Lost my bet: the PC isn’t dead… yet

I can’t predict the future, nobody can… but I think betting on the future is a good intellectual workout. It forces you to think beyond your day-to-day activities.

In 2012, I bet $100 with Greg Linden that by 2015, tablets would have replaced PCs. Greg wrote an analysis of the outcome but the short story is that I lost.

What is a PC? We did not define it back in 2012, but I think we should, just to clarify. You have a screen, a physical keyboard… and, typically, a Microsoft operating system coupled with an Intel processor.

I hoped in 2012 that tablets would somehow find a way to supplant PCs in the office. I did not know how it might happen, but I hoped. Tablets haven’t managed it. I think that tablets have had their chance.

The big story in the last three years was neither PCs or tablets, but smartphones… They have gotten huge and extremely powerful. Let us review the specifications of the latest Samsung phone: 2560 x 1440 pixels, 4GB of RAM, 8-core processor. Though we can’t compare a phone with a PC, it does feel like smartphones from today could give PCs from 2012 a run for their money. In 2016, nobody cares for PC software. Not even Microsoft I think. If your app does not work on mobile, it is broken, period. I even had to reconfigure my blog so that it looks nice on mobile… The managers I know all work from their smartphones. I have not seen a manager carry a laptop in a long time. I am told that everybody owns a smartphone.

I was wrong, but I don’t think I was wrong in the sense that I overestimated the rate of change. I mispredicted the direction of the change. I could not imagine smartphones getting so big and powerful… I could have imagined millions of refugees roaming the Earth due to a terrible civil war in Syria… but I could not have imagined, only three years ago, that most refugees in 2015 would carry smartphones. I feel that the change was as drastic as I imagined it, but in a way that I could not foresee.

I still think that the PC is falling off a cliff. There is lots of inertia, but with everything available on the cloud and through mobile apps, most people would be better served by something other than a PC. I feel that betting again for the demise of the PC would be boring: though I did not win my bet, the PC is effectively irrelevant today.

My interest has moved to forms of interactions. What holds us back to the PC is the keyboard. There is simply not a good substitute for the keyboard right now. We could do away with the monitors. We could do away with Microsoft Windows and Intel processors… But we are still shackled to the keyboard.

Voice and gesture recognition is probably the future for most people. I still fill out paper forms weekly, even though it is thoroughly obsolete technology. I think that in 10 years (circa 2025) the keyboard will be similarly obsolete.

In any case, I will be donating $100 to Wikipedia. Greg had decided to match my donation even though he won the bet. That’s very gracious of him.

Tweet about this on TwitterShare on Facebook8Share on Google+13Share on Reddit0Email this to someone

14 thoughts on “Lost my bet: the PC isn’t dead… yet”

  1. You are right to say that the reason you lost your bet (and the reason you would lose it again if you made a similar bet about 2019) is the keyboard.

    Voice isn’t going to replace it, either, for one simple reason. It doesn’t matter how good the recognition technology is. The simple fact is that no one is going to move over to voice as long as other people standing around can hear your private emails because you are dictating them to the computer.

    So no, the PC isn’t irrelevant and won’t be anytime soon.

  2. I also thought, that people would shy away from talking in public about their private things on the phone. But i experience it often. So they could also dictate openly,i guess.

    But i think it is not very efficient. I also am slow to type on a touchscreen. And spooked by watching a screen in general, which for me is an issue with VR.

    I want to see electronics disappear. I am tired of things,updates, upgrades,replacements. I would like to have them simply integrated into my life. My own sensors work. I want to use them all. Make my extremities glow, my eyes sharp, my ears effective. And allow me to switch it off.

    I totally love those head up displays in cars. I also wanna be driven home. I dont want to type or say where i live. It simply knows what i just think where that is.

    And then i sit in front of my powerful PC and think about expressing myself. I feel very limited. I can program, but i often lay in bed and try to think a problem through, because it is easier. But i get lost. Something that holds a thought and later connects the dots.

    I am very sad about interfacing with electronics these days.

    blubb

    Maybe

  3. I sure hope the keyboard is fading in relevance by 2025. It’s had a long enough run, and it’s holding us back now. There was a fad of chorded glove “keyboards” in the late 90s, and we could easily do much better with today’s sensor technology and machine learning. And I predict that eyesight-based text entry will be a lot more efficient than people currently expect, again due to improvements in sensor technology and ML. And I definitely think we have the technology today to build text input driven by electrical signals from the muscles, allowing for chorded glove keyboards that require no glove. We are controlling robots today through neuromuscular signals that don’t even require the person’s arm to move (the person just needs to imagine moving), so I could see rapid text entry with no visible movement ala “Rainbows End” within 10 years. And of course, we are already testing neural implants with rapid advances in materials science.

    Note that the point about keyboards being private is becoming less reliable with each passing year. There now exist ML models that can reverse what a person is typing simply from the audio, and models to reverse from electrical emissions have long existed. As sensors become more ubiquitous, it will be easier and easier for machines to passively know what is being entered on nearby keyboards. Hell, you could imagine a standard software feature of phones in 2025 being the “keyboard mode”, where you could bang away on any detached keyboard of any defunct device (or even on any surface), and the phone would automatically use the audio/video/electrical as an input device for reliably entering text. We could build that with today’s technology.

    1. The success of the keyboard is based on human biology. We’re really good at using our hands for fine manipulation of physical objects. A ridiculous fraction of our brain capacity is devoted to controlling hands and sensing with them.

      When something eventually replaces the keyboard, it’s probably going to be some kind of bidirectional neural interface or an even more efficient way of using our hands for manipulating and sensing. Clever tricks like eyesight-based input may be useful in some special cases or for augmenting primary input methods, but ultimately they’re just replacing something we’re good at with something we’re bad at.

  4. I had a very nice desktop computer (for a software developer). It died a few years ago. I only noticed as my grandson was using it.

    I have two high-end MacBook Pros, two Chromebooks, and three tablets. (In truth one Chromebook and two tablets are for my grandson.)

    Not sure I will ever build another PC.

    You were not far wrong. In past years, I needed a beefy local machine for development. Going forward, that beef looks to be pulled from the cloud. But I still need a good keyboard, and lots of display acreage.

    The dots do not all quite connect, as yet.

    From extended sessions, I need the excellent keyboard and multiple large displays I had in my study, with the dead PC. Might wire up a Chromebox, but … not quite there. Want to fire up local containers on the Chromebook, or remote containers in the cloud, as needed. Possible … but lots of unfinished edges.

    Tablets are suited for consumption, not production. For production you still need a good keyboard. At the same time, the existing desktop PC depends on rapidly vanishing basis.

    A transition point, if still rough.

  5. “typically, a Microsoft operating system coupled with an Intel processor.”

    😀 Glad to know I made it past the “PC era”, I have two boxes both of them Linux and AMD.

  6. I think we’re all talking about a model where users have a thin client and pay for computing/disk/etc in the cloud. This is what I would call death of PCs.

    In a very similar vein, desktop application landscape and demand for programmers is changing. Traditional heavy client are replaced with lighter, web versions connecting to mother ship. These apps won’t even have to be converted to mobile version. They run in the browser.

  7. I’ve been using Emacs for 23 years now. I believe I will still be using Emacs, with a keyboard, in 2025. However, all bets are off if we talk about 2035. Part of me hopes I won’t be needing a keyboard in 2035, although I expect to still play the piano then and I expect the piano to survive for another hundred years as a preferred input device for creating music.

  8. Producing text is a huge industry and the keyboard has no real competition. I cannot imagine one.

    Voice does not work in mass on a train, in a cafe, in open-floor offices, and other packed spaces. Text must be created in silence.

    Gestures are too tiring. Find an input method, which requires less movement than my fingers on my keyboard.

    Writing via brain waves will still take some time. Maybe AI text prediction will advance in the next decade, so my phone not only correctly predicts the next word, but whole sentences and paragraphs after a few characters.

    1. (…) the keyboard has no real competition. I cannot imagine one.

      It is really hard to imagine the future.

      Voice does not work in mass on a train, in a cafe, in open-floor offices, and other packed spaces. Text must be created in silence.

      In a lot of collaborative scenarios, the keyboard gets in the way. If we are chatting about some foreign country and I want to check its GDP. Why would I stop the conversation and start typing?

      Find an input method, which requires less movement than my fingers on my keyboard.

      I agree with that.

      Writing via brain waves will still take some time.

      Right now, measuring brain waves is intrusive. Who wants to wear a silly hat to type?

      Maybe AI text prediction will advance in the next decade, so my phone not only correctly predicts the next word, but whole sentences and paragraphs after a few characters.

      That’s entirely credible but might not replace the keyboard.

  9. Phone screens are just too tiny — and always will be — to get any real work done.

    No dual monitors, either.

    I do hope most people try to be “productive” on their phone, because that means I can easily and utterly outcompete them by using a real computer and some decent 24″ or larger 4K monitors.

    Like shooting fish in a barrel that will be.

    1. Phone screens are just too tiny — and always will be — to get any real work done.

      Visual bandwidth is important but would argue that even a 24” monitor is insufficient. It is just a smallish square surface in the end.

Leave a Reply

Your email address will not be published. Required fields are marked *