We are approaching critical mass on a number of technologies that we will see increasingly this year. This wave will start at CES, where we will see an impressive number of attempts at personal robots and AI-powered digital assistants. Most will fail. However, both the failures and the few successes will set the stage for the first true mobile personal robots that will arrive in the following years.
The 5G rollout will enable desktop cloud computing, and the Microsoft Virtual Desktop will become real for millions of us shortly thereafter. Ironically, many of us will access it with Chromebooks.
If you are located in a city center, you’ll increasingly see both electric car charging stations (largely thanks to VW) and autonomous cars. There has been an increase in violent activity toward the latter, as well as passive demonstrations against the former, which just seems weird.
We’ll have AI pretty much everyplace, but most of it will still suck. Deep learning applications will surprise us, though, with their advances in capability and knowledge. Efforts to make sure these systems aren’t corrupted will get more focus.
I’ll touch on all of that and then close with my product of the week: the only treadmill that you are likely to use for more than an emergency coat rack, the Peloton Tread.
I’ve been reviewing the announcements slated to go up next week, and I found a surprising number of robots and digital assistants (which I expect to be obsolesced by robots that are mobile).
Developers of the new digital assistants seem to have realized that we like to gesture, and talking to a tube therefore isn’t natural to us. Many entrants in the coming wave of digital assistants can emote, mostly by moving around something that looks kind of like a head or a big eye. This ability does make the device appear more human. As the back-end AI improves, the voices and responses sound more human as well.
The other vector is mobility, and an increasing number of these things are on wheels. The advantage is that the technology can follow you around without your having to buy one of these things for every room. The disadvantage is that wheeled appliances don’t deal with obstacles, especially stairs, very well.
Early price points suggest you could buy a traditional digital assistant for every room for less than what one of these mobile assistants would cost. Given you likely could create a cloud feature for this class of device, allowing the personality of your assistant to move virtually with you room to room, I’m questioning this initial focus on mobility. Most of these things don’t have arms — or if they do, they don’t have hands — that would allow this mobility to become a true advantage.
The 5G Revolution
What makes 5G different is that it addresses not only bandwidth shortcomings, with its largest impact at the network’s edge, but also latency issues, which should force improvements to the overall network.
There will be some initial adverse impacts in the market with Apple, of the big players, likely getting hit hardest. This is because informed buyers likely will want to wait until the 5G versions of smartphones are out before they refresh to avoid premature obsolescence, and those devices won’t show up in volume until after mid-year.
The reason Apple is likely to get hit hardest is that its battle with Qualcomm largely locks it out of this technology until its 2020 refresh. Given that churn already has been slowing for Apple (people just can’t justify replacing an iPhone that works for the few advantages of a new one now), this should reduce sales dramatically until the company has a solution.
The big impact, though, is that this will make concepts like the “Always-Connected PC” real, and allow us to get the benefits of Windows in the cloud. (It also should open the door for the anticipated MacBook to iPad Pro pivot Apple has been hinting is coming.)
This drive for computing in the cloud, something that Microsoft was restructured to anticipate, will change dramatically what sits on your lap and desk over the next five years.
AI has been around in some form or other since the 1950s, and much of it has sucked. The most recent trend to use deep learning promises to give us AI that is not only intelligent, but also adaptable to our needs and capable of advancing at computer speeds.
Granted, there is a lot of concern — well founded — that this eventually will make us obsolete, but companies like IBM have been working to mitigate that. IBM’s Watson may be evolving to become the best doctor in the world.
Given that much of this is cloud-based, the promise is that our connected devices will start evolving at a pace that is even faster than we have seen so far. Deep learning should transition our existing digital assistants automatically into something far more powerful.
Let’s just hope these systems remain on the “help humans” path, and that they don’t watch too many Terminator movies.
I’m still looking for a personal robot like my namesake, Robby the Robot, and this year we’ll take the next major step to getting there. By the end of the year we’ll be more connected, an increasing number of us will be riding in cars driven by robots, and all of this will be enhanced by the spread of deep learning as the birth of true AI progresses. I’m just hoping the new Terminator movie isn’t prophetic.
If you are at CES this week, best of luck getting around. This is one show that not only will showcase the future of transportation, but also could really use it. Getting around Las Vegas this week once again will be a never-ending nightmare.