In the article Here’s What Is Limiting Twitter’s Mainstream Potential, Mills Baker makes an intriguing proposal: Twitter’s chronological fire-hose format is limiting its ability to reach a mainstream audience. He says bringing in an algorithm à la Facebook’s automatic timeline curation will bring joy and contentment to all twits worldwide. I respectfully disagree; I’m not up for any similar feed-wrecking Twitter “enhancement.”
That’s the idea the author pushes, though. The story makes the analogy of a “rude editor:”
If Twitter is your dynamic real-time newspaper, it’s also a newspaper with a single editor: you. While that feels empowering, it’s also burdensome; you are now required to tailor your timeline to your taste in perpetuity. Worse, you have just one signal you can send to your “newsroom” to indicate the kinds of content you want: follow / unfollow a user. You’re an editor whose only power is to hire or fire.
The author submits that you would be obliged to appreciate a machine-learning algorithm to take the role of “assistant editor,” curating your feed to your own likes and dislikes, and never showing anything outside your preferences. Their theory is that the solution to bringing Twitter to the masses will “involve personalization through algorithmic feeds” instead of adding features to actually allow users to customize their own news.
The article hails the Facebook overhaul as a great new feed-aggregating superhero: “[G]one are the political screeds, and instead I see photos of babies, dogs, and trips taken by people I actually like!”
Perfecting the feed? Not for this Webhead. What if I actually prefer thoughtful discussions and journal entries instead of endless graphic memes and photo galleries, cute as the babies are? Can I change it? Nope, the process is all invisible and unreachable.
I grant that Twitter’s format does produce a number of irrelevant items in the news mix, and that the full feed can be overwhelming at times. Just don’t give a mechanic a car and then superglue the hood.
Definitely lots of great points; I think you’d be crazy *not* to be worried that many of the coming ML feeds will be garbage, for some time anyway. I also think that many of us will be unhappy with the future of the web, but I think more people will be pleased than displeased by a wide margin. Facebook isn’t erring with their feed; the data is clear on that. And remember: a chronological Facebook feed isn’t an option. Most people have too many friends making too much content.
But this is odd to me:
Of course you can change it! Facebook already has loads of signals at your command: they’re not just looking at scroll speed, what you stop to view, what you click on, where you comment, and what you like; they’re *also* looking at what you say to unfollow; what you say to unfriend; what you say to block; what you reject; and so on. In fact, often when I feed back into FB, they ask me to complete a little survey of *why* I didn’t like something, presumably so they can improve the feed.
What interests me here is the idea that there is currently some unedited, untampered with supply of great information that I solely command and direct, free from any institutional limit or programmatic controls or levers, but that I’m about to lose that for something in which I’m a coerced, captive user. That’s a nice political idea, but it’s not accurate at all. I neither have as much control as that implies currently nor will I have as little as it implies I will in the future.
To be frank, I think us nerds —I’m assuming you’re a nerd; if not, I apologize; I am a nerd of 80s/90s extraction— really over-index on “control” and under-index on “ease” when we think about software. That’s nothing new; you’d rather manage an RSS reader, find blogs using a URL bar and Google —neither of which ever suggests anything to you, neither of which exposes you to new and diverse info, incidentally— and manage them manually. You get total control both in action *and* conceptually: you are the chooser.
But regular people —and in a democracy, we must respect them as we do ourselves— do not care about precisely controlling their hardware, software, systems; they don’t find RSS easy enough (it isn’t). They don’t think the web is easy enough (it isn’t). They don’t want to curate their Twitter lists and manage their RSS subscriptions. And what they want is no less important than what we want.
I would prefer to have people visit my web page, which I used to hand code and upload over FTP. But *people* don’t prefer that. And I can’t grouse over the fact that technology reflects the desires of my peers; in fact, I don’t feel I can even gainsay their desires.
Anyway, sorry for the long comment, and thank you very much for reading and replying; I appreciate it!
No worries on the comment length; it’s good to hear your thoughts.
Before this auto-curation began, the Facebook firehose showed me what at least appeared to be the bulk of friends’ posts. Then the feed was revamped, and at that time my number of friends’ updates slowed to a relative trickle. Yes, I get that you can tweak the timeline with blocks, likes and so forth, but you can’t vote on a post you never see in the first place.
I prefer to see everything and then whittle that down to my preferences explicitly, instead of an algorithm trying to guess what I’ll pick. I understand that the majority may very well prefer the reverse, though. You have a valid point, and I appreciate your sharing your perspective.