2017 Tech Predictions

It’s that time of the year again, when I make predictions for the upcoming year. As has become my tradition now for nigh-on a decade, I will first go back over last years' predictions, to see how well I called it (and keep me honest), then wax prophetic on what I think the new year has to offer us.

As per previous years, I’m giving myself either a +1 or a -1 based on a purely subjective and highly-biased evaluational criteria as to whether it actually happened (or in some cases at least started to happen before 31 Dec 2016 ended).

Bear with me for a moment, though. This is just too good.

In 2015…

I said:

Microsoft acquires Xamarin.

Oh, baby. Off by a year. I’m should go back and give myself a +1 for this one. It was really surprising that they hadn’t. As a matter of fact, if Microsoft had listened to me and done it in 2015, they’d probably have saved themselves a TON of money compared to what they actually paid for Xamarin in 2016. But they made the acquisition, Xamarin is now part of the Microsoft family, and (finally!) .NET developers have access to the Xamarin toolchain and can build native iOS and Android apps without having to shell out additional cash to do so. ‘Bout time, Microsoft. (I suspect this had everything to do with Satya, to be honest.)

OK, gloat over.

In 2016…

I said:

  • Microsoft will continue to roll out features on Azure, and start closing the gap between it and AWS. Calling this one a +1; it doesn’t take much research to see this has definitely been happening in 2016. However, it’s not necessarily a true statement that they’ve been closing the gap; Amazon keeps adding stuff as well, and the feature-parity lists are starting to get ridiculous. Whether these features are actually of use, however, is an important distinction, and something for the second half of this post.
  • (X)-as-a-Service providers will continue to proliferate. Oh my, yes, Ted gets another +1 for this. When running a gaming convention has an (X)-as-a-Service for it (seriously, here) then you know the proliferation is in full swing. PaaS providers are exploding everywhere, and while a few have disappeared (farewell, Parse!), it’s clear that this was the gold rush of 2016.
  • Apple will put out two, maybe three new products, and they’ll all be “meh” at best. I should’ve broken this into two predictions: one about Apple’s “meh” products, and one about wearables. If I’d done that, I’d have scored two +1’s for it, because not only have wearables not really gone very far (show me somebody wearing a smart watch, and I’ll show you a geek with too much time on their hands and not enough “discrimination” in their discriminatory income), but Apple’s product releases have been… “meh”! I’m looking at you, iPhone 7, and I’m really looking at you, MacBook Pro. (When Consumer Reports doesn’t give the MBP its top rating, you know the luster has failed.) More on Apple in the second half.
  • iOS 10 will be called iOSX. Dangit. Such an opportunity wasted. -1
  • Android N will be code-named “Nougat”. Why, hello there, Android 7.0 Nougat. So pleased to make your acquaintance. +1
  • Java9 will ship. As I noted last year, @olivergierke pointed out that Java9 had already slipped to 2017, so this one was already a -1. Sigh. And I called it a “no duh” event, too—I’m going to let this one cancel out the extra +1 I’d have given myself for the Apple/wearables thing, just to keep the math safe (and my ego relatively sized). The article he cited says that Oracle “blamed the delay on complexities in developing modularization”, a la Project Jigsaw.
  • Facebook will start looking for other things to do. Welllllllll, it’d be really tempting to say that Facebook’s now “things to do” was “Be the deciding factor in who gets elected by passively encouraging the widespread dissemination of fake news and outright falsehoods!”, but seriously, who would’ve believed that even if I had predicted it? Which I didn’t. -1
  • Google will continue to quietly just sort of lay there. A year ago, I wrote, “Google, for all that they are on the top of everybody’s minds since that’s the search engine most of us use, hasn’t really done much by way of software product invention recently. … I suspect the same will be true of 2016–they will continue to do lots of innovative things, but it’ll all be “big” and “visionary” stuff, like the Google Car, that won’t have immediate impact or be something we can use in 2016 (or 2017).” And…. yeah. +1 More emphasis around the existing products they’ve built, but as a company, they’ve clearly spent most of 2016 on the Alphabet/Google restructure (which accomplished… what, exactly?), and anything new has been either way quiet or way removed from the business.
  • Oracle will quietly continue to work on Java. A year ago, I wrote, “[Oracle is not] going to kill it, but there’s really not a whole lot of need to go around preaching its message, either. So they let the evangelists go, and they’ll just keep on keepin’ on.” Score a +1 for the long-haired geek in Seattle; they just keep posting new code.
  • C# 7 will be a confused morass. If we permit me the freedom to call it “.NET Core” instead of just “C# 7”, then wow do I get a +1 on this one. Even if I just constrain my prediction to C# 7/Roslyn, I still score one, but once you throw in the CoreCLR and “dotnetcore” and the different profiles and…. Holy spaghetti web browser history, Batman! The demarcation lines of the different project teams working on this whole thing are starting to become really clear as the different OSS projects each look really consistent within themselves, but then, when you get to the borders, things just…. fall apart.
  • Another version of Visual Basic will ship, and nobody will really notice. Alas, there was no new version of Visual Basic, since it would be in lockstep with the release of C# 7 (which didn’t ship), but nobody really noticed. Or cared. Still, nothing shipped, so -1.
  • Apple will now learn the “joys” of growing a language in the public as well. First there was Swift 2, which was itself source-incompatible with Swift 1, and then during the summer, Apple shipped Swift 3, which was… source-incompatible with Swift 2, owing to some language changes that the community effectively decided was necessary. +1. (And thanks for that, by the way—made teaching iOS this Fall a royal PITA.)
  • Ted will continue to layer in a few features into the blog engine. You’ve got comments! And I’ll take that +1, thank you very much.

Nine up (ten, if we count my Xamarin prediction from 2015), four down. Not bad. But now, we move on to the more interesting part of the post: 2017.

2017 Predictions

The calendar year 2017 is going to be a wild one for the tech industry, largely owing to the rather large orange elephant in the room—Donald Trump’s election to President of the United States is a huge wildcard whose randomness simply cannot be understated. The man thrives on being unpredictable, and like most industries, the tech industry (for all that it cherishes “innovation” and “disruption”) thrives on predictability. His collection of “tech titans” at Trump Tower last month yielded absolutely zero positive traction that I can see, and I suspect that the various corporate tech leaders (Nadella, Bezos, Cook, etc) are all looking at him right now the way humans do a rogue elephant—he could be good for them, so long as he doesn’t go wild and start tramping everything in his path out of spite, anger, fear, or any other of a half- dozen emotions. There’s no prediction here, though—just a “wow, this is an X-factor” that in turn makes predictions that much harder.

But on that note….

  • The Congress will call for an investigation into the ‘hacking’ of the 2016 US election. (0.8 probability) To be honest, I’m not sure if anybody knows exactly what we mean when we say “the Russians ‘hacked’ the US election” in casual conversation. There’s no clear evidence that the voter machines themselves were cracked or tampered with, but it’s fairly easy to see a correlation with the DNC hacks and Wikileaks disclosures and Trump’s corresponding favorability gains in the polls. That said, though, the five-hundred or so US politicians that make up the Congress (and excluding Trump himself and his transition team) are not comfortable with the idea that somebody outside the US engaged in some kind of manipulation of the election, and they are going to want answers. Just yesterday or the day before, though, Trump made the comment that hacking is “extremely hard to prove”, and he’s right about that—without some kind of “smoking gun” found in a Russian government employee’s possession, it’s going to remain a major point of contention in the coming year, and investigation or not, it’s not going to go away regardless of what the investigation finds.
  • Security becomes a HUGE deal for the industry. (0.8) The election is just the tip of the iceberg; consumers may have gotten used to (and complacent about) corporate security disclosures, but the idea that the election could be hacked is sending shivers down the collective spines of anyone who does anything online. The downside is that it’s such a complex topic, it’s hard for anyone who’s not a computer security expert to really understand what to do; even among experts, there’s a fair amount of disagreement, even on simple issues like scope (how widespread is it) or actual facts vs hype. Pair that with the paranoia that is inherent in any security professional (if you think computer security types are paranoid, try talking to physical security professionals for a while), and you have an industry that’s ripe for a lot of snake oil and hyperbole. My prediction, then, is that the industry starts to see the first set of “security snake oil” products somewhere within the calendar year 2017. And by that, I mean products that claim to provide security for your interactions online but in fact do nothing of the sort. (Late-night infomercials about downloadable web pages that can “clean your system” of viruses and malware, move over—it’s time for late-night infomercials about downloadable web page that can “secure you against even the most determined attacker”!)
  • Apple continues to plummet. (0.7) Their products this year were merely slightly-enhanced copies of the previous line of products (iPhone 7 vs iPhone 6) or containing gimmicky “enhancements” while the core of the product remained essentially unchanged from prior generations (MacBook Pro). Sorry, folks, the TouchBar does not qualify as “disruption” or “innovation”; it’s a strip of touch-sensitive glass from an iPad designed to start prepping you for the idea that Apple can remove the keyboard entirely, replace it with a touchpad, and then put a hinge in between two iPad Pros and call it a “MacBookPad Pro” and charge you $10k for it. (And by the way, if you’re thinking about one of the new MacBook Pro machines, make sure you go into an Apple Store and try it out–the keyboard is definitely not the same as its been for years. It feels like they took about half of the keys' “press depth” away, and it totally changes the “touch” on the keyboard. I imagine somebody could get used to it in time, but… ugh.)
  • Apple doesn’t introduce any new products this year. (0.6) And by new, I mean something that’s not an incremental improvement on what they’ve already got. Heck, I’ll even go so far as to say that this means that there’s no new form factors to the existing product line. (Meaning, no new-sized iPad or iPhone or laptop.)
  • PC manufacturers double their efforts to build a MacBook Pro. (0.8) The MBP is vulnerable, for the first time in a half-decade, and PC manufacturers are going to look for ways to capitalize on that. Somebody is going to put out a similarly-sized, similarly-weighted non-touch-screen Windows 10 laptop with 32GB of RAM and a 1 or 2 TB SSD, the usual collection of ports, and price it around the same as MacBook Pro ($2k to $4k), and developers will start buying them. (Bonus points to that manufacturer if they offer Linux as an out-of-the-box option.) I know I will….
  • Apple rumors about Tim Cook’s departure begin. (0.6) Cook has proven that he’s no Steve Jobs; in fact, the comparisons between his and Steve Ballmer’s reign at Microsoft are proving eerily and entirely similar. Both basically took companies that were defining the marketplace and shepherded them into a position of trying to manage the cost structures and find better price-points, and in doing so, killed off much of the mojo that drove both firms. Ballmer took close to a decade to be run out of Microsoft (and even then, it took BillG’s intervention behind the scenes, from what I can tell), but I don’t think the Apple Board is going to wait that long—I think by the end of 2017 we’re going to start hearing serious rumors about Cook being offered a golden parachute to give up the center chair and let somebody else in to run the show.
  • Oracle will continue to just write Java. (0.7) Oracle, despite the best efforts of media and journalists everywhere, just refuses to get drawn into “techno-drama”. Java hasn’t been the Trojan Horse into corporate pocketbooks that all the Java-doomsdayers were predicting back when Oracle acquired Sun, and releases of Java just keep coming through both commercial and OSS channels. There’s really no reason at this point to doubt that Oracle is going to do anything but continue down that path. Make no mistake, I’m sure they’re looking for ways to monetize Java in some way so that they can try to earn back the cash they spent to buy Sun, but I don’t think it’s going to be through selling or charging for the JDK or JRE anytime soon.
  • Oracle Cloud emerges onto the cloud scene in a big splash. (0.6) IBM now has Bluemix and Watson, and they were really the last of the “big-iron” holdouts around the cloud. (What I mean by that is that all companies have been quietly flirting with cloud, but some push it loud and clear, a la Microsoft or Google, and some were playing it very quietly for a while.) With IBM acquiring Loopback (a NodeJS server-side stack) last year, it’s clear that IBM is going to push JavaScript as their main cloud development play, essentially ceding the Java-cloud development ground to somebody else. Amazon has historically been the place that Java developers have gone to run their Java code in the cloud, but if Oracle can build a compelling offering (particularly with a free tier that AWS currently lacks), this could be a relatively big splash. Between Oracle’s reputation in the database world, if they have a solid “stack” offering that basically makes a Java-based back-end a snap to start up, Oracle could essentially claim the Java-favored cloud play from Amazon. (Yes, Heroku is out there and holds a fair amount of Java and Scala love, but now that they’re owned by Salesforce I suspect the Java-leaning flavor of Heroku to wane a bit.)
  • Salesforce makes a major database acquisition. (0.5) Salesforce is growing, and they’re clearly interested in expanding their cloud to be more than just the CRM. With Heroku, they have a Platform that developers can feel comfortable on, but they don’t have a big-name database (relational or otherwise) that complements that play. They currently are sitting on a ton of cash, and last year’s crop of acquisitions didn’t include a big database storage name. There’s not a ton of players left out there, but I could see them making a strong push to get something like Cassandra or Couchbase. (Yes, they have Data.com, but that doesn’t seem to be making much headway in the developer mindset space.)
  • Salesforce releases a new programming language. (0.4) Let’s call the spade a spade: Apex is a Java knock-off, and it shows a lot of warts, particularly since it hasn’t really kept up with what few improvements Java-the-language has made in recent years. The last company to be in this position—a red-hot platform but a language feeling a little creaky at the corners and just plain “old” everyhwere else—was Apple right before they released Swift. Salesforce has the engineering power, they are looking to command more of the developer mindshare, and they have a ton of cash to blow, so…. Whether this happens this year, next year, or 2019, I’m not sure, but if it doesn’t happen this year, the odds go up each year after that.
  • LinkedIn Learning starts to make a serious dent in online developer training. (0.5) Between the fact that LinkedIn Learning (formerly Lynda.com) is growing out its library to a pretty respectable degree, and the fact that Microsoft now owns LinkedIn, it’s pretty reasonable to assume that Microsoft is going to start making this available to its developer community in various ways. This may happen in 2018, though, depending on how swiftly Microsoft moves to incporate LinkedIn assets across the rest of the firm; if they bought LinkedIn solely for the CRM data to go with Dynamics, for example, then this probably won’t happen for a few years.
  • Swift doesn’t go to 4. (0.7) Swift 3 held breaking changes from Swift 2, and the folks at Apple are not stupid. Swift 4 will be far, far down the horizon for a few years yet, given that each major version number bump has heralded incompatibilities. Apple will not want to call anything “Swift 4” and dredge up memories of incompatibilities in their customers' minds for a while. Swift might get a 3.1 in the summer, but that’s as far as it’ll go.
  • Microsoft ships C# 7. (0.8) Roslyn needs to ship in 2017 if Microsoft is going to be able to call this open-source process a success. Otherwise it’ll start a lot of people grumbling. (Yes, a new version of Visual Basic will come with it, and it will make basically no news.)
  • No new Android version. (0.4) Android-N is still slowly making its way through the networks, and while we’ll probably start hearing rumors of what Android-8 (Oreo?) will include, with a targeted ship date of 2018, probably 1Q or 2Q.
  • Twitter will continue its slide into irrelevancy. (0.5) Let’s face it, Twitter’s days are numbered. If you’re holding Twitter stock, now’s a good time to sell—when Twitter was left out of Trump’s “tech summit” last month, the stated reason was that it was “too small”. Put that into your brain-pan and circulate for a while—the service that invented microblogging and is one of the core founders of “social media” was “too small” for the PEOTUS' time. Twitter hasn’t really done anything “new” or “interesting”, but simply continued to be the 140-character microblogging platform it’s always been. It’s reaching commodity status, in fact. That’s not a good sign for a company that wants to be more than it is. I suspect Jack Dorsey gets tossed on his can, the company starts looking for a new CEO, and the “new vision” will start to take shape by the end of the year (2017), and then in 2018 we find out that the “new vision” is terrible, takes them out of their “core business”, and the slide accelerates. But nobody buys them this year, not yet.
  • The “Internet of Things” continues to draw hype, and continues to fail to deliver. (0.6) It’s been how many years we’ve heard about IoT now, and how it will revolutionize our lives, and all we’ve really seen thus far is the wide variety of Internet-enabled devices being subverted for a widespread DDoS attack. Wearables, “smart refrigerators” and other IP-enabled devices are proliferating, but—to perhaps everybody’s surprise but mine—nobody’s quite sure what to DO with these things once you have them. Your thermostat is online; terrific. Does it have an API that will let me query meter usage? No, that’s a different thing, and a different API, and a different connection endpoint, and…. Oh, and be careful, somebody could remote-hack your thermostat and hold your house hostage. Because that’s worth the risk.
  • Tech “unicorns” will start to watch the bubble pop. (0.3) Uber, Lyft, all these companies that are valued at double-digit billions with zero profits, major losses, and no real assets to sell in the event of a bankruptcy…. All of this is going to start to make some investors nervous, particularly when they look around and realize that the tech sector has been carrying the country’s economy through its “recovery” (yes, we’ve been in a recovery for the last half-decade!). All it takes is a few small stones to start the avalanche.
  • Voice-controlled fart apps will emerge. (0.6) Seriously. As Alexa and Siri and these other voice-activated systems start to move into stationary devices in your home, and as the SDKs for these systems start to become more widespread, the first thing developers will do is build some kind of ridiculously silly app (it would be a kindness to call it a game) that will somehow sweep everybody’s sense of humor into the toilet. (Seriously. Imagine it. “Alexa, did you have beans for dinner?” “Yes, I did, and– BRAAAAAAAAAAP!” It’s exactly the kind of thing that would get people giggling for hours on end, particularly in a weed-induced state. Did I mention I live in Seattle?)
  • Facebook will find that preventing ‘fake-news sites’ is a lot easier said than done. (0.8) As a result, they’ll put some kind of “AI” filter on linked sites, declare a victory, and try to get out of the political game entirely. It’s a lose-lose scenario for them: one man’s “fake news” site is another man’s “revolutionary take” backed by the First Amendment, and Facebook does not want to be anywhere near a court trying to justify their actions against Free Speech. (Old-timers like me will remember Prodigy, an online service that started censoring content, which started its slide into doom.) Zuckerberg doesn’t want to be held responsible for swaying important political events one way or another, but neither does he want to be the target of numerous political activist lawsuits (from all directions). As Joshua (the AI in the WOPR, back in the 80s movies that every geek my age openly worshipped) learned, Zuck will discover that sometimes “the only winning move is not to play”.
  • A driverless car will kill somebody. (0.5) It’s only a matter of time. The circumstances may not be the software’s fault—and in fact it’s likely that it won’t be, when the final analysis comes back—but the headlines will scream, and the widespread fear of a human “not being in the loop” will set driverless cars back by years. Expert testimony and repeated demonstrations will do nothing to shake the public’s fear that a computer-driven car could “hit a bug and kill me”.
  • The topic of ethics and programming will begin to become fashionable. (0.3) Somewhere alongside the driverless car’s first fatality, people will start asking how the car’s programming makes decisions that most humans make in a split-second without even thinking about it. Case in point: the car detects that a motorcycle rider has had a problem and the rider has laid the bike down in the road right in front of the car. (For discussion purposes, there is no room left to brake; the rider is too close.) The car can either swerve to the side to avoid the now-helpless rider, potentially causing a major accident involving multiple people; or the car can simply continue forward, running over (and very likely killing) the motorcycle rider but avoiding the possibility of multiple fatalities from a larger accident. Most humans would swerve—but is that the “right” decision? More to the point, what should the software be programmed to do? Once the public gets wind of these kinds of decisions being made by geeks behind flat-screen LCDs, it’s going to cause a major outcry. (And yes, these kinds of decisions are going to be encoded in the software, somewhere.)
  • “The cloud” continues to grow, even as consumers wonder what the hell it is. (0.7) Let’s be clear—as of right now, the cloud is basically a developer thing. My parents really don’t “get” the cloud, largely because there’s really nothing they get from it. Sure, one can argue that GMail is the world’s most popular cloud email service…. but your email is just stored on a server that Google owns, as opposed to a server that your ISP owns. (If that’s your definition of “cloud”, then pretty much all client-server computing is “cloud” in your world.) People are looking at more online services for things like bill payment, true, but those are basically services being offered by vendors with whom these people are already doing business–again, that’s not “cloud”. Cloud offerings have basically found a home in the developer world, but general-purpose cloud, the way that cloud was first being sold, is losing its window of opportunity to get hold of general consumers' minds. (I lose this prediction if my parents are suddenly smitten with a product that stores or computes for them and isn’t a vendor they already have a relationship with.)
  • “Blockchain” remains the most opaque ‘thing’ of the year. (0.8) Everybody will go on and on about its huge technical advantages and obvious benefits, while never actually describing what it is or how it could work to change the world it’s so clearly destined to change. It’s the ultimate hype machine, and it will show no signs of slowing down until maybe the end of the year. By that time, something will emerge out of it (the way blockchain emerged out of bitcoins and cryptocurrency) that will carry forward the legacy of “changing the world” without actually changing anything.
  • Artificial intelligence will continue to remain a ‘future’ thing. (0.8) Part of the reason I say this is because AI is like magic—if you can understand it, it’s not interesting anymore and it’s just an implementation detail. We’ve had rules engines and natural language processing for years. When Amazon started doing “predictive analysis” of what you would like to buy, we pulled “data science” and “behavioral analytics” out of the “AI” world and into its own category. When AI figured out how to make the spoken word make sense, we called it “speech-to-text” and it was a feature on Android alreday back in the v2 days. (Marry speech-to-text up with a natural language parser, and you have Siri—which, remember, was its own company before Apple acquired them.) No, Alexa is not going to revolutionize the world any more than Siri did—the act of talking to a machine is not particularly new, and it’s only as good as the services that sit behind the parser and can “hook in” to the parsed text. “Cortana, fire up StarCraft 2” is easy to parse and start an application; “Cortana, fire up StarCraft 2, and find me a random Hard co-op match as Artanis” requires not just firing up an application, but also “hooking” inside the application to know how to carry out the rest of the request. That requires an API platform that all applications can hook into, provide, and describe (in natural-text terms) to the voice-control system. That is not going to be easy to define, adopt, or test.

On a personal note, several predictions come to mind:

  • Ted will celebrate his one-year anniversary at Smartsheet in September. I’m optimistic about these guys, and the things we can do together. I’m looking forward to taking them into the developer limelight in a variety of different ways.
  • Ted will do less speaking this year. My new role actually encourages me to help develop new talent for my employer to go out and do the actual speaking, so while I’m definitely down for doing a few conferences this year, it’s not going to be more than 12, total, for the calendar year. I enjoy speaking, but I’m looking to be a lot more careful about where I speak now.
  • Ted will not be renewed as a Microsoft MVP. Actually, this appears to be fact, not a prediction. MVP renewals for the January cycle went out already, and I didn’t receive one. Fortunately, most of the stuff I care about in the Microsoft world is all open-source (or moving that way) anyway, and while it’s been nice being on the MVP mailing lists, there’s really been nothing there that’s been all that insightful or amazing. (And, fortunately, living in Redmond makes it trivially easy to get together with anybody on a product team if I really want or need to, and I am privileged to call many of the people on those teams “friend”.) It would’ve been 14 years, but as we Stoics say, “All good things, in time, must come to an end.”
  • Ted will look to engage with other tech companies beyond Microsoft. Google just started a new MVP-like program, and I’ve been teaching Android and Angular and some Google Cloud Platform stuff for a while, so perhaps they’ll welcome me into their fold.
  • Ted will continue to teach at UW. I’ve been guest-lecturing at UW for the past three years now, and I’m loving it. The students are bright, eager, and a helluvalot smarter than I was at that age. They’re an incredible joy to teach.
  • Ted will look to publish a few mobile apps. I’ve had a few ideas floating around for a while, but just never really made the time to do it. Even if they never turn a dime in profit, I’m long overdue for having a few apps in the respective mobile stores.
  • Ted will continue to write for various tech ‘zines. I love having the back-page editorial at CODE Magazine, the column in MSDN, and the various series on developerWorks, among others. I fully intend to keep all that going at full speed. (And I’m always looking for new outlets, if anybody has any leads on paid technical content gigs!)
  • And finally, Ted will try to blog more. The perennial projection. I’ve got much to blog about, including the patterns series, as well as some interesting themes and ideas floating around the ol’ brain pan.

Happy Holidays, and thanks for reading!