[api]

TL;DR:

The number of cars on the road is growing much faster than trucks, and far more people drive them. Even most truck drivers also own cars and often drive their cars in preference to their trucks. Most ground-transportation-related investments revolve around cars and not trucks, and there are far more car-related products than truck-related products. Therefore trucks will soon be obsolete, and all cargo will be moved with cars.

He's not totally wrong. Mobile is the future. But desktop is also the future, and server, and IoT, and special purpose devices, and hacker devices like the Pi, and smart TVs, and smart cars, and smart appliances, and ...

... and and and and and ...

What's really happening is a great diversification of computing and a massive explosion in the amount of "silicon per capita." The part about "how many electric motors do you own?" is fairly spot on. But he falls down when he tries to push his argument too far and argue that mobile will eclipse desktop/laptop completely and we'll all be staring at 3.5" screens. I will say this until someone shows me a viable port of SolidWorks, Visual Studio, or Revit Architect that is usable on a smartphone. I'm not holding my breath for that.

The only way I see mobile replacing the laptop/PC is either some breakthrough in interaction metaphor (VR? AR that is really useful? brainwaves? who knows) or a mobile device that converts into a desktop when you plug a monitor into it and pair it with a keyboard. It would also have to be capable of being used with as much versatility as a PC, which would mean significant evolution of mobile OSes. Personally I doubt convergence... because why? Silicon is so cheap there is no economic driver for anyone to invest the development necessary for mobile to "swallow PC." It's cheaper and easier to just use two separate devices and let them deeply specialize and optimize for their respective roles. A car/truck convertible is possible too, but the result would probably be an oversized gas guzzler car with poor handling that converts into a shitty underpowered truck with poor capacity.

I also think he's wrong about mobile being the center of gravity. Mobile devices are dumb terminals for accessing the cloud. If there's an emerging new "sun" to replace the PC it's the cloud, not mobile. Increasingly I use both PCs with their big monitors and keyboards and mobile devices to access common cloud resources.

[roymurdock]

Couldn't agree more with all of your observations.

I agree that cloud is currently the center of the ecosystem, and will be over the short term. Data is the blood and the cloud is the heart that stores and pumps that blood. Will become increasingly important as silicon per capita ratio increases and people look to analyze data for optimization purposes.

As another poster commented, a paradigm shift in I/O (VR/AR?) will drive the next surge in productivity as new tech adopts to a better/faster way to get it to do what we want.

[api]

Tangential but:

As a cyberpunk hippie who thinks people should own their data and control their computers (personal comput-ing regardless of form factor), that's why I focus mostly on the transport and the cloud. The cloud is where your data mostly lives today and probably will live even more in the future, and having private transport envelopes to get that data to and from your multiple devices is key. In the long term I'm excited about personal cloud devices and homomorphic encryption, which would enable your own cloud resources to be hosted in commodity cloud data centers without compromising security. The next "PC" will really be a kind of amorphous blob of silicon orbiting cloud resources linked by an encrypted communication envelope.

You'll be able to access all that from a variety of different I/O devices: things with big monitors and keyboards, things with little screens and touch panels, refrigerators, cars, thermostats, etc.

[roymurdock]

Thing is, homomorphic encryption is expensive in terms of time and computing power. I can't see it being implemented for the majority of consumer grade data, which will probably continue to be mined by large corporations and governments in exchange for "free" hardware, software, apps, etc. Financial data, medical data, VIP data, crypto-junkie data - I think there's a strong use case there.

The guys developing enigma said they'd be putting out another paper at the end of the summer but I didn't see anything...any insights into what's going on over there?

[api]

I follow it a bit. Right now we know that it's theoretically possible but it's very slow, too slow for uses other than maybe very niche things.

In the next five years I fully expect to see an acceptably fast implementation that will broaden use to the application areas you mention: high security stuff and crypto junkies.

At that point is where things will get interesting. To make homomorphic crypto really fast will take custom silicon. As soon as I see acceptably fast software homomorphic, I expect to see either a startup or one or more projects inside big companies very quickly pop up with the goal of producing either homomorphic silicon to pair with an existing CPU or -- more interestingly -- a CPU that computes homomorphically as its native mode of operation. That could get interesting.

I predict that by 2030 you will be able to spin up a 100% "blind" sealed VM in the cloud for <$20/month that is fast enough to run what can run on an average cloud instance today.

All compute doesn't have to be homomorphic to get the benefits. You only really need it for encryption, decryption, key generation, authentication, and indexing of security critical data. Boilerplate stuff can be done in the clear. By 2030 standard cleartext compute will be so cheap I wonder if cloud providers will even bother charging for it below very large levels.

Edit: can't make this any deeper but to reply to your reply:

I'm not sure. I think that's up in the air. Right now you can do a whole lot with a small cloud instance with <1gb RAM and 20gb local storage. Software bloat seems to be increasing less rapidly than it was 10 years ago, and there are even movements toward more efficient languages today. So I just have to shrug on that one. We'll see.

[roymurdock]

Well the one thing I would push back on is...how useful will the computing power of the average cloud instance today be in 2030?