Facebook and Cambridge Analytical Screwed Up, But Not How You Think

Facebook: Open for Advertising, Open for Business

There’s a huge kerfuffle over the recent revelation that Cambridge Analytica, a political research and advertising firm, used data from millions of Facebook accounts in order to build behavioral profiles and then, using those profiling models, advertise to people and manipulate their voting behavior.

People are upset over this, but, it’s not for the reasons that most people are talking about, that is, it isn’t that either Facebook or Cambridge Analytica behaved unethically. They didn’t. They behaved exactly as they were supposed to. The mistake they made was that Facebook users were reminded of how it all works.

What do I mean?

Let’s back up a minute, and look at both Facebook’s business model and also how advertising works.

Facebook is a social networking site, where you can connect with lots of people. That’s not a business model—there is no mention of how the rent is paid (salaries, their giant electrical bill, money to buy new servers and replace old ones). The business model is actually a very old one, one that’s been around for decades: advertising. As any old-school American television scholar will tell you, “If the product is free, you are the product.” This has been true for broadcast television and radio for decades. But as television offerings and the idea of “radio” have complexified, the idea “over-the-air” broadcasting has been somewhat forgotten: many people get their local, i.e., broadcast, television via cable, satellite, or the internet. The local airwave component is not part of that picture even though it’s still there.

The advertising industry, for decades, has wanted perfect consumer profiles for each and every individual. This way they can advertise to you perfectly. “Advertise” is just a polite word for manipulate. Yes, manipulate. Not influence, not coerce, but manipulate. It’s behavioral propaganda. Credit card data and frequent shopper cards were a good start, but the internet, once it reached the right scale, was a bonanza of behavioral data. With big data capabilities, perfect advertising may finally be within reach.

Most advertising that we think about is consumer-based, trying to get you to buy a product. Yes, you, buy this. That’s the idea. It is that basic. And political advertising is exactly the same: behavioral acceptance of a product. But in this case, your behavior is voting, not buying, and the product is a politician, not a good or service (that could be argued but doesn’t matter). It’s the same: advertisements try to manipulate your behavior to the advertiser’s desired outcome.

There is one other important aspect to advertising: that you not think about the advertising. And that’s where the failure is for Facebook and Cambridge Analytica. When you think about the advertising and how it really works, you realize you are the product. If you realize your vote, which is supposed to be sacrosanct, can be manipulated, then it throws into question what you might want to think about democracy and free will. (If you do not live in a democracy, this explanation probably seems short-sighted.) Capitalism is triumphing over democracy, and it’s not supposed to be that way.

Except, it is. Facebook thrives on a user-based advertising model where you are the product. All of your data and access to you is sold to advertisers, or, if Facebook can keep your data, Facebook itself can make all the behavioral models and then sell access to you (the advertisers will still need to make the “right” advertisements, but Facebook will provide the “right” people). Advertising works by showing you what will manipulate you into the desired behavior, be it buying or voting—they’re the same, it’s just a selection on your part.

Facebook will continue to make its service available for free to users, and can do so because the users are the product. Again, this is not new. Facebook sells access to its users to advertisers with precision targeting because Facebook has so much data about your behaviors and preferences, and although there are other financial models this is the one Facebook is currently using. Advertisers try to manipulate your preferences and behaviors based on your past preferences and behaviors.

None of this is new. (Suddenly, my media studies degree is looking pretty hot!)

Where they screwed up was that users remembered that they, the users, were the product, and that they, the users, were the subject of manipulation in the political domain where an individual’s voting preference is supposed to be pristine, determined solely by the voter, beyond manipulation and not for sale. Except, that isn’t true. Just most of the time we don’t want to think about it.

Addendum 1: Google's business model is also advertising.
Addendum 2: Zuckerberg's first website.

Google Search and Coffee

Google search, via Google Maps, gives some bizarre results for "coffee", as you can see with these three result sets. (And now Blogger is being horrible with layout, yes, I know, Medium or something, but I started this blog a long time ago, switching costs.)

Here the search is just "coffee", so, obviously looking for somewhere to get coffee (perhaps in liquid form). You see Starbucks, which I am not a fan of, and at this zoom level there is one other name that is showing and a few results where the name of the location isn't shown.

I want something without Starbucks (which is the next search), but also note there is nothing showing in the triangle made by those roads there in the middle/left.

So, the same search, with "-starbucks" which should, I have been told, exclude Starbucks. Except it didn't do that -- it removed all results but one. Although I hadn't zoomed in on that previous search, I can tell you that not all the results were Starbucks. We know at least one result, on the left, was for "Pavement Coffeehouse", which is not a Starbucks. This search result is completely incorrect and thus rather useless. Notice there are still no results in the triangle space there in the middle/left.

So maybe you were wondering, "This is Boston, where are the Dunkin' Donuts!" Good question. Google wasn't showing them, despite Google's description of DD: "Chain known for donuts & coffee". So, right there, AND COFFEE. Yet DD is not a Google Maps result for "coffee", as shown above. (Although the pin shows a fork and knife, not a cup and saucer.)



So, what is going on?

1. Google appears to have not associated "coffee" with Dunkin' Donuts, which is wrong.
2. Google seems to think that "-starbucks" means "get rid of all results for things Google thinks are like starbucks", which is also just wrong.

In conclusion, you can't use Google Maps to search accurately for coffee.

Post at Esoteric Gaming

Exciting news, Mark Chen just rolled out the second issue of Esoteric Gaming (along with his team, of which I am proud to say I am one), and I have a piece on EQ2, multi-boxing, and the prison server! Super cool, and all the articles are well-worth reading. You could head over there now!

The articles in the second issue are:

• The Guilt of Cheating
• Co-Playing with My Mom: Mega Man 2
• Celebrating a Portal to The Nether with Cake! Emergent Play in Minecraft
• Death of a Guild, Birth of a Child: Multi-Boxing and Self-Imposed Exile to the Prison Server in EQ2

I love the Minecraft/Portal mashup there in that one title.

More EQ2 Homage

I was looking around for some images of EQ2 to use, since when I played it I was just playing it (I rarely took screenshots), and found a great directory of old images posted by a gamer here: http://img.surbane.net/index.php?/category/13 -- here's one they took, cropped down to just the one window.

Hilarious! That is, if you know that A Flock of Seagulls was a new wave band in the 1980s and their big song was "I Ran (So Far Away)". EQ2 is so full of homage to culture, it's great fun in that way.

Some MMO/Gaming Blogs, and Leveling

I've run into some gaming blogs and posts I've quite enjoyed recently.

1. 18 Years Later, Why Are People Still Playing Ultima Online? The comments are where the heart of it is, although they go a bit off in the middle, stick with it.

2. Raph Koster's Website, and two of those pages, A UO Postmortem Of Sorts and Popular Posts.

3. The Ancient Gaming Noob, and two of his pages on in-game economies: Yelling and Selling in Waterdeep and, relatedly, Charting the Relative Natures of MMO Economies.

I want to pull a quote from Koster's UO write-up:

We also wanted to get away from levels...

For quite some time, I've felt that leveling is ridiculous. It's a useless, sparkling hamster wheel. In WoW and EQ2, the designers had to come up with ways to deal with it--eventually in EQ2, there was the "mentoring" system, where higher-level characters could mentor down in level to the level of a lower-level character for the purpose of grouping (thus the name mentoring). In ESO currently, they have essentially done away with leveling although it is still very much a part of the game. What I mean is that, your level is meaningless in terms of encounters--everything is, behind the scenes, scaled to max level--and so levels are meaningless for grouping (which itself isn't even needed much of the time). Yes leveling is still relevant for skills and gear, but it's just a shiny carrot.

(Edit: I just caught this post from Koster, Do Levels Suck? Really nice thoughts there.)

So I thought about it, and went back in my mind to the cannon for the mechanics of much of the fantasy-based computer gaming today: AD&D (let's be real here, 1st edition, which wasn't even called "1st edition" at the time since it was the only edition). This is true for single-player, single-player party (like Wizardry or Icewind Dale, or even multi-boxing in an MMO), multiplayer, and MMO scenarios. What was leveling like in D&D?

It totally made sense for players. No, wait, it was also horrifically problematic within the storyline of the game itself. Why?

Because elves. (Blame the elves, poor things.) Remember? They could live for hundreds of years, so in theory they could gain level after level, hundreds of levels beyond the capabilities and lifespans of their original human companions. Thus, the longer-lived races (like elves and dwarves) were level-capped, while short-lived humans had no level caps. This made absolutely no sense within the fantasy world, it was completely and solely because it would lead to 200 year old elves who were also level 200 magic users and would be like demi-gods, and that would imbalance the game. Of course there's a big "if" on that last one: only if DMs allowed it, which they didn't have to. Remember, in D&D1, IIRC in the DM's Guide, there was a paragraph at the back (or maybe the front) saying how these are not rules, they are just guidelines.

So one on hand, the origin of today's leveling (which probably has a history beyond AD&D, and does make some sense within lived experience) is terribly flawed. Yet it could have been dealt with differently (how many people actually played characters forever and would have actually run into level caps?).

And today's leveling has another issue, which is that every level is the same. At higher levels you need more XP, but the things you do give you more XP. It's not exponential, it's just linear.

This is part of a larger discussion in game design about where the skill lies and how to get more skill. Is it like Halo, entirely within the player? Does it mostly lie in the character, like WoW? Is increasing skill dependent on time (levels), or perhaps spending money (pay-to-play)?

I think it could be done differently. They never level up in the other fantasy game canonical source, Tolkien's imaginary worlds, although Gandalf is old and powerful with an intentional connection between age as experience and power as level. Skills could erode over time partially, or parts of them could and other parts could be maintained, or re-learning forgotten skills could come with a discount compared to learning them the first time.

Blogging, Ephemeral

So, I've been blogging for well over ten years now (but mostly the "readers" are myself and, I would guess given the consistent 30-40 views per post, webcrawlers), but I was blogging at UM before the earliest posts listed here (2007), which are re-posted UM posts IIRC.

Which made me think about how ephemeral this all is, although I might have a folder with the UMich blog material in it.

Relatedly, I've been thinking about digital photos, and what is Facebook, Instagram, and Tumblr, for instance, it's all ephemeral. Someone a while ago pointed out how, in terms of records and photographs, we are in a terrible period, as none of it will be saved for posterity. I'm debating taking an afternoon and ordering up some print books of my photos, but am not even sure what I have anymore. I think I've lost my Belgrade trip photos, back in the days of my Canon Elf (~2003?), although I might have a CD with the photo from a friend who was on the trip as well (but neither my desktop nor laptop have a CD drive, however I have an old Macbook with a CD drive that I have saved just in case).

So, no more finding ancestors' diaries and photos.

My great-grandfather and great-grandmother made a scrapbook of their trip to England in the summer of 1914, just before WWI started. My great-grandmother's parents were German and had emigrated to the US. Oops. Being German-ish in England during WWI, not a great idea. They managed to get some money somehow and got on a boat home, but it wasn't easy. Sure, if this had been 100 years later, a blog, but would it even exist for a great-grandchild to read 100 years after the fact? Our digital world hasn't been around long enough to tell, but, given the rate of digital decay we've seen, the answer is no.

Minitel!

A working Minitel! This one was built in 1985 and is still going strong, cared for by Julien Mailland and Kevin Driscoll, who spoke recently at MIT's Comparative Media Studies program weekly seminar. They have an awesome new book out about the Minitel that hit on many of the issues I also ran into studying related technologies in my dissertation back in ~2003.

But, as they told me, all the specs for the Minitel were released when it came out in the 1980s so all the service providers could connect to it. Those specs are still available today, so they have a working Minitel and an Arduino device sending it some data. Wow! Super cool.

Minitel, front

Minitel, back, and Arduino device

Closeup, Arduino device