In a follow up to the last decade in review this thread is looking at some of the disruptive technology just around the corner. View: https://www.youtube.com/watch?v=y916mxoio0E This is a good introduction into how to predict the future. The TL;DR is look at the rate of change in an industry. https://medium.com/slalom-technology/the-2-5-on-understanding-innovation-691a0c466e26 Over the last century the time it takes for a disruptive technology to go from less than 10% of the market to over 90% has been coming down. Cars took 40 years or more to replace the horse. Electric vehicles could replace all combustion vehicles within a decade. How could it possibly do that? Well here is my prediction... By 2022 electric vehicles become more or less the same price as the ICE alternative with the same specs... Or worse for the ICE version. By 2025 there is almost no demand for new ICE vehicles. By 2027 we will start noticing there are less petrol pumps around as less people use them, range anxiety starts to switch to filling up you liquid fuel tank. At the end of the decade you will be able to get a cheap electric vehicle for about the same price as a second hand ICE vehicle would cost now. Alongside this, I do think we will get self driving cars. I think most likely Google doing what they did with Android and Tesla probably being the Apple do it yourself alternative. Probably with whichever of the old guard get it right there will be some 'used for a brand' systems like WebOS is to TVs. Also very likely there will be an open source alternative either in development or working in the wild. Tesla is currently claiming they have an 8.8x better safety record vs humans and 3 billion miles driven. Waymo just broke their 20 million mile and said "It took us a decade to drive the first 10 million miles but just over one year to complete these last 10 million". The amount of money, data and experience in the field is increasing exponentially. By 2025 we could well have had 100 billion miles driven around the world on some sort of self driving platform. Encountered most of those one in a million scenarios and worked out how to do the right thing there. Probably well on the way to working out the one in a billion really rare things too. I think it will probably take most of the decade to work out the laws, solve the inevitable problems and get the reliability right. I see it becoming like the aeroplane industry. In the early days there were a lot of aeroplane companies. You had the de Havilland Aircraft Company Limited who designed the world's first commercial jet airliner the DH 106 Comet in 1949. By 1964 the company was gone. One serous problem can shatter people's trust in you, cost a lot to fix and sink you in an instance. Think Nokia when they tried to quickly build an OS to rival iOS and Android. Uber learnt the hard way that it's not something to half do right. You have to really nail it from the off and not ask more of it than it can cope with. Google and Tesla are working from opposite ends. Tesla is going for the clock up a lot of miles on the easier roads with very little unusual obstacles. Whereas Google is going more for a really get to know the area and work out how to get around it safely. I think the ultimate self driving platform will have a very detailed map of the roads all over the world, downloading the data for your route / local area on the fly. It will have a series of sensors that allows it to read the environment. I think the humble video camera will be king and the rest will be to backup these cameras. The main trick of the OS / software will be to map the video feed onto the map, work out where the movable obstacles are and determine the safe path. Like with the 747 MAX problem, the real trick is reliably handling sensor failure or false information. This is where that map comes in handy. In the event of a problem or confusion, the vehicle should be able to predict how to safely slow down and stop. Probably sending out a car to car alert saying what it's intentions are. Once you can safely do all that with very little lag then you are likely to only be a small step away from building a system that is 100x better than a human. So back to my prediction of almost no ICE cars in 10 years. I think in 10 years time you will be able to buy an electric car cheaper than even a second hand ICE car. Self driving cars will be real, there will be the Uber type services vastly undercutting any human operated vehicle. These two will combine so almost anyone could roll out a self driving taxi service with very little money. The reduction in demand for petrol will make it more expensive and harder to find and about 80%-90% of all ICE vehicles could have been replaced with an electric drive chain. If not globally at least in the developed world but I think the model will more likely follow the smart phones where electrification and self driving is going to open up travelling around like the mobiles did with communications.