In traditional data processing, information is formulated in terms of sequences of zeros and ones – bits. In a quantum context, we can allow for bits which are a mixture of both – both 0 and 1 simultaneously.
More importantly: When several quantum bits are put together, we are able to construct mixes of all possible sequences of zeros and ones. Consequently, the information content of quantum information does not increase just a little as you add one more bit. In fact, it doubles with every quantum bit!
In a quantum computer, such mixtures of sequences may be processed in one single computation. In such a process, some bit sequences may be reinforced while others may be diminished – or even fully cancelled. Exploiting these possibilities for constructing efficient algorithms is far from trivial. But it may be done. Many examples are known, and devising new ones is ongoing work.
Improving the quality of quantum computers is also ongoing work. As larger and less noisy quantum computers are being built, this may actually change the world – for everyone, not just those with a particular interest.
What is quantum computing not
A quantum computer is not a computer which is faster per se. And quantum algorithms are not regular algorithms with a magic quantum speedup.
Quantum computing is different, more general. It gives us more leeway when it comes to processing data. It may be exploited to find more efficient solution to specific problems.
You will not replace your home computer with a quantum computer any time soon, probably never. Most of the specific problems for which there is a quantum advantage belongs to the realm of high performance computing.
Not even in this realm will quantum computers take over completely. When quantum computers run quantum algorithms, they will do so in tandem with traditional, classical computers.