Automated learning algorithms (ML) allow computers to define and apply rules that are not expressly explained by the developer.
There are a lot of articles devoted to machine learning algorithms. Below is an attempt to provide a "helicopter view" description of how these algorithms are implemented in different business areas. This list is of course not a comprehensive list.
The first point is that ML algorithms can help people by helping them find patterns or dependencies, which cannot be seen by humans.
The digital prediction appears to be the most popular area here. For a long time computers have been used actively to predict the behavior of financial markets. Most models were developed before the 1980s, when financial markets managed to achieve sufficient computing power. These technologies later spread to other industries. Since computing power is now cheap, it can even be used by even small businesses in all kinds of forecasts, such as traffic (people, cars, users), sales forecast and more.
Anomaly detection algorithms help people check a lot of data and identify which cases should be examined as anomalies. In the area of finance, they can identify fraudulent transactions. In infrastructure monitoring, they allow identification of problems before affecting businesses. It is used in QC manufacturing.
The main idea here is that you should not describe each type of anomaly. You can give a large list of different known cases (learning group) to the system and the system used to identify anomalies.
Aggregate algorithms allow the collection of a large amount of data using a wide range of targeted criteria. A man cannot efficiently work with more than a few hundred objects with many parameters. The machine can perform more efficient assembly, for example, for qualified customers / clients, product lists segmentation, classification of customer support cases, etc.
Prediction recommendations / preferences / behavior algorithms give us the opportunity to be more effective in interacting with customers or users by providing exactly what they need, even if they have never thought about it before. Recommendation systems are already working poorly in most services now, but this sector will be improved quickly in the very near future.
The second point is that machine learning algorithms can replace people. The system analyzes people's behavior, builds rules based on this information (i.e. learn from people) and applies those rules that work in place of people.
First of all, this concerns all kinds of standard decision making. There are a lot of activities that require standard procedures in standard situations. People make some "normative decisions" and escalate into non-standard situations. There are no reasons why machines can't do this: document processing, cold calls, bookkeeping, first-line customer support etc.
Again, the main advantage here is that ML does not require a clear definition of rules. It "learns" from situations, which have already been solved by people during their work, and makes the learning process cheaper. Such systems will save a lot of money for business owners, but many people will lose their jobs.
Another productive area is all sorts of data collection / web scraping. Google knows a lot. But when you need to get some structured information gathered from the web, you still need to get someone to do it (and there's a great chance that the result isn't really good). Based on your preferences and requirements, information and cross-authentication architecture will be automatically collected thanks to ML. Qualitative analysis of information will continue to be done by people.
Finally, all of these methods can be used in almost any industry. We must take this into consideration when we anticipate the future of some markets and our society in general.