I may write something later. Here is what I wrote 4 years ago what the world of 2025 might look like:
2025 is about 6 years away. If you count late 2025, it's closer to 7.
I wrote that before Google Home, and just after Alexa was released, back when it was barely functional. Since then, it has improved "exponentially". And remember, we now have Google Duplex, and Alibaba's version, which is even better (by a mile, according to reports).
Use of Home and Alexa, and also virtual assistants via voice has been increasing exponentially, particularly among young people. Most of my Assistant interactions are by voice.
Some of what I predicted has already come to pass, such as the news summary feature, which is now possible, though maybe not yet deployed (Salesforce had some research on this that made a splash a year or two ago; Facebook, too).
Machine reading has also greatly improved. It's not yet good enough to read documents to check complicated numerical logic; but it can now answer factoid -- and even more sophisticated -- questions with good accuracy. This is how Google Home is able to answer so many questions -- it can read Wikipedia and find answers, saying, "Here's what I found... [quotes Wikipedia or gives direct answer]."
Probably the tech is good enough now to translate domain-restricted documents to formal representations with good accuracy; and formal reasoning engines can check the logic. So, the machine-reading part of my prediction is also possible, given that restriction. But you would need somebody to actually write the domain-specific reading and reasoning system.
This is probably a dead-end, though. The future will be to use a unified reading-reasoning system with a giant neural net. They will probably either be built by infusing brain data, and/or pretraining on large amounts of text + fine-tuning of some kind. I don't see a future in external, hard-coded symbolic reasoning systems.
The self-driving cars will be here by 2025. Level 4, not level 5. And some small percent of the population will be using them -- next-gen TESLAs.
The construction robots will be here, too. They are starting to take over brick-laying, surveying, and other jobs. Many more will be done by robot by 2025. They're not general-purpose, though.
Digital driver's licenses are starting to appear. I don't know how many states use them, probably about 10 to 15 right now. In the past year or two many more have been looking into it. Maybe most states will allow it in 5 years. Many now allow you to use digital car insurance cards; and some don't even ask you for it, if you get pulled over -- they have it in their computers, already.
China is far ahead of the U.S. in terms of biometric identification. You really can buy stuff in some stores there, just using your face. Some European countries, like Sweden, are going cashless. And more and more restaurants and stores in the U.S. are going cashless, too. A few of the restaurants I eat at are cashless. Occasionally, I pull money to pay, and they remind me they are cashless.
We now have Apple's FaceID. The tech will get better and better, and harder and harder to hack or defeat.
We also have cashierless stores starting to appear, such as Amazon Go. Who could have predicted it would arrive so soon?
Farmbots are getting better, too. There are experimental farms managed entirely by robot. Maybe by 2025 we will see this expand. Maybe I was too optimistic there.
As to people moving out of cities and using VR, I was probably too optimistic. I was probably also too optimistic about AI prescribing medicine; but I think diagnosis will be pretty good by 2025. Babylon Health is working on it; and maybe Deepmind will make a big splash on this -- not really the kind of project they tend to focus on, but they have signalled an interest in helping advance automated diagnosis.
I was probably too pessimistic about tech unemployment, though it depends on whether there is an economic downturn around 2025, as these tend to cause high unemployment and rapid automation.
One thing I didn't predict at all was the coming BCI revolution. I only first became aware of what is coming late in 2015, early 2016.
I'll get back to this with some predictions about 2029.
The main things to keep in mind are: don't focus on a specific company, as companies are fickle; and be mindful of the difference between what's possible in theory, and what will actually be deployed.
Addendum: one thing to keep in mind is that when this future arrives, it won't feel futuristic. It will seem as though nothing has really changed. That's just how the mind works. "We were promised a glorious future. Sure, we've got supercomputer smartphones thay can talk to us and answer our questions, but that's not really very futuristic."
Some of the things might be here already, that people aren't even aware of. Digital driver's licenses is one example. How many know whether their states allow them?