16 January 2020
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 2019 ended).
... I wrote a lot of stuff. (That always seems to happen when I do these.) Let's start. I'll include the full
text of what I wrote in each bullet point first, then put the Result after it with whatever insights or
comments seem relevant. (Arguably none of them are, but hey, it's my set of predictions, so....)
Blockchain might actually start being understood. (Probability 0.4) Blockchain itself is not going away—if anything, companies seem to be putting more resources into it, not less—which suggests that enough people are seeing something of value buried deep (very deep) inside the hype. It may not be clear yet what it is, but that message is going to start trickling and filtering out and a few people might actually start to clue in on how best to use it. Result: 0 Honestly, I'm not sure how to score this one. On the one hand, we are starting to see some companies coming out with tooling and platforms that are centered around blockchains, which lends itself by definition to allowing the technology to be better understood. On the other hand, though, can anybody really, actually, describe what the hell it is? Rather than make the strong case for either point, I'm just going to call this a zero and move on.
AI and ML will begin to start separating... a little. (0.6) There’s lots of things that fall under the umbrella term “artificial intelligence”, and machine learning is only one of them. As we get more experience with ML in practical usage, we’ll start to see that the different branches of AI are useful for different things, and we’ll begin to use those terms a little more carefully. Look for “natural language” and “expert systems” and “machine learning” to start getting used independently, rather than the catch-all “AI”. Result: -1 I just can't give myself a point here. There's a few places that are starting to discern between some of the terms, but for the most part, "AI" and "ML" are still entirely synonymous, and that is just confusing the hell out of everybody.
AI will start to move out of the startup and into the enterprise. (0.7) Lots of people have kicked their AI startup off, and lots of big companies have started AI initiatives within their own companies, and they’re all finding… this stuff is hard. And what’s worse, the kind of skills required to do “AI” (whichever flavor you care to name) are rare, and those departments simply will resist scaling to the rest of the company. What usually happens after that first wave of interest, however, is that a small group of companies start to figure out how to apply AI to more mundane and practical purposes inside the big companies’ IT space, and that’s where the serious money lives. Sure, there’s going to be some flashy bits about robots and image recognition and stuff, but the real money is waiting to be made in machine learning being used to discover patterns in the data that up until this point mostly got discovered by accident. Look for a few companies to start selling AI not as a raw technology, but as part of a toolchain that a company can buy to do “something else”. Result: +1 Oh, yeah, it's happening. It's hard to point to examples, but just about everybody I know in the space is talking about how they're using AI inside of a large enterprise, in many cases by virtue of the large enterprise having bought into one of the fourteen-million different AI startups. Bunch of it will probably fail, miserably, but that's really never stopped anything on the enterprise hype curve, and frankly, is all part of the cycle.
Amazon gets hauled in front of the DOJ. (0.8) If it doesn’t happen this year, it’ll happen in 2020. Jeff Bezos has been doing a great job of staying out of the limelight when it comes to the ugly whispers of “monopoly”, but good Lord, folks, Amazon is clearly making no secret of the fact that it wants to take over the entire retail experience—and when you couple that with the fact that Amazon is now opening brick-and-mortar stores, there’s not a huge leap from “Company using its monopoly power to drive competitors out of business by taking losses on sales” to “Amazon founder Jeff Bezos was in Washington DC today to sit in front of a Congressional inquiry…“. Particularly when Amazon just bought Whole Foods, a non-technology company if ever there was one. Result: -1 Oh, but not for lack of trying---Trumpie has been doing everything short of putting Bezos in front of a Congressional impeachment trial to try and shoot the guy down, and it somewhat surprises me that putting them in front of a monoply accusation hasn't occurred to any of his followers yet. Either way, though, it didn't happen, so no point for me.
Facebook undergoes some kind of radical transformation. (0.8) Zuckerberg is on the ropes, whether he realizes it or not. People are quitting the app in droves. Milennials aren’t taking it up. Every single day, various political groups scream about Facebook’s censorship of their particular pet issue, and the shine has come off of Sheryl Sandberg’s star. Facebook’s various other branded platforms, notably Instagram, are also starting to come under fire as people begin to openly question exactly how much influence these “social media influencers” have on them—or should. Congress may or may not get involved here, but either way, what was once pretty much a lock on a market has shown some serious cracks, and I suspect several competitors will start sharpening their knives, including several startups that may or may not already exist. Zuck will need to show some determined leadership and strike off in a bold direction, or the calls will come for his head faster than they did for any of IBM’s CEOs in the 90s. Result: -1 No, Facebook seems bound and determined to just keep on keepin' on, as they say, and it's likely going to hurt them in a big way. Zuck seems completely bound to the idea that if they just make a few tweaks here and there, they can just keep on being who they are, when in fact, who they are is by very definition a vehicle for bad mojo. Look, in a lot of ways the social media world is following the same arc as early journalism (check out "yellow journalism" and see if you recognize anything), before some of the core laws and rules and conventions governing it came into being. Although, honestly, given the current political climate, we may be losing those rules and laws and conventions before too long, too.
Google will look to capitalize on Facebook’s struggles... and fail miserably. (0.8; 0.6) This is another two-fer. Right now, deep in the heart of Menlo Park, a Google executive is putting the finishing touches on a presentation to the Alphabet board around building yet another social media network app, complete with numbers about how many GMail users want to see a union of their email and social media profiles, and how that could drive additional AdWords revenue, and…. And the people to whom that presentation is shown will nod sagely, approve the effort, and…. Google will announce the beta of yet another Google social media app that will somehow be different than all the other social media Google-branded apps, because this time, it will be different… somehow…. And yet, it won’t, and it won’t, and somewhere in 2023 Google will kill it because this time, it’s replacement will be different…. Result: -1 Google has, remarkably, stayed just as clear of the social media space as it ever has, which is to say they didn't throw a new social media product into the fray, and frankly, that surprises the hell out of me.
Microsoft will, too. (0.6) Unfortunately, here in Seattle, a Microsoft exec is putting the finishing touches on almost exactly the same presentation. With any luck, Satya will toss the whole idea out the window before the presentation is even finished, but... Result: -1 And I've almost never been happier to be wrong. Satya, you did good on this one. (Of course, Microsoft probably put all that effort into Teams, so maybe this one is a wash?)
Uber is going to struggle even more. (0.8) Uber was once the darling of the Silicon Valley and Wall Street—now, it’s having a hard time getting the time of day from either group. This past year was not kind to them (a net loss of close to US$3 billion, if we leave out one-time investments), and people have figured out that Lyft and other Uber-competitors are just as convenient, but without the soul-selling that using Uber brings to mind. Uber Eats might be a saving grace here, but again, lots of competitors, and those competitors don’t have the “Uber” logo that is increasingly turning from balloon to boat anchor around their branding. Result: +1 Oh, my, yes. They're not alone, to be sure, in that Lyft and several other ride-sharing firms are expressing equal kinds of difficulties. And the whole wider collection of "unicorn companies" that were basing themselves off the idea that people would want to partake in the entirely ungoverned, uncontrolled, unmonitored, and employee-unfriendly "gig economy" (or "side hustle" if you prefer) are facing some problems. Most of which stem from the fact that these companies aren't "disrupting" anything, they're just trying to reinvent an already-existing thing by calling themselves something different and trying to eke out a profit on the narrow margins that they can find by not having to pay government oversight costs. That was, is, and never will be, sustainable.
AWS (and Azure) will only continue to add services. (0.9) There’s no end in sight! AWS already has ten database services, but is there another angle to data storage that needs to be in the cloud? Go create it! Brand it! Sell it! Turn it into a thing that Bezos can show off at re:Invent 2019! There’s no end in sight! Result: +1 Of course they did.
Calls of “enough already!” will begin to echo in the clouds. (0.9) It’s becoming far more than anyone can keep up with. AWS, Azure, Google, all of them have way too many services for anybody to reasonably keep track of, and as a result, we will start to see a new industry of “cloud advisors” begin to form—--speakers, online courses, and the like—--simply geared to guide people through the choices of which of Amazon’s dozen or so data services you might actually need to use. Result: 0 It's starting to happen, but quietly and in dark corners. Hence, I can't really cop a point for it.
C# 8.0 will ship in Visual Studio 2019. (0.9) This one’s a gimme, but I’m never one to take a free point or two. Result: +1 April 2, 2019, baby. Or late September, take your pick.
Java12 will ship. (0.9) Oracle has announced that they want to move to a six-month release cadence, so given that Java11 shipped in September of 2018, it would make sense that Java12 will follow in March of 2019. One thing that needs saying, though: it’s not clear what features would actually ship in Java12, particularly since some of the proposed features could take a year or two to ensure that they meet the high quality bar that a language and platform as widely-used as Java requires. (Note that the same thing is true of C# and .NET, but Microsoft hedges their bets some by doing major releases as part of the two-year release cadence of Visual Studio; Oracle lacks that “ship vehicle” channel, since they don’t really own a widely-used IDE—--and no, sorry, NetBeans doesn’t count there.) Result: +1 March 19, 2019, baby. Oracle has been almost nefarious in making their six-month ship cycle for Java, and whether that's a good thing or a bad thing has yet to be determined. I think the bigger question, though, is whether or not anybody actually cares about Java's ship dates anymore. I mean, I used to look at each upcoming ship date with a certain amount of excitement and attention, but now... And more to the point, there's no "Here's what's in Java 12" presentations happening at conferences much anymore, which leads me to wonder if anybody actually cares.
Android ships... Q-something. (0.8) Android definitely seems on a 1-year release cadence, but hell if I can think of a dessert that starts with “Q”. Fossbytes suggests names like “quarabiya” or “quince”, but only one of the names is actually found inside the US (“quesito”, from Puerto Rico), and all of Android’s previous code-names have been US-centric names (with one exception, the “Petit Four” release, 1.1, which was also before they started using the alphabetical order, too). OK, OK, I know, “Eclair” is technically French, but it’s widely-known and -recognized in the US, so it counts. So Android has to come up with a “Q”-based dessert that is widely-known in the US and won’t be impossible to pronounce…. or they just skip it and go to “R”. Either way, I don’t envy the person who has to make that decision. Result: -1 They gave up. THEY GAVE UP. They got to "Q", and rather than even try to shoehorn something in place, they bailed entirely and just called it "Android 10". I mean, they should've done this from the beginning, so this is wildly overdue, but damn, man, I wanted to see what they would come up with for "Q".
iOS 13 is skipped, and 14 ships. (0.4) Hey, nobody in the Western world likes the number “13”, and Apple is not one to try to convince them otherwise. Look for Apple to pull another marketing shift and try to call iOS 13 “X3” or something, just to avoid the superstitious’ wrath. Result: -1 2019 was not a good year for Ted trying to predict marketing moves. Not only did Android move to numbering, but iOS decided to stick with numbering and just ship iOS 13. sigh
Security is going to quietly become a must-have skill. (0.6) The public is tired of the bleeding. Companies are tired of the scrutiny and managing PR fallout. Developers are tired of the second- guessing and accusations. Everybody is tired of security holes biting them in the ass. Slowly, quietly, companies are going to start demanding their developers be held accountable somehow—either in interviews or in internal code reviews or in any other form they can actually use—to make sure code doesn’t get compromised. It’s become clear that the hackers are only getting more and more organized, including state-level backing, and that “thoughts and prayers” are not effective as a security strategy any more than they are as a gun strategy. Result: +1 Let's be clear, if your company isn't starting to take security seriously (particularly in the face of GDPR and its ilk), then you're really just begging to be hacked and owned and bankrupted. We haven't quiet yet reached the point where we have answers to the problems that plague us, though, and that's where the problem lies---too much expertise is required to secure bespoke applications, and right now the best answers are "outsource your authentication strategy" (to places like Auth0 or Okta), which doesn't take us very far.
More security legislation like GDPR will come. (0.7) For the same reason as the previous one, governments will start laying down regulations about data privacy and use, and many of those laws will be contradictory to one another, if not to itself. (Even GDPR has a few edge-case flaws inside of it.) What’s worse, because it will be new legislation, and the penalties will be stiff enough to merit actually doing something about it, lots of companies will find it impossible to understand what to do, obtain help from “experts”, and choose to take the most conservative route that will only make end-users’ lives harder—much like what happened with GDPR. (You think accepting the privacy-policy-to-use-cookies messages are bad now—just wait. This is only getting started.) Result: +1 Hello, California. Should've known you'd be the first. Chances are good you won't be the last, since now this forces anybody doing business anywhere in the US to have to bring themselves up to California's standards... or refuse to do business in the largest domestic economy in the country. Expect a few more states to foloow suit, and then it'll just be waiting for the rest of the bowling pins to come down. Except for Mississippi---that place is just too bass-ackwards to care. (Seriously, twelve years for a cellphone? That is some fucked-up shit right there, and I don't mind swearing that strongly about it.)
Developers start finding their voice on ethics. (0.7) It started with Microsoft and Google, but it won’t end there—more and more developers are finding that they have a certain amount of power at their fingertips, and that unlike a number of their non-technical colleagues, if they threaten to abandon work on a project, their employers will listen to them. (So long as the demand for developers exceeds the supply, this will continue to hold true.) While there will always be some numbers of developers who will not care what they work on—so long as the checks cash—a similar percentage of developers are growing more and more aware that what they do could be used for less-than-innocent purposes… and will start to become an active voice. Look for a few more attempts (successful or otherwise) at coordinated walkouts by developers and other IT staff over easily-identifiable egregiously-unethical uses of some technology or other. Result: +1 Oh, my, yes. It's happening. Google, Microsoft, other companies are finding that their developers have a voice and are growing more willing to use it. Of course, this also means that companies are starting to find ways to try and shut that voice down, which is setting up for some very interesting employer/labor fights to come.
Developers are going to start a backlash against coding-school graduates. (0.5) This is one I hope I’m wrong about… but I’m guessing I won’t be. Here’s the thinking: More and more, “coding” is being seen as the next-great-way-to-get-out-of-poverty by millions of people who have little to no idea what “coding” means. They are, however, putting thousands of dollars onto credit cards and loans in an effort to learn “coding”, so that they, too, can get at this untold paradise of billions of dollars that we technology-types have been enjoying for the past two decades. But, as any average economist will tell you, the equilibrium point depends on the intersection of supply and demand—and so long as the demand for software was high, and supply of those who could deliver it relatively low, then the price (the equilibrium point) would stay high. Floods of new developers threatens to increase the supply—shifting the supply curve to the right—and bring down the equilibrium (price) point. Salaries will fall. Developers’ power relative to the rest of the company will fall with it. And developers are analytical enough to know this, see through how the rest of that story ends, and look to take steps to solidify their “hold” that they’ve enjoyed for so long. (For the record, I believe this has already been happening for several years now—why else would we as an industry be so concerned with discerning between the true “craftsmen” and the casual practitioner of code?) Look for more and more tech-minded folks to start openly resisting the idea of hiring and mentoring junior developers as they start to figure out (or guess) that their position and high income will be threatened if they don’t. Result: +1 Ironically, though, the backlash is coming not from the people who would hire coding-school graduates, but other coding-school graduates and/or their allies, who are rallying against predatory practices on the part of said coding schools. Some of them (like the linked story) were clearly never going to fulfill on their promises, but other schools that are genuinely trying to be a force for good will likely get tarred with the same brush.
2020 is the start of a new decade, and that usually means predictors and prophets and tech pundits and other
folks who think they know a bunch of stuff try to convince you that this is the start of something entirely
new. (It's those kinds of folks who predicted flying cars and hoverboards, by the way.)
Before we get too deep into this, though, I saw an interesting Twitter meme that went something like this: In
the movie "Back to the Future", Marty McFly journeys into the past by 30 years, from 1985 to 1955. If Marty were
to make that journey today, he'd journey back to... 1990.
Shall we talk about what this wondrous new year is going to bring?
Everybody is going to go ga-ga over 5G, and it won't make a whit of difference. (0.6) Apparently, we have to go learn the Eight Fallacies of Distributed Computing all over again, again. On top of which, 5G is rumored to play havoc with weather prediction, so don't expect this one to go through with the ease of the 3G or 4G upgrades. I suspect that this is going to face some major backlash, probably in 2021 or 2022, particularly if it gets deployed widely in 2020.
Quantum computing will be the next major hype machine. (0.6) From the "our blockchain cannot possibly be defeated by any crypto-cracking tool known to man," we whiplash over to, "our quantum computing platform can crack any modern cryptography in seconds" without missing a beat. Quantum will be the Huge New Thing, and everybody will want to take a crack at the quintillions or whatever of CPU cycles that quantum computing represents... and like most of these kinds of hype waves, really, quantum computing will be useful in a small set of niche verticals, and useless to everybody else.
Data privacy is going to start to separate from the larger "security" set of issues. (0.5) Companies are having to take "security" seriously, but there's "serious" and then there's "if we don't do this we can't do business anymore" serious. Data privacy will fall into that latter category, given the legal penalties that will face violators of GDPR and the new California laws. THis means that tools and products (likely databases and its related ilk) will start advertising and promoting their "privacy-by-design" kinds of features, as a way of making it "easy" to implement privacy in your enterprise, big or small. But they won't want to be held liable in the event of a general hack, so they'll very carefully and clearly outline differences between "privacy" and the larger field of "security".
Kubernetes all the things... to whatever comes next. (0.4) Kubernetes has pretty much locked up the container world, which means that we will now start to see what peple are suggesting we need to do after Kubernetes, because the stack has to keep growing deeper, or else vendors have to start competing with something in the middle of the stack, which means having to get into a feature war or fight for standardization somehow, neither of which are proven routes to success.
The demand for "full-stack" developer unicorns will reach its peak. (0.7) Anybody who's watching the job sites and LinkedIn listings can see that everybody wants a developer who knows a major enteprise language (Java, C#, etc), Docker, Kubernetes, SQL and NoSQL skills, JavaScript (preferably knowing Angular, React, or Vue, and possibly all three), HTML, CSS, XML, JSON, one of the major cloud platforms (AWS, Azure, or GCP, certificates from certification authorities in each of those preferred), and sometimes even more beyond that. Folks, that's not a developer, that's an entire IT department (thank you, Twitter meme), and at a certain point that bubble is going to burst. But not this year.
New programming languages are going to start the R&D cycle on languages again. (0.6) They may be "service-oriented", they may be something else, but languages like Ballerina, Jolie, and Dark are basically suggesting that there's a new level of abstraction to be sought as a linguistic first-class citizen. Frankly, if you've been doing any sort of service development (micro- or otherwise), you've felt the costs and pain (in terms of complexity and the number of moving parts you have to track) of doing so. Cloud vendors are trying to suggest that "serverless" is the way to go, but in a lot of ways it just trades the complexity off to other ares of the app; a new paradigm is what's needed, to bring the complexities back down to human-manageable levels.
Automation will take an even stronger seat at the conceptual table. (0.7) "If I can't find developers to do these things I need to get done," fumes the CEO, "I'll go out and find some tools that can do all that stuff without requiring them!" Automation, particularly that enabled by low-code or no-code tools, will begin to skyrocket as a desirable target, even though this will mean that non-developers will start doing developer-ish things, and before long, will run afoul of the same basic problems that always strike when non-developers start doing developer-ish things. (Truthfully, this is related to the programming languages issue; companies need better "technical agility" out of their IT departments, and right now, nothing is providing that. They will look for anything, no matter how outlandish or outrageous, to provide that, and in 2020 that pressure will really make itself felt.)
Lastly, one more:
Despite my best intentions, I totally abandoned the blog in 2019. With the shutdown of MSDN Magazine, and my choice to withdraw from the back-page editorials in CODE Magazine, with no drawdown in my interest in writing and no clear outlet by which to publish it, the blog should see some better love in the year to come. But... I still won't give that a 1.0 probability.
I will, however, state that this last point---the one about businesses needing to find better technical agility in their IT and technology departments---is one that I'm very, very interested in, and I now work someplace that has the resources and interest in solving some of those problems, at least internally (but hopefully in an FOSS manner). We could always use a few more talented engineers, if you're of a mind...
Last modified 16 January 2020