IBM published a video where it predicts what the world will look like in 2016 (see bottom of this post for the link). It includes the following five predictions:
- You will make your own energy: Anything that moves has the potential to create energy. Your running shoes, your bicycle and even the water flowing through your pipes can create energy.
- You will not need a password: Your biological makeup is the key to your individual identity, and soon, it will become the key to safeguarding it.
- Mind reading is no longer science fiction: Scientists are researching how to link your brain to your devices, such as a computer or a smartphone, so you just need to think about calling someone and it happens.
- The digital divide will cease to exist: In five years, the gap between information haves and have-nots will narrow considerably due to advances in mobile technology.
- Junk mail will become priority mail: Think about how often we're flooded with advertisements we consider to be irrelevant or unwanted - it doesn't have to be that way anymore.
I want to start with the last one – that junk mail (i.e., spam) will disappear. You’ll need to watch the video to get the nuances of the prediction, but IBM says that in five years, Junk Mail will become a thing of the past. Instead, what will happen is that spam filters will become so good at knowing what type of mail you want to receive that it will filter out everything that it knows you don’t want and deliver you the mail it knows that you will want. In this way, junk mail becomes priority mail.
Imagine that your phone syncs up somewhere and sees that your favorite band is coming to town. Your personalized spam filter would know that you like this band and either (a) allow emails like this to pass through your spam filter to your inbox or (b) actively go out and find the information, delivering it to you.
How likely is this to occur?
Bold predictions about spam filters have occurred before. As everyone likes to point out, Bill Gates predicted in 2004 that spam would become a thing of the past. Yet here we are, 8 years later, and spam is still a problem.
But it’s not the same problem that it was before, now is it?
Let’s take a look at this. The spam problem - in email - has changed over the past few years. We used to see a lot of botnet spam with illegal content but we see much less botnet spam these days. If you read any report about the state of spam, you’ll know that it has declined considerably over the past year. However, what has replaced it (in terms of how annoying it is and how many complaints it generates) is snowshoe spam which is smaller and lighter and looks a lot like marketing mail. My prediction is that the next big revolution in antispam technology is figuring out a way to effectively deal with snowshoers (now that we’ve gotten pretty good at stopping botnet spam).
Snowshoe spam is annoying. But, if spam filters do get good at stopping snowshoe spam, in addition to remaining good at stopping botnet spam (or botnet spam stays down), then IBM’s prediction becomes possible. Just think about it for a moment:
- Spam filters are good at blocking most spam so few people get it.
- Spam filters are good at detecting legitimate marketing mail.
- Social networks and search engines are becoming more and more personalized. When you login to Facebook, the ads are targeted to you. If a spam filter talked to a social network, then it would be able to automatically decide which marketing mail to get to your inbox based upon a best guess of the things you are interested in. For example, my wife and I regularly attend lecture series put on by National Geographic. If I “liked” National Geographic on Facebook, then if they ever sent mail to me, my spam filter (after talking to Facebook) would let the mail through to me. And I’d say “Hey, this upcoming talk looks pretty interesting!”
- The principles we have learned over the past 10 years still apply. A spam filter would guess what the person would like to see, but the senders of the mail still need reputation to ensure their delivery. They’d need to sign their mail with DKIM, publish SPF records and have low levels of spam complaints, ensure opt-in best practices, and so forth.
Thus, the next big trend in spam filtering, according to IBM, is theoretically possible. Is it possible to do within 5 years?
Spam still hasn’t been totally solved. Just in the past week we’ve seen an eTrade spam blitz and then a Bank of America spam blitz, and these were cases of botnet spam with relays behind relays. We haven’t managed to eradicate that type of spam yet, but it isn’t the problem it once was.
But looking over how something like this might be accomplished, it’s not tough to visualize. Imagine someone (let’s say me) had a Gmail account and used their Google+ account actively. If they +1’ed things like Minyanville and TheStreet.com, and then went into Gmail and said “Bring me stuff that’s relevant,” it’s not difficult for Gmail to sift through their mountains of mail and bring you relevant things.
But on the other hand, there’s the problem of permission. Would you want Gmail to give you updates from financial services (for example) that you never subscribed to? For me personally, if I owned stock in Apple, I might want news alerts brought to me even if I never wanted to hear everything from The Motley Fool. But perhaps I’d want to hear everything from Minyanville when they talk about Apple. After all, I +1’ed Minyanville so I must like it. But Minyanville only sends mail to people who signed up. But I want Gmail to bring me stuff that is relevant. What do I do?
I’m sure people will figure it out eventually. It’s an area that is ripe for exploration.