by an extensive outage Thursday caused by a construction crew cutting through a fiber-optic line in South Lake Union.
(That headline, by the way, used to read: DAMAGE TO FIBER-OPTICS CABLE CAUSES COMCAST OUTAGE. Now it reads: COMCAST SERVICE RESTORED TO THOUSANDS OF SEATTLE-AREA CUSTOMERS. The happy-ending approach to journalism.)
Anyway, that was my first exciting day working from home. Another reason—as if I needed one—why Charlie Brown is my patron saint.
(But Who is Alan Turing?)
The first time Alan Turing was referenced in The New York Times was 20 years after his death, and his name was just a passing reference in an essay by Guy Davenport over the shameful obscurity of another man, the Canadian critic Hugh Kenner, whom Davenport assumed, or hoped, future generations would admire more than the rabble in 1973. Of Kenner's 1968 book of literary and cultural criticism, Davenport writes:
“The Counterfeiters” is a lesson in how to see. Not how to see surfaces but the inside of things and the astounding affinities of things which heretofore seemed to have nothing to do with each other. Vaucanson and Yeats, for instance (but who is Vaucanson?), metaphysical poetry and Babbage (but who is Babbage?), Buster Keaton and Alan Turing (but who is Alan Turing?).
Do first references get any better? A writer mocking critics for not knowing the name of a man the publication he's writing in—and which calls itself “the paper of record”—has never printed.
At this point, even to Kenner, Turing was a genius and/or “eccentric” mathematician and no more. The story of Ultra, and the Enigma machine, didn't break until a year later, 1974, with the publication of F.W. Winterbotham's “The Ultra Secret.” That story, which makes up most of the Oscar-nominated “The Imitation Game,” was unknown but to a few. Indeed, in the mid-1960s, in a long essay on computers in The New Yorker, in which Turing is liberally mentioned, we get this puzzled query near the end:
Turing's story has gained traction as computers became part of everyday life and mainstream culture became more accepting of homosexuality. In the 21st century, when you first hear his background, it seems impossibly dramatic. Wait, a father of the modern computer? And gay? And builder of the machine that broke the code that brought down the Nazis and saved millions of lives and potentially all of us from speaking German—or not speaking at all? And persecuted for his homosexuality after the war? Despite saving all of those lives? Why have I not heard of this guy before?
Derek Jacobi was the first to play Turing, on a BBC-2 TV series called “Micro Live,” in 1983. He was also the second, in 1996, recreating his Tony-nominated performance from the 1987 Broadway play, “Breaking the Code.” Benedict Cumberbatch is the seventh. There will be more.
Derek Jacobi as Alan Turing, with Claudius' stutter.