I think car ownership might change to subscription service model vs. outright buying a car, but still mostly private vehicle "ownership" in 2030 outside of dense urban areas.

Maybe the avg. car ownership per household will drop dramatically, from ~2 to 1 or less, with the rise of ridesharing/public transit.

> might change to subscription service model vs. outright buying a car

This is what I mean. Car ownership will be a hold-over from folks who still own a car from a long while ago.

Most people will just subscribe to a subscription rideshare service.

I'm skeptical about this. This past year we've seen two major car sharing services take a huge hit. ReachNow was suddenly shut down and Car2Go pulled out of the US. If you end up being right, the car sharing companies will have to go back to the drawing board to create a new business model.

> the car sharing companies will have to go back to the drawing board to create a new business model.

I'll bet on this. Either self-driving cars dramatically lowers the cost of operations. Or startups like Nomad Rides change the business model entirely. Micro-mobility via scooters / bikes might also be a valid reason here.

But the consumer desire for ride-share is there and it's unlikely to go anywhere.

> Micro-mobility via scooters / bikes might also be a valid reason here.

I think this is the winner for dense cities with mild winters.

During rush hour in many cities you're about as fast walking as you are driving. People barely walk though and cycling requires physical effort which makes it inconvenient enough for many people to prefer cars. Scooters fix these problems and get you there really fast, except they suck in winter, except again in a few years winters may not be that bad anymore. The only (practical, non-luxury) reason you might want a car then is transporting a lot of stuff. For those few situations, carsharing might be an acceptable solution.

It's not far from that now, at least in the UK I think 80% of people get their cars on some kind of personal contract plan.

> very few desktop applications will exist

Probably true for Windows and Mac. Meanwhile, Linux and BSD people will still be arguing GTK vs. Qt.

> Meanwhile, Linux and BSD people will still be arguing GTK vs. Qt.

100%. I doubt *nix apps will change much. But you DAWs, photoshops, and microsoft words will no longer exist on the desktop.

Re: DAWs .... rendering a physically modelled synthetic orchestra, let's say, 50 players, would probably tax even my current Threadripper system. It has a cooling system the size of ... let's be polite and say two very large coffee mugs.

Remind me again how you're going to do this on something other than "the desktop"?

Oh, and don't think about offloading it to the cloud. It's going to be on the order of 100GB of data when finished.

Oh, and did I mention that in 2020, I'm going to be conducting this thing in realtime (or some ML-based agent will be) ...

My guess is either one of two scenarios, possibly both:

1. WASM develops into a security model that's much closer to native hardware, more like a container that's shipped with a browser. It's probably more accurate to say that your desktop might still be important in this model, but from a user perspective, the app will "live" inside your browser.

2. Apps become more like consoles that stream the output of what you're running rather than the entirety of your data. The computing will happen on cloud-compute, but you could interact with the final product in your browser. Similar to Stadia.

> very few desktop applications will exist, they will have moved to web assembly powered browser applications

Oh, I hope you're wrong about this one. That would render computers largely worthless to me due to a lack of software.

He's off on this one. There's still stuff life video editors or video games or 3D apps that will always make more sense to run locally.

Just the trivial apps will be moving to the cloud.

> That would render computers largely worthless to me due to a lack of software.

Curious, why?

Because PWAs are not acceptable for my use. They involve far too much risk and exposure, and they almost always provide an inferior experience.

What, no 'true AI' (yet)?

Assuming you mean GAI, I think very convincing, powerful AI already exists inside organizations with the resources to build it (google/microsoft), and they're trying to figure out how to commercialize it without doing anything risky. Models have already been publicly demonstrated which can leave some with eerie impressions until you understand how it works, and even if you do, the results are continuing to gain complexity and intricacy.

If Google, Microsoft, Amazon or Apple have GAI, how do you explain the fact that Siri, Cortana, Google assistant and Alexa are all so very, very, very far from that?