In the 1940s Thomas Watson, then the CEO of IBM, infamously forecasted that the world wouldn’t need more than "about 5" computers. Thankfully for IBM and Youtube-famous vloggers, Watson was a little more than wrong.

To be fair, at that point computers cost millions of dollars and took up an entire room to do anything a graphing calculator today can do. It makes sense that he'd think the cost wouldn’t be justifiable for most companies—much less families or individual people.

Back when people started thinking about computers, all they wanted was something to automate number-crunching. By the late 1800s, the US population had grown so large that it took the U.S. government more than seven years to manually summarize the results of each census.

By the time they were done, they were three years away from the next census. There was more time spent processing the census than not, which is a bit much if you ask us. We needed some way to process data faster.

The first computers were built for just that: number crunching. These calculators were run on vacuum tubes. When electrons flowed in one direction in a tube, that tube represented a binary 1. When the electrons flowed in the opposite direction, that same tube became a 0.
(Source)

Vacuum tubes had many problems, though. They were large and heavy, burned out often, and created a lot of heat. If we had continued to use vacuum tubes, Watson's prediction probably would've come true. They were a mess.

Even if they created some sweet music.

Despite the creepy music you can get from vacuum tubes, the first commercial computer produced was the UNIVAC 1 (the UNIVersal Automatic Computer). The first UNIVAC computer to be produced was the one bought by the US Census bureau in 1951.

It wasn't until IBM's 650 and 701 mainframe computers in 1953 that computers were made widely available for the first time to simplify number-crunching in businesses. They were used for things like maintaining employee information and managing records. They were still so expensive, though, that sending personal messages or holding scores for the bowling league would've cost a company hundreds of thousands of dollars.

Yikes.

Then came the transistor. Tiny transistors replaced those bulky vacuum tubes, making computers much easier to produce commercially. These parts made computer memory start its shrink from room-sized mainframe computer to teeny-tiny smart phone. Computers with transistors could not only process numbers, they could also store instructions (read: programs), making them much more powerful than before.

The next big jump came when computers went from transistor to integrated circuit, which combined the CPU, memory, and input and output all into one small chip. These things were also much more adaptable than the circuits we'd been using. Before, each circuit was custom-built to handle a unique computer. Now the same chip could be used in any number of different devices. That integrated circuit was the final step towards a personal computer that could actually fit

• in someone's house.
• in a person's budget. (Scrooge McDuck could've replaced his big pile o' money any time he wanted, but we're talking about a middle class American family here.)

The first computers—mainframes—were the size of large rooms, mainly to house and cool large pieces of hardware. Today, smart phones small enough to fit in your pocket have more computing capability than NASA needed in order to send Apollo 11 to the moon. And it's all thanks to transistors and integrated circuits.
(Source)