Partially correct. Technology companies fall into three main categories:
1.) Companies that let their customers do what they were already doing, but more efficiently, such that they need less manpower. Eg. self-checkout machines, manufacturing robotics, enterprise IT, backend logistics, data warehousing.
2.) Companies that let previously unskilled groups perform the jobs of skilled workers. Eg. EBay (disrupts auction-houses and collectibles), blogging (disrupts the publishing industry), AirBnB (disrupts hotels & hostels), Uber (disrupts taxicabs), StackOverflow (disrupts IT consultants).
3.) Companies that let people perform entirely new things that they couldn't do before. Eg. Google (researching topics that would've been prohibitively expensive to learn about), Kickstarter (funding creative projects), PayPal/Square (accepting payment as a small independent vendor).
The uses of a transformative technology tend to progress from 1 to 2 to 3, because that is the rough order of obviousness and risk to the entrepreneur. When a new technology comes out, all you need is familiarity with the technology and with existing processes to understand where a manual process could be automated outright. This is the domain of enterprise startups founded by people with deep knowledge of an industry and a keen interest in new technology.
Once those opportunities have been picked clean, entrepreneurs look for cases where they could use technology to let a disenfranchised group do the work of an enfranchised group. They take the people who have been rendered out of work in stage 1, and then give them the tools to put higher-skilled workers out of work in stage 2. Doing this requires a strong understanding of both the processes of the industry that will be cannibalized, and an understanding of the capabilities of the workers that will take their job, and an understanding of the technology that lets you bridge the former and the latter. So entrepreneurs here need to do a lot of synthesis and need a bit of "in the right place at the right time" as well.
Once those opportunities are picked clean, then entrepreneurs start looking for things that people want to do but currently can't because they lack the means. This is where the great consumer companies of the next age are formed. It's very risky business for the entrepreneur, though, because there is essentially zero data available on what things people would like to do if they could but can't currently. Usually entrepreneurs that succeed here do so via vision, creativity, and a good sense of what they themselves want to do.
We're currently just starting stage 2 in the Internet/mobile/cloud revolution, so I would expect more job destruction and more short-term social pain in the near term. But it stops when all the easy opportunities of stage 1 and stage 2 have been picked clean, and people are forced to hunt for consumer business models that employ all the people put out of work by the earlier stages. Stage 1 and 2 destroy jobs, but stage 3 creates them.
BTW, for the last major economic transition (the Gilded Age and Industrial Revolution), the stages looked like this:
1. Sailors => Steamboats. Weavers => power looms. Blacksmiths => Steelmaking. Coachmen => Railroads. Miners => bulldozers, excavators, and heavy equipment. Domestic servants => home appliances.
2. Railroads => Automobiles. Whalers => Oil. Coal miners => Oil. Messengers => telegraphs. Farmers => Meatpacking/agribusiness.
3. Electricity. Home appliances (part of this industry replaced domestic servants, but part let consumers do genuinely new things, like microwaves and blenders). Penicillin. Air conditioning. Airliners. Elevators. Refrigerators. Computers (kicking off the next revolution).
Note that there's a large overlap between the stages. The last stage 1 business, replacing domestic servants with vacuum cleaners and washing machines, didn't happen until after WW2, while the first stage 3 business, electricity, happened in the late 19th century. I think we've seen the same thing with the Internet age, where Google and PayPal were among the first Stage 3 businesses, but Stage 2 hadn't even happened yet.
The general pattern is based on who're the substitutes for each industry. Stage 1 substitutes machines for labor. Stage 2 substitutes labor for labor with the help of machines to do things more efficiently. Stage 3 uses machines (built by labor) to solve unfulfilled human desires.