1. 27

Predictions are hard:

  • The Beatles’ Decca audition
  • There’s no chance that the iPhone is going to get any significant market share

So at the risk of being totally wrong, what do you think is next in your field?

Personally I’m focused on programming and entrepreneurship. Also think it’s interesting to hear from people in the different fields: hardware, devops, machine learning etc.

  1.  

  2. 35

    I’ll bite.

    General industry trends
    • (5 years) Ready VC will dry up, advertising revenue will bottomout, and companies will have to tighten their belts, disgorging legions of middlingly-skilled developers onto the market–salaries will plummet.
    • (10 years) There will be a loud and messy legal discrimination case ruling in favor of protecting political beliefs and out-of-work activities (probably defending some skinhead). This will accelerate an avalanche of HR drama. People not from the American coasts will continue business as usual.
    • (10 years) There will be at least two major unions for software engineers with proper collective bargaining.
    • (10 years) Increasingly, we’ll see more “coop” teams. The average size will be about half of what it is today, organized around smaller and more cohesive business ideas. These teams will have equal ownership in the profits of their projects.
    Education
    • (5 years) All schools will have some form of programming taught. Most will be garbage.
    • (10 years) Workforce starts getting hit with students who grew up on touchscreens and walled gardens. They are worse at programming than the folks that came before them. They are also more pleasant to work with, when they’re not looking at their phones.
    • (10 years) Some schools will ban social media and communications devices to promote classroom focus.
    • (15 years) There will be a serious retrospective analysis in an academic journal pointing out that web development was almost deliberately constructed to make teaching it as a craft as hard as possible.
    Networking
    • (5 years) Mesh networks still don’t matter. :(
    • (10 years) Mesh networks matter, but are a great way to get in trouble with the government.
    • (10 years) IPv6 still isn’t rolled out properly.
    • (15 years) It is impossible to host your own server on the “public” internet unless you’re a business.
    Devops
    • (5 years) Security, cost, and regulatory concerns are going to move people back towards running their own hardware.
    • (10 years) Containers will be stuck in Big Enterprise, and everybody else will realize they were a mistake made to compensate for unskilled developers.
    • (15 years) There will still be work available for legacy Rails applications.
    Hardware
    • (5 years) Alternative battery and PCB techniques allow for more flexible electronics. This initially only shows up in toys, later spreads to fashion. Limited use otherwise.
    • (5 years) VR fails to revitalize the wounded videocard market. Videocard manufacturers are on permanent decline due to pathologies of selling to the cryptobutts folks at expense of building reliable customer base. Gamers have decided graphics are Good Enough, and don’t pay for new gear.
    • (10 years) No significant changes in core count or clock speed will be practical, focus will be shifted instead to power consumption, heat dissipation, and DRM. Chipmakers slash R&D budgets in favor of legal team sizes, since that’s what actually ensures income.

    ~

    I’ve got other fun ones, but that’s a good start I think.

    1. 7

      (5 years) Security, cost, and regulatory concerns are going to move people back towards running their own hardware.

      As of today, public cloud is actually solving several (and way more than people running their own hardware) of these issues.

      (10 years) Containers will be stuck in Big Enterprise, and everybody else will realize they were a mistake made to compensate for unskilled developers.

      Containers are actually solving some real problems, several of them already were independently solved, but containers bring a more cohesive solution.

      1. 1

        Containers are actually solving some real problems, several of them already were independently solved, but containers bring a more cohesive solution.

        I am interested, could you elaborate?

        1. 1

          The two main ones that I often mention in favor of containers (trying to stay concise):

          • Isolation: We previously had VMs on a virtualization level but they’re heavy, potentially slow to boot and obscure (try to launch xen and manage vms your pet server), and jail/chroot are way harder to setup and specific to each of your application and do not allow you to restrict resources (to my knowledge).
          • Standard interface: Very useful for orchestration as an example, several tool existed to deploy applications with an orchestrator, but it was mostly executables and suffered from the lack of isolation. Statically compiling solved some of theses issues, but not every application can be.

          Containers are a solution to some problems but not the solution to everything. I just think that wishing they weren’t there, probably means the interlocutor didn’t understand the benefits of it.

          1. 2

            I just think that wishing they weren’t there, probably means the interlocutor didn’t understand the benefits of it.

            I’ve been using FreeBSD jails since 2000, and Solaris zones since Solaris 10, circa 2005. I’ve been writing alternative front-ends for containers in Linux. I think I understand containers and their benefits pretty well.

            That doesn’t mean I don’t think docker, and kubernetes, and all the “modern” stuff are not a steaming pile, both the idea and especially the implementation.

            There is nothing wrong with container technology, containers are great. But there is something fundamentally wrong with the way software is deployed today, using containers.

            1. 1

              But there is something fundamentally wrong with the way software is deployed today, using containers.

              Can you elaborate? Do you have resources to share on that? I feel a comment on Lobsters might a be a bit light to explain such a statement.

            2. 1

              You can actually set resource isolation on various levels; classic Unix quotas, priorities (“nice” in sh) and setrusage() (“ulimit” in sh); Linux cgroups etc. (which is what Docker uses, IIUC); and/or more-specific solutions such as java -Xmx […].

              1. 2

                So you have to use X different tools and syntax to, set the CPU/RAM/IO/… limits, and why using cgroups when you can have cgroups + other features using containers? I mean, your answer is correct but in reality, it’s deeply annoying to work with these at large scale.

                1. 4

                  Eh, I’m a pretty decent old-school sysadmin, and Docker isn’t what I’d consider stable. (Or supported on OpenBSD.) I think this is more of a choose-your-own-pain case.

                  1. 3

                    I really feel this debate is exactly like debates about programming languages. It all depends of your use-cases and experience with each technologies!

                    1. 2

                      I’ll second that. We use Docker for some internal stuff and it’s not very stable in my experience.

                      1. 1

                        If you have <10 applications to run for decades, don’t use Docker. If you have +100 applications to launch and update regularly, or at scale, you often don’t care if 1 or 2 containers die sometimes. You just restart them and it’s almost expected that you won’t reach 100% stability.

                        1. 1

                          I’m not sure I buy that.

                          Out testing infrastructure uses docker containers. I don’t think we’re doing anything unusual, but we still run into problems once or twice a week that require somebody to “sudo killall docker” because it’s completely hung up and unresponsive.

                          1. 1

                            We run at $job thousands of container everyday and it’s very uncommon to have containers crashing because of Docker.

              2. 1

                Easier local development is a big one - developers being able to quickly bring up a full stack of services on their machines. In a world of many services this can be really valuable - you don’t want to be mocking out interfaces if you can avoid it, and better still is calling out to the same code that’s going to be running in production. Another is the fact that the container that’s built by your build system after your tests pass is exactly what runs in production.

            3. 7

              (5 years) VR fails to revitalize the wounded videocard market. Videocard manufacturers are on permanent decline due to pathologies of selling to the cryptobutts folks at expense of building reliable customer base. Gamers have decided graphics are Good Enough, and don’t pay for new gear.

              While I might accept that VR may fail, I don’t think video card companies are reliant on VR succeeding. They have autonomous cars and machine learning to look forward to.

              1. 2

                (10 years) No significant changes in core count or clock speed will be practical, focus will be shifted instead to power consumption, heat dissipation, and DRM. Chipmakers slash R&D budgets in favor of legal team sizes, since that’s what actually ensures income.

                This trend also supports a shift away from scripting languages towards Rust, Go, etc. A focus on hardware extensions (eg deep learning hardware) goes with it.

                1. 1

                  (10 years) Containers will be stuck in Big Enterprise, and everybody else will realize they were a mistake made to compensate for unskilled developers.

                  One can dream!

                  1. 2

                    Would you (or anyone) be able to help me understand this point please? My current job uses containers heavily, and previously I’ve used Solaris Zones and FreeBSD jails. What I see is that developers are able to very closely emulate the deployment environment in development, and don’t have to do “cross platform” tricks just to get a desktop that isn’t running their server OS. I see that particular “skill” as unnecessary unless the software being cross-platform is truly a business goal.

                    1. 1

                      I think Jessie Frazelle perfectly answer to this concern here: https://blog.jessfraz.com/post/containers-zones-jails-vms/

                      P.S.: I have the same question to people that are against containers…

                  2. 1

                    (5 years) Mesh networks still don’t matter. :( (10 years) Mesh networks matter, but are a great way to get in trouble with the government.

                    Serious attempts at mesh networks basically don’t exist since the 200#s when everyone discovered it’s way easier to deploy an overlay net on top of Comcast instead of making mid-distance hops with RONJA/etc.

                    It would be so cool to build a hybrid USPS/UPS/Fedex batch + local realtime link powered national scale network capable of, say, 100mB per user per day, with ~ 3 day max latency. All attempts I’ve found are either very small scale, or just boil down to sending encrypted packets over Comcast.

                    1. 1

                      Everyone’s definition of mesh different, but today there are many serious mesh networks, the main ones being Freifunk and Guifi

                    2. 1

                      (10 years) There will be at least two major unions for software engineers with proper collective bargaining.

                      What leads you to this conclusion? From what I hear, it’s rather the opposite trend, not only in the software industry…

                      (5 years) All schools will have some form of programming taught. Most will be garbage.

                      …especially if this is taken into account, I’d argue.

                      (10 years) Some schools will ban social media and communications devices to promote classroom focus.

                      Aren’t these already banned from schools? Or are you talking about general bans?

                      1. 1

                        I like the container one, I also don’t see the point

                        1. 1

                          It’s really easy to see what state a container is in because you can read a 200 line text file and see that it’s just alpine linux with X Y Z installed and this config changed. On a VM it’s next to impossible to see what has been changed since it was installed.

                          1. 3

                            ate a container is in because you can read a 200 line text file and see that it’s just alpine linux with X Y Z in

                            I just check the puppet manifest

                            1. 2

                              It’s still possible to change other things outside of that config. With a container having almost no persistent memory if you change something outside of the dockerfile it will be blown away soon.

                          2. 1

                            Containers wont be needed because unikernels.

                          3. 1

                            All schools will have some form of programming taught. Most will be garbage.

                            and will therefore be highly desirable hires to full stack shops.

                            1. 1

                              I would add the bottom falling out of the PC market, making PCs more expensive as gamers and enterprise, the entire reason why it still maintains economies of scale, just don’t buy new HW anymore.

                              1. 1

                                I used to always buy PCs, but indeed the last 5 years I haven’t used a desktop PC.

                                1. 1

                                  If it does happen, It’ll probably affect laptops as well, but desktops especially.

                              2. 1

                                (5 years) All schools will have some form of programming taught. Most will be garbage.

                                My prediction: Whether the programming language is garbage or not, provided some reasonable amount of time is spent on these courses we will see a general improvement in the logical thinking and deductive reasoning skills of those students.

                                (at least, I hope so)

                              3. 15
                                • (5 Years) People will start having an “oh shit” moment in tech when they realize global warming is raining on the utopian parade.
                                • (10 Years) Large lobbying effort to put money into fruitless technical projects to address global warming.
                                • (15 Years) They call him the road warrior….
                                1. 7

                                  I think Alan Kay’s observation on pink and blue planes is relevant here. I’m too beholden to the status quo to see the left-field changes, only the incremental changes.

                                  With that context in mind:

                                  • the AGAF (Apple, Google, Amazon, Facebook) axis will thrive until the thing that displaces smartphones comes along. We give them far more latitude in terms of lock-in than ever have before (consider particularly the MS antitrust case, and compare with either iOS or Android restrictions on choice), they’ll milk the money until it dries up.
                                  • that next platform is not yet evident. Voice is limited to environments where I’m not already talking and don’t mind others talking, AR (currently, at least) seems too intrusive, wearables have not yet found the problem to solve.
                                  • it feels safe to say that node only has a few years left before cool start-ups looking to overspend on innovation tokens start meetup groups and medium posts about the next hot thing. There has been no meaningful advancement in ‘backend’ frameworks for decades, they just go in and out of fashion.
                                  • notifications will continue to grow beyond manageability, as every non-technical cofounder, growth hacker and product owner determines that their new feature needs to give itself a shout out.
                                  • all conversation moves to Slack. They combat the growing hate speech problem on their platform in the tried and tested way, by verifying Nazis and saying that in this case their community standards have not been breached.
                                  • ES2021 will contain the new feature everybody knows will be the decisive victory for JavaScript over its detractors. You know, that one, like that thing in the other language except does weird things if the second argument is an array. The new feature has entirely new semantics for this-binding.
                                  • In the thread for Alan Kay’s obituary on hacker news, people will share their adulation and describe how they carefully read all of his writing and that is what made them choose to go down the OOP route and get their first job using C++. The energy derived from him spinning in his grave will power us-west.
                                  • the Juicero founders will introduce hand-crafted, artisinal water made by carefully binding oxygen to hydrogen. They have accidentally invented a highly efficient rocket propulsion system, and have not noticed this.
                                  • by 2033 there is enough memory in computers for every end-user application to be made of Electron. Maybe, with no dependence on the native platform, this can truly be the year of the Linux desktop.
                                  • speaking of the Linux desktop, systemd continues to grow until it manages ACPI, audio, display management and autonomous weapons features. A bug in the last subsystem accidentally ends all war in the world, but a patch is swiftly released.
                                  1. 1

                                    all conversation moves to Slack.

                                    hahaha. You may want to qualify that statement.

                                  2. 6

                                    Not in a “field” per se, but an observation:

                                    The end of personal computing. Maybe in next twenty years or so. Artificial intelligence becomes more mainstream thanks to the hellbent nature of tech giants. “Personal assistant devices” would usher smartphones out the door and stick a fork in laptops for they’re done with. They’d resemble smartphones and tablets with stereo cameras and microphones with software meant to identify objects and people, play your music, pass messages to friends, and whatnot (all requiring an internet connection). No doubt these assistants will be locked down and unrootable for years. When augmented/virtual reality gets rebooted for the third time after the 3D movies and TV fad comes and passes, it’ll kick off for real this time. Eventually the screen will disappear from the face of the assistant and it’d even be possible to broadcast what one sees.

                                    Actual hardware to use becomes rarer to find as “it’s 204X, man, get with the times!”. There will be fewer software devs who grew up experimenting with computers. Most of these devs will have come from a game development background using some popular Game Maker-esque app. These future developers, system administrators, and software architects will be quickly hired (also likely by said tech giants) until AI is tweaked perfectly enough to replace them. A fringe movement will show up assembling homemade computers as a form of rebellion. Some plans may even catch on. The rest of the populace will eye these tinkerers suspiciously.

                                    tl;dr Smartphones/tablets replace laptops and desktops for 99% of people, laptops and desktops reserved for businesses (thus prices get jacked up), AI and {A,V}R take off once more after another metaphorical winter, and consumers almost end up like sheep herded blindly by their personal assistant.

                                    1. 6

                                      I don’t know what will happen in the next few years, but having been in technology professionally for almost 30 years, I think these things will remain true and help find the “next big thing”:

                                      • Universities will continue to crank out students that are too verbose in their writing. The goal in school is “2000 word essay on such and such”. In the working world, brevity and clarity are better.
                                      • Helping others is the best way I know how to keep up-to-date. People reciprocate when they know you are approachable.
                                      • Age ought not to matter. Do not talk down to folks, and don’t put people on a pedistal either. Treat everyone with respect. I live by the motto I learned from the late Dr. Walter Martin: “I respect a person’s right to believe whatever they want, all I ask is they respect my right to belive whatever I want”.
                                      • Attending and/or running a local meetup/club is worth the time and effort.
                                      • Care about your tools. Master some, be familiar with others, and don’t be afraid to ask for help with things you do not use very often.
                                      • If you tend to be an introvert, learn to talk more. If you tend to be an extrovert, learn to listen more.
                                      • Vim and Emacs will still be around. So will email.
                                      • Code reviews are hard, but they’re awesome. Keep doing them.
                                      • Attend conferences. If you have not given a talk and are afraid to, join a local Toastmasters.org club. People that can do public speaking will always go further in their career than those who chose not to speak.
                                      • Learn the basics of negotiating. I recommend an old series entitled “Roger Dawson’s Power Negotiating”. It’s corny, and only about the first half is worth listening to. You’ll be surprised how many people who are technically inferior to you will do better and have less conflict in life who know how to structure win/win negotiations. You’ll also be shocked how often we are always negotiating, in virtually every aspect of life.
                                      • It still will not be the “Year of the Linux Desktop” :)
                                      1. 6

                                        Lua will replace all scripting languages.

                                        1. 4

                                          I think in 5 years or so we might see first proper IPv6 rollouts as IPv4 space begins to dry up completely. Encrypted Protocols will probably be largely available but not widely deployed yet. In 10 years AI and Automation will IMO start taking over mechanical, simply jobs like garbage trucks or sorting toothpicks by size. In about 15 years people will still be writing in javascript and nobody knows why.

                                          In the meantime we might see some serious change in how people handle their opinions and political orientation. I’m however not willing to bet which side of the forked road we’ll go down but it’ll be a bumpy ride on either side. I’m not even certain which side is the better one at the current moment.

                                          Decentralization might start to matter. I realize people have been saying that for a good while now but with mastodon and similar efforts we’re getting mass-market ready decentralized tools without the overhead of being distributed. I predict that Mastodon will take atleast 5% of Twitter marketshare worldwide by 2030 or die out completely.

                                          There will (by nature of gut feeling) be a revolution in digital (and IRL) privacy, the EU might get a ball rolling with GDPR that other governments (Canada and the Asian Market) might follow. But it could just as likely die out and we get a 1984 surveillance google.

                                          1. 2

                                            In about 15 years people will still be writing in javascript and nobody knows why.

                                            Because there is so much Javascript being written now. An awful lot of it is still going to be around, and worse, it’ll be on the back-end AND the front-end.

                                            1. 1

                                              I dread the day when Javascript programmers will be needed like COBOL programmers are needed today.

                                          2. 3
                                            • Probabilistic programming will become much more important. Neural networks work very well if you have a lot of data, power, and time. Not only do humans not need millions and millions of hours of driving time to become proficient, but often all of data, power, and time are limited.
                                            • Message passing–and so microkernels–will stage a comeback. Processors just aren’t getting much faster anymore, but datasets keep getting bigger.
                                            • Not entirely tech related, but the cost of access to space will continue to drop, which will lead to malicious actors. Just take a look at this article from a few days ago. If you know people in aerospace, chances are they’ve at least at some point mentioned how vulnerable basically all the infrastructure is in space, to an even greater degree than on the ground.

                                            I think generally technology will keep getting better. Personal computing isn’t going to go away, software development will likely remain as easy to get into. I think software will continue to improve to the point that it knows essentially everything about you. The downside is that it then knows absolutely everything about you. While we lobsters might be prudent enough to make a show like Black Mirror as a reminder of people’s humanity, I guarantee that worldwide everybody is not as sensible.

                                            1. 3

                                              I’m restricting myself to predictions that seem to be already happening. So consider this in the spirit of Gibson’s the future is already here and just not fully distributed yet. Also historically, I usually get all my predictions spectacularly wrong and this might be more predictive as a set of things that don’t happen.

                                              Virtual Reality

                                              As pixel density increases in the next 5 years you are going to see VR in more common places. Expect to see stuff like Google Daydream being used on flights to watch movies. Expect to see pop-up sports centers doing VR games and events. Think stuff like laser tag, mini golf, or squash.

                                              Programming Languages

                                              Over the 5 - 10 year standpoint expect to see gradual typing enter the mainstream. As the languages figure out a way not to check the annotations constantly at runtime they will start to have a performance profile like today’s scripting languages.

                                              Expect in 5 years the first practical applications of dependent types to appear in the wild.

                                              Mobile

                                              In 5 years, a large set of user will do all their computer through their cell phones. They will have accessories for connecting them to a bigger screen and keyboard but no dedicated desktop / laptop.

                                              Cryptocurrency

                                              In 5 years, another cryptocurrency will supplant Bitcoin by basically fixing all the engineering issues in the protocol, the governance issues in the project, and delivering on the claims the technology originally made. This won’t be obvious until about a year before it happens. Most blockchain startups will be out of business or have been acquired by a bank at this point.

                                              Machine Learning

                                              In 5 years, you will start to see machine learning algorithms and systems become truly engineered. Between the need for fairness in ML, GDPR, and a general need for reliable diagnostics in devops expect lots of work into making ML system easier to debug and engineer with strong reliability guarantees.

                                              Expect to see more regular products using image recognition. People will create fan-films featuring celebrities that didn’t actually act in them, or are still alive.

                                              In 5 years, conversational agents will still be a niche without clear successes.

                                              In 5 years, there will be a startup which compellingly writes news articles for multiple languages.

                                              ML will start making inroads into other areas of CS. Expect to hear about new state of the art results where deep learning is used to augment network protocols, program synthesis, theorem provers, and compiler optimizations.

                                              With the exception of small geofenced areas, we will not have self-driving cars widely deployed.

                                              1. 2

                                                DevOps:

                                                • (5years) We will have more configurable and standardised hardware (especially networking). For example: hard firewall rules with k8s yaml files, JBOD/NAS configured the same way….
                                                • (10years) Google will probably rulez the internet.
                                                • (+10years) People will still try to hire consultants or any specific position to fix bad culture or bad management.
                                                1. 1

                                                  My field is ‘search’ (not SEO, but real low level search tech).

                                                  If I had to put a big what’s next prediction related to my field - we’re almost on the cusp of something passing a domain-specific Turing test (chatbots are really a search problem). I say domain specific - because we’re not even close to a general chatbot turing test pass yet. But when restricted to a narrow field of content and context, it’s going to happen soon.

                                                  1. 2

                                                    (chatbots are really a search problem)

                                                    Could you explain what you mean here?

                                                    1. 3

                                                      For the purpose of my comment, let’s say search is more or less synonymous with Information Retrieval. IR is the science of matching a query with data in a repository for returning the correct response to the query. A response is one or more pieces of data that is most relevant given the query context, intent, and substance. Chatbots extend this to a dialog of multiple query response pairs, and typically limits the response to either a single relevant answer, or a counter interrogative to gain more context, intent, or substance.

                                                      Many times the context and intent are unknown (the substance is part of the query), so the search is two part: finding the context and intent during query analysis, and then using that along with the substance to find the best response.

                                                      So take for example, the interrogative in an ecommerce chatbot: “what is the status of my order?” The context can be derived from the logged in user (the person asking the question), their data in the system, and the area of the site in which they are asking the question. The intent can be derived from the domain (this is important to my prediction) and query structure as ‘Order Status’. The substance is contributing to intent derivation in this query, as ‘status of my order’ can be an altLabel of the concept ‘order status’ (an important distinction for the prediction, because you have an ontology specific to the domain). In this example, when all three pieces are available, then you can structure the search to return an order (or list of orders) in your repository that are most relevant (relevant is likely recent unfulfilled orders). If you prefer you can have the bot ask a question to narrow the result set like ‘I see these three orders…which one are you interested in?’ and then fulfilling with a search based on the response to the follow up question.

                                                      The context and intent are the difficult part to solve. Restricting to a domain narrows the possibilities for both.

                                                  2. 1

                                                    Predictions:

                                                    • APL operations will start to sneak into mainstream programming languages
                                                    • Some platform change will occur that will allow a massively simplified stack.
                                                    • People will rediscover on-prem/embedded for some problems and migrate to them from the cloud.
                                                    • ML will generate code we can’t understand for common programming tasks, thus fulfilling two of the common cases for programmers.
                                                    • Scaling will become less important as more businesses move toward niches.
                                                    • There will be a push toward air gapped software.
                                                    • The AI personal assistant space will boom but is likely already captured by the big players.
                                                    • Culturally: the tech backlash will deepen with unforeseeable results.
                                                    • Culturally: the migration from large social media spaces to small private ones will continue.
                                                    • Blockchain-ish tech won’t be a big player in currency. It will, however, have a niche in the creation of durable records.
                                                    • The conversation around typing in program languages will become more nuanced (if I have anything to do with it).
                                                    1. 1

                                                      APL operations will start to sneak into mainstream programming languages

                                                      APL is a fairly old language. Why do you antifipate this happening now?

                                                      1. 1

                                                        Because:

                                                        1. The operations are useful.
                                                        2. People stayed away from APL/J for years because they are lexically scary.
                                                        3. The chaining style of programming seen in point-free Haskell/F#, LINQ/RX, Java Streams, and Rust have made a less control-flow style of programming seem less alien.
                                                    2. 1

                                                      In 5 years using vanilla Javascript to write backend systems will be widely regarded as a big mistake, as the burden of maintenance increasingly outweighs the supposed ease of development (personally, I’m there already, though I was never a Node fan).

                                                      In 10-15 years server-side, pre ES-2015 vanilla Javascript will be the new COBOL. There will be money to be made maintaining these systems, but few will want to do it.

                                                      1. 1

                                                        Assuming a tangent of the current trends

                                                        2 years: The cloud is just someone else’s computer. It was all a big mistake. The next big thing is very similar and it’s called distributed computing because it’s a more technical term.

                                                        5 years: augmented reality is cheap enough to replace touch screens. Cars drive themselves. Quantum computers are a viable purchase for any self-respecting company. Security methods are meaningless if you want to protect yourself from corporations or the government unless you’d like to pay thousands for them.

                                                        10 years: everyone is wearing some sort of AR device in their every day life to actually always have a screen in their face, at the convenience of still seeing what’s in front of them. These also do cool things like predicting every object’s path in space in real time. For example: if that car doesn’t speed up, there’s no way in hell it’s going to hit your car, here’s where that ball’s going to pass, here are the top 10 reasons to buy a new car this summer. #9: that probable dent wouldn’t have an effect on this brand’s new paint. Distributed computing is done with quantum computers and serves an actual purpose. Previous methods of security are 100% meaningless.

                                                        15 years: Everyone will be augmented in some way, everyone is constantly distracted with things that keeps their brain from being stimulated whatsoever, leaving no room for imagination or any form of thinking. Nobody knows how to drive a car and nobody can catch a ball by themselves anymore. Nobody really cares about that fact, because it really isn’t important. Everyone has a QPU (Quantum Processing Unit) along with their CPU and GPU. It really isn’t useful in many computers other than to offload some things. This feature is managed by operating systems developed by people who really don’t care about it and only worked on that feature for one month so that it looks good on the patch notes and gives people a reason to buy a QPU before becoming absolutely abandoned.

                                                        20 years: The last Hurd developer dies by coincidence. The operating system itself is 45 years behind anything considered modern.

                                                        30 years: A war fought exactly 200 years ago is completely forgotten, as are all of its effects.

                                                        1. 4

                                                          What war in 1848 are you getting at? Mexican-American war? Opium war? 1848 uprisings in Europe?