1. 17

A certain comment on HN recently triggered me to write down a few thoughts and off the sofa predictions/suspicions I have for long/medium term stuff that I think may happen in tech (software/hardware) in our lifetimes. Given that some earlier predictions I kept purely in my mind in the past did actually happen, I thought it could be a fun game to put them down this time, and share with you, and also ask what are your thoughts and armchair predictions! I’d be very interested in reading them. Especially with the ending year, the time seems really well suited for such kinds of pondering…! :)

Very sorry if that kind of thread is already expected to be posted more officially at the very New Year’s Eve or any other later time here; I don’t know if there’s such a tradition in the community; I’d be grateful if you could kindly reply here if that is so, and probably disregard this post then; I’m open to deleting the question in such a case, or having it merged to the potential future thread by moderators, if that’s possible and deemed fair. Also, please let me know if this thread is somehow not-really-OK for this community.


I don’t want to hijack the main post with my own thoughts, nor to railroad/influence your answers prematurely, so I decided to write my own “predictions” in a regular response below. If, on the other hand, you’d like to understand better what I mean, or just would like to read a trigger for your thoughts (as happened to me), or would want to discuss, please do see my, or other, responses below! Also, please feel free to see it as a very open question, and with as long or short timeframes as you prefer (though personally, I’m mainly interested in “long”/“medium-length” shots; more precisely, what I might actually witness in my lifetime, but not small-scale short-term stuff like “Go modules will win over Godep, or reverse”).

  1.  

  2. 17

    I expect the AI hype bubble to burst, and self-driving car projects to be mothballed as being impractical and too expensive.

    1. 4

      The self-driving car thing has a long tail. Getting a self-driving car to 98% capability is “easy”, but I think the last 2% (without which the project is useless) is still decades away.

      1. 6

        I guess it depends on what the last 2% is, but I don’t think they’re useless without every single thing working perfectly? Perhaps they wouldn’t be suitable for sale to the general public as go-anywhere vehicles (which is certainly what I want!) but I can imagine that you might be able to use them in limited cases as fleets of taxis, geo-gated into only the areas they’re capable of working reliably in. Or perhaps for long-haul trucking - autonomous on the interstate, driver required for city roads and for parking.

        1. 7

          I do see self-driving long-haul trucking coming in the next decade, perhaps with a human-operated lead truck and definitely with a remote operator somewhere that can give instructions to a truck that gets itself into a situation it doesn’t know how to handle (e.g. it pulls over because its GPS map doesn’t match up with what it’s seeing on the ground).

          The final 2% are things like handling an ambulance behind you with its lights on, roads without painted lanes or with soft shoulders, roads that are closed, lanes that are diverted, etc.

          The final-final 0.5% is handling special instructions. What does a self-driving car do if there’s a police officer in the middle of the road directing traffic? A light-up sign that says “all traffic left-lane?” A “bridge out ahead!” sign, etc?

          1. 3

            It’s already seeing production adoption in mining and agriculture, especially in dangerous areas that have OSHA limits on how long someone can drive. John Deere and Caterpillar have been working on this for a long time.

          2. 1

            A 98% solution would be far from useless, if it leads to a better safety record.

            The places where automated driving could improve safety (fast response to emergencies, warnings of dangerous traffic conditions) are separate from places where automation is likely to break down (communication between drivers in complex traffic interactions.)

            1. 1

              Id use it like Google Maps uses it. It drives in a situation computer can handle. If not, I drive. Hell, just letting it do highways and Interstates for long hauls while I chill would be great.

            2. 3

              If you have a 98% solution, you could deploy it to a limited region where it is a 100% solution. For example, Voyage deploying to The Villages, FL.

              1. 2

                Exactly! How much time and money are people willing to dunk in this bottomless pit? At some point it should dawn on people in the industry that this simply isn’t cost-effective.

            3. 9

              Moore’s Law will end in the next decade. This triggers a crisis in the sense that just parallelizing your software is not enough anymore. You have to deal with heterogeneous hardware in addition to that. Some cores/nodes/CPUs are optimized for crypto, others for media en-/decoding, others for low energy IO, etc. Weird stuff like ARMs LITTLEbig is just the beginning.

              I don’t see predictions as binary but as probabilities. For this prediction I would assign 70-80% that it comes true. In other words, I would bet 2:1 for or 1:5 against it.

              1. 4

                Hasn’t Moore’s Law already ended? Transistors are already on the scale of dozens of molecules wide, there’s not much room left to go down physically, CPU clock speeds (which is what I think most people actually care about when they talk about Moore’s Law) haven’t been increasing for a decade or so, and chip manufacturers have been having a hard time scaling their process down further for years now.

                1. 3

                  The thing with clock speeds is Dennard scaling and that has stopped over 15 years ago. Moores Law is about transistors per area. We scaled further through parallel cores instead of clock speed. The last time I checked was with the AMD Ryzen release in 2017 and Moores Law has continued until then.

              2. 7

                I’ll only speculate on the next half-decade (0-5 years).

                Industry trends:

                The tech bubble collapses. Interest rates change, some high-profile company collapses (my money is on one of Uber, Twitter, Tumblr, Facebook, Magic Leap, Coinbase), and investors suddenly realize that basic business economics actually matter again. In turn, portfolio companies are encourage to tigthen belts, and a surplus of nominally skilled workers drives wages down across the industry–and removes the tolerance for vanity projects on company time.

                Mozilla and Firefox die. The mission creep of Mozilla as well as its dependence on ad revenue and search partnerships is going to kill it once the bubble collapses. Some useful things, like the MDN docs and possibly the Firefox codebase itself, are gonna survive through various forks. Rustlang probably survives through its community.

                There is gonna be a massive drop in wages for programmers. Either while dealing with unions (explained below) or with visa stuff or a change in the type of work, we’re gonna see market forces make all of us poorer.

                Language trends:

                Rust fails to provide a compelling alternative to C, Go, or Haskell, but remains an existential threat to C++. Between the bubble collapse taking out the main source of funding for the Rust Evangelion Strike Force, the decrease in wages for programmers, and the fact that honestly most of the useful systems stuff is already implemented there just isn’t going to be a vibrant niche. Rustlang is going to be a source of continued exasperation for the true believers for decades to come, as were lisp and smalltalk before it.

                JS is going to continue to be the dominant language in production around the world, despite a cancerous language spec and anemic standard library. The community continues to be full of perpetual novices, but the language strength for banging things together under budget and on a tight deadline cause it to continue to flourish.

                Go continues to be a boring, mediocre, and utterly successful language used by people that need to get things done–it’ll probably make up most of the daily ops and ‘real work’ tooling we rely on. Not much to explain here, and its main sponsor (GOOG) isn’t likely to stop making payroll soon.

                Elixir/Phoenix will supplant Ruby/Rails, and to a good extent frontend JS, and possibly show up as a major player in embedded systems. Not too much to say here…I don’t expect it’ll be vastly more popular than it is, I don’t think the web braindamage is completely set in yet, and I think people will just be quietly productive in it. You’ll see it a lot in contract work, but it’s not going to be a crazy huge deal. Those that know know, and those that don’t won’t.

                Assembly will still be useful. The magical myth of a competent autovectorizing compiler will continue to evade language ecosystems that aren’t Fortran, C, or C++.

                Formal methods will be a secret weapon, but not a useful one in most cases. With the collapse of the market, there will be even more emphasis on lean and sloppy engineering, and in the vast majority of domains the resulting lack of quality and correctness will be the market’s expressed preference.

                Ops trends:

                Containers/k8s will turn out to be a huge boondoggle outside of very specific usecases. The bubble forces developers to focus on their jobs instead of buzzword chasing, and all of the sudden the mundane solution of “run this locally and document your shit” becomes preferable to teams when compared with “pay for dedicated devops people, pay a hosting platform megabucks to babysit containers, add lots of overhead to solve problems that don’t exist in simple systems”.

                BSD derivatives become the standard ops platform, because of Linux douchebaggery. This is a little more optimistic, but the renewed emphasis on careful systems administration mentioned previously suddenly mean that spinning up and shutting down machines and containers instead of actually maintaining them becomes a cost-ineffective choice. That, and the bazaar having become too bizarre in its offerings (systemd, etc.).

                Monitoring remains mostly the same. Dashboards and automated alerting look basically the same they do now, especially as market contractions kill off really novel/weird business ideas that would create demand for something else.

                Cultural trends:

                We’re gonna see our first big ideological purge. My money is split on which project, but I think we’ll see the first big push to deplatform/remove/purge/destroy some integral group or BFDL. Further, I’m pretty sure that this is gonna come from the left. I wager it’ll take another few years to see if it cripples the project in question.

                We’re gonna see increased work to make programmer guilds or unions. This will be fought tooth-and-nail by righties in tech, but it’ll happen. It will bring some definite benefits, but is going to hurt wages for a time while companies adjust. There’s gonna come a time where things come to a head and during a general strike we find out that, honestly, we’ve been living off of a really inefficient system and companies haven’t looked too hard.

                We’re gonna see increased backlash from users as they realize just how much of their data we’ve been giving away. There will be show trials, some engineers are gonna be served up by the companies, and entire chunks of the industry are going to have a Very Bad Day when legislation finally catches up with them.

                The pendulum is gonna swing back away from liberal/progressive policies. With the collapse of the bubble, keeping your job is going to be more important than tweeting your beliefs. Blacklists formed for unionbusting will be just as useful for getting rid of people who are outspoken about progressive things (and especially at first these lists will overlap a lot). Within the ranks of programmers, the follow-on from the purges will make people realize the existential threat of allowing certain folks in their projects, in turn having a chilling effect on otherwise healthy perspectives.


                In my opinion, we all need to buckle up and get ready for a really rough ride.

                1. 3

                  The tech bubble collapses. Interest rates change, some high-profile company collapses (my money is on one of Uber, Twitter, Tumblr, Facebook, Magic Leap, Coinbase), and investors suddenly realize that basic business economics actually matter again. In turn, portfolio companies are encourage to tighten belts, and a surplus of nominally skilled workers drives wages down across the industry–and removes the tolerance for vanity projects on company time.

                  I tentatively agree with this, and I wonder if some of it could possibly be driven by fines/legislation against big corps due to pressure from constituents for both privacy and voter manipulation reasons. As to which company… my current bet is on either Facebook or Amazon being forcibly split up. I could see also see Uber running out of gas (pun intended).

                  Aside: I don’t think Tumblr is even worth discussing though, as it was bought by Verizon as part of yahoo, and it appears they (verizon) are already trying their hardest to kill it.

                  1. 2

                    There is gonna be a massive drop in wages for programmers. Either while dealing with unions (explained below) or with visa stuff or a change in the type of work, we’re gonna see market forces make all of us poorer.

                    What kind of market forces could these be? I’m trying to think from the point of view of pure supply/demand. If there are enough people who need software to do their work better, the supply should remain fairly consistent, no? Of course, if the jobs that others are doing are in trouble (for whatever reason), that would directly mean programmers are in trouble. But curious what your chain of thought here is.

                    1. 4

                      If companies that are paying outsize salaries to programmers cease to be viable due to market forces (a collapse in value of ad sales or something), then they’ll be forced to fire employees, leading to more programmers vying for the same job. So supply will outpace demands for jobs.

                      Really tho, wages for programmers in the American tech sector are aberrantly high. It seems likely there will be a correction. Wages will stay high, but not astronomical. And that’s fine.

                      1. 1

                        If a massive employer like Accenture, IBM, or DXC goes belly up it’ll flood the market with 100’s of thousands of unemployed programmers, and the supply/demand ratio will change.

                        1. 1

                          But are there any market forces acting in the next 5-10 years that could make those big companies go out of business?

                          1. 1

                            Maybe? Enron, WorldCom, and Bear Stearns didn’t look like they were going out of business 5-10 years before they declared bankruptcy.

                            1. 3

                              “When the tide goes out, we’ll see who’s been swimming naked”.

                      2. 1

                        Christ, those cultural predictions are pessimistic friendlysock, and some of them are things I worry about myself. Do you have any idea about what practical steps “getting ready for a rough ride” might look like for the average programmer?

                      3. 6

                        The pessimistic me would say:

                        • A ever increasing set of the web will be offered as a walled service, only using the web as a foundation for a more controlled network.
                        • Hardware and software will become more closed in practice (even if they are open source in name).
                        • Traditional desktops/laptops will loose, while more questionable portable devices will gain in popularity
                        • Energy-consumption by server-farms/consumer-devices will probably rise
                        • New social networks will arise and fall at greater speeds, not always transparently and will be fueled by simple but addictive offers.

                        The optimistic me would say:

                        • With the rise of RISC-V a market and a consciousness for open-hardware will arise, lowering the barrier for pure free software systems
                        • Distributions like Nix and Guix that make it easier to build software reproducibly will become more popular
                        • Alternative networks around the Fediverse will gain more popularity, and (for some reason) will still be able to survive in a actually distributed fashion avoiding addiction-tricks
                        • The split between users and developers will shrink, as the abilityof “programming”/understanding of computational thinking becomes more widespread due to a higher focus on it in public schools (as a form of “literacy” and ability to self-emancipate oneself).
                        • Energy-consumption by server-farms/consumer-devices will probably rise less that the pessimistic me belives
                        • Blockchain/AI-Hype will loose steam (connected to the previous point about literacy).

                        The idealistic me would say:

                        • The modern web will realize that many of it’s developments have been mistakes and there will be high-level attempts at trying to moderate it to become more sane (web-engine wise, the centrality of browsers, overuse of HTTP, etc.)
                        1. 2

                          Hardware and software will become more closed in practice (even if they are open source in name).

                          Intel’s Software Guard Extensions and the Intel Attestation Service are super-scary in this respect. At least they seem to be opening up to third-party attestation services, now.

                          1. 2

                            Blockchain/AI-Hype will loose steam (connected to the previous point about literacy).

                            Possibly unpopular opinion, but I’m not too sure about blockchain losing steam. This video explains the reasoning much better than I can.

                          2. 4

                            Here are my crazy predictions. Keep in mind all my tech predictions have been wrong so far. Get this, I told one of the co-founders of GitHub (before there was GitHub) that ruby is slow and git has a terrible user experience and nobody will use it. That web programming is a fad and C++ will make a comeback (maybe latter is slowly coming true? now I can do C++ in the browser with webassembly!). With all that in mind and my terrible record, here goes nothing.

                            short term

                            • mobile computers saturate user base and growth slows.
                            • bigger/better batteries in mobile devices mean dynamic languages are good enough again for mobile.
                            • cloud computing becomes saturated and growth slows.
                            • desktops/laptops become more expensive with fewer options.
                            • AI finds little use outside of “public relations”.
                            • The decline of Apple starts.

                            medium term

                            • power management tech (smart grid) becomes popular because governments attempt to use carbon-taxes making electricity more expensive.
                            • efficient computing becomes important again.
                            • farm tech becomes more important.
                            • web companies start collapsing.
                            • advertising industry collapses taking down google/FB.
                            • development of decentralized/ad-hoc networks.

                            long term

                            • we fail to properly address global warming resulting in huge social upheaval.
                            • we enter a new dark age putting a stop to most technological development.
                            • efficient computing, recycled computing is the only computer. All software is FOSS.
                            • no more software or computer industry.
                            • organized human life is no longer possible.
                            • humanity was a nice experiment.
                            1. 4

                              Publicly-writable social media (basically every online community that isn’t paywalled or invite-only) gets regulated out of existence, due to a combination of the adtech bubble bursting and public outcry over privacy violations, voter manipulation, astroturfing, offensive crap, and spam.

                              1. 6

                                Writing fast code will be actually be pointless.

                                What got me thinking about this is this: My webserver is probably the fastest HTTP/1.1 implementation that is possible - there’s nothing to remove, it’s 100 or so lines of code, and seriously: I’d love to see anything someone thinks is faster.

                                The next fastest, is within a few percent though. The next fastest is thousands of lines of code; they’re doing much more than 10x more work than me - they have more branch mis-predictions, they can spill L1. None of it seems to matter anymore.

                                That means our modern CPUs are so quick they can do at least 10x more work, throw it away, and still get the packet out to a 1Gb network interface in time for the network buffer to flush (or a disk buffer, or any other kind of IO). When the new 7nm process popularises I expect the gap will be gone entirely, and that’s just a few years away.

                                We’ve got 10Gb adapters though, that’ll push things back maybe as long as a decade. Hopefully we’ll get those 100Gb adapters before then…

                                1. 3

                                  “I’d love to see anything someone thinks is faster.”

                                  One idea I saw in an embedded server was pre-converting the web pages into TCP packets that just get fired off directly. I cant remember if I tried or timed that. Might want to experiment with caching TCP packets generated on startup for all content or first request for Might be hardware versions of this concept with RDMA or something.

                                  1. 3

                                    Xok exokernel from MIT. File system block size was aligned and offset so you could basically mmap a file and splat it into packets.

                                    1. 1

                                      That’s clever. Thanks for the tip.

                                  2. 2

                                    I think this is only one facet of a bigger change. I think “programming” will no longer be seen as one monolithic pursuit. Fast code will continue to be extremely important for some activities, e.g. gaming, trading, simulations.

                                    1. 2

                                      With CPU performance increasing at a snail’s pace, writing fast and efficient software might become an important feature again.

                                    2. 1

                                      My webserver is probably the fastest HTTP/1.1 implementation that is possible - there’s nothing to remove, it’s 100 or so lines of code, and seriously: I’d love to see anything someone thinks is faster.

                                      I would love to see a webserver in K ;)

                                      1. 2

                                        We’d need sockets for K first :)

                                    3. 3

                                      I’m glad that this thread came up because we are getting our predictions for 2019 at KOPN Tech Radio and I am hoping I can burrow a couple of predictions that are better than mine (due credit will be given of course). In the spirit of contributions here are some of my personal predictions for 2019 that I’ll be talking about on the show:

                                      1. Bitcoin and Crypto Currencies will raise again pushed by developments in blockchain based applications. There has been a lot of exploration of the use of blockchain-based systems in the industry. Some examples are uses in the Shipping industry, the AAIS considering using a hyperledger based system to streamline state-required reporting.

                                      2. The A.I./Machine Learning winter is coming. Over the last couple of years there has been a lot of growth in Machine Learning, and a lot of success in that area. However, there are a couple of factors that will curb the enthusiasm surrounding Machine Learning

                                      One of the issues facing Machine Learning is that as it’s popularity grows, the average quality of it’s practitioners goes done. We are already seeing corporate consulting companies peddling standard statistical analysis packages as “Machine Learning” and failing to deliver on their promises. Another example of this deterioration is that with the large number of vacant “Machine Learning” position any one that has completed a Tensorflow tutorial can now call them self a Machine Learning Researcher/Engineer/Practitioner.

                                      Another factor facing Machine Learnings growth is that (IMHO) we are reaching the limits of what can be done given our current tools and paradigms. The No Free Lunch Theorem makes is clear that there is no “Universal Learner” and the better an algorithm is at solving one problem, the worse it is at solving another.

                                      1. The web will become both more open and more closed, with both parties becoming more and more detached from one another.
                                      1. 3

                                        Mid term future—Google will fade. In the 60s, IBM was controlling the industry such that it’s hard to imagine if you weren’t there. Their time has passed. In the 90s, Microsoft was controlling the industry such that it’s hard to imagine if you weren’t there. Their time has passed. So too, will Google. I’m expecting in ten to fifteen years, Google falls to background noise.

                                        Mid to long term—general purpose computing is dead, or will be a really expensive hobby. There’s no market for it (as most people don’t want to know programming) and securing computers is still a hard thing. Given that, it’s just cheaper over all to produce dedicated hardware for a particular job. Much better if you can make it last only a year or so, in order to get recurring revenue (aka “rents”).

                                        What I hope never comes to pass—all protocols are just HTTPS, and the only programming language is JavaScript, and x86 still reigns surpreme.

                                        1. 2

                                          I predict we’ll see a lot more Typescript on the server side in the next year. If that happens I think it could be a nail in the coffin for Elm, ReasonML, and the other statically-typed compile-to-JS languages, which are front-end only [0]. I predict Typescript will be good enough for most people who want one statically typed language to do (most) everything in. I think the deno project, with its focus on security and fixing the warts of NodeJS, might be the tipping point for TS adoption.

                                          [0] Ok, ReasonML isn’t front-end only, being OCaml, but sadly I don’t see a whole lot of momentum there.

                                          1. 1

                                            Rust and Typescript, together, I see dominating the world very soon, if not already behind the shadows.

                                          2. 2

                                            In the long term, computers will largely program themselves, and the job formerly known as “programmer” will be replaced by “archaeologist”, requiring a very different set of skills.

                                            1. 3

                                              This is a subplot in Vinge’s A Deepness in the Sky.

                                              1. 1

                                                I didn’t know that, but it seems so obvious to me.

                                            2. 2

                                              Long term: datacenters move to outerspace for simple reasons, 1. superconductivity allows raising clock speeds to around 700Ghz, 2. distance to geostationary orbits is rougly equivalent to distance to nearest datacenter, 3. lifting matter into orbital space becomes faster and cheaper, 4. near absolute zero temperature is cheaper to keep computers cool in, compared to typical temperatures near earths surface, 5. energy is provided by nuclear reactor, since less environmental problems, 6. larger elliptical orbits are used for non-realtime computations, effectively filling space: shadowing the sun on earths surface controls global temperatures.

                                              When we speak of the “cloud,” the next generation will point upwards into the sky. And for good reasons.

                                              1. 9

                                                Dumping heat in outer space is actually really hard – where do you put the heat?

                                                Data centers submerged in oceans would be a far likely scenario.

                                                1. 1

                                                  Current satellites vent heat via radiation, right? Maybe it wouldn’t be practical at “DC scale”, but I feel like with a big enough surface area (?) you could radiate heat out into space, assuming that you don’t mind being lit up like an IR beacon :-)

                                                  1. 1

                                                    Point 1 prevents generating heat, otherwise you loose superconductivity. Therefore, computations are most of the time reversible, eliminating the generation of heat in the first place.

                                                  2. 0

                                                    I am glad to hear that this idea is still living Hans! I especially like number 6!

                                                  3. 2

                                                    My original pondering that made me start this thread — I’m copying it verbatim from HN:


                                                    Personally, in my heart, I’m kinda keeping fingers crossed for a future where if one needs to/has to, one uses Rust (mostly where one would use C/C++ today, including for “squeezing the last drop of performance”, and especially high risk everpresent libraries like libpng; plus as an extra, probably for highly ambitious parallelism like in Servo); otherwise, one just goes with Go (for the amazing general productivity and readability).

                                                    Though then there’s also Luna-lang, which I’m hoping even stronger to become long-term a new breakthrough paradigm/tech in programming & scripting…

                                                    On a mostly unrelated note, but continuing with off the sofa predictions, I’m curious on a few more developments:

                                                    • whether WebAsm will become the new de-facto universal architecture — at least for VMs, but I wonder if we won’t eventually end up living on purely WebAsm CPUs at some point in future? (Unless WebAsm has some inherently hardware-hostile design decisions; I’m not a CPU/ISA guy to know that.)
                                                    • then RISC-V; will it become the ultimate personal computer/device CPU in the meanwhile? Also, what may happen if both RISC-V and WebAsm start naturally competing in this domain eventually?
                                                    • Fuchsia off-handedly dethroning the Linux kernel & ecosystem (and maybe even Windows) as the sudden standard mainstream OS (and thus also, amusingly, suddenly swinging the Tannenbaum’s argument up through a semi-random caprice of Google’s deep pockets)?

                                                    Especially per the last point, I wonder if we’ll eventually end up in a world with much more… “secure”… personal devices. And if yes, will it end up being net good or bad (or just neutral/hard to say/nuanced, a.k.a. both, as usually) for humanity. Given that I’m recently realizing that “secure” unfortunately seems to be a very close sibling, or even maybe just an other face of, “closed” — think DRM, walled-garden ecosystems, no more rooted Android or XBOX… or even no more control at all over your PC, a.k.a. no general computing for the masses, potentially. Though I personally stay mostly hopeful on this front; in part because I think I believe people are as a sum too chaotic for this to be able to happen completely.


                                                    As an extra, now that I posted the above, one more thing I thought of: on the front of potentially breakthrough hardware developments, I personally currently don’t really believe in Quantum Computing becoming “usable” in my lifetime (meaning being practical enough to be applicable to real problems, in a way more effective than “classical” computing, i.e. to break primes-based cryptography, or stuff like that).

                                                    1. 1

                                                      long term, complete speculation (~10 years)

                                                      • one or more of FAANG level corps will be sold off at fire sale prices due to user exodus and/or investor exodus
                                                      • the smartphone paradigm led by the iPhone will dis-integrate again into special purpose devices
                                                      • commercial (and startup) drug companies will start marketing and synthesizing new drugs for recreational drug users
                                                      1. 1

                                                        Platforms like YouTube, Facebook, Google, Twitter will struggle to remain “neutral” as people continue to post vast amounts of offensive junk, radical political stuff, porn, etc. There will be a lot of controversy around “AI” systems for censoring content. There will be high profile court cases and legal battles around public content platforms and their responsibilities.

                                                        Encryption and privacy will become household topics of importance, maybe especially after some new authoritarian governments in Europe use unencrypted stored data in some awful way (religious oppression, etc).

                                                        1. 1

                                                          Payment services will continue to grow and diversify for some time until we end up with a duopoly or similar there as well.