B2 reading: Who chooses what you watch?

1 Open a streaming service or a social media app and you are immediately shown a selection of films, videos or posts that have been chosen for you. This selection is not random. It is produced by recommendation algorithms: computer programs that analyse what you have watched, liked, skipped and searched for, and compare it with the behaviour of millions of other users in order to predict what you will want next.

2 There is no doubt that these systems can be useful. Faced with thousands of hours of content, few of us want to search through everything ourselves. A good recommendation can introduce you to a documentary or a band you would never have found on your own. For the companies, the benefits are even clearer: the longer people stay on a platform, the more advertising they see, or the more likely they are to keep paying for their subscription.

3 That commercial goal, however, is exactly what concerns some researchers. An algorithm designed to maximise the time users spend on a platform will tend to promote whatever keeps them watching, and that is not always what is good for them or for society. Content that provokes strong emotions, particularly anger or outrage, often attracts more clicks and comments than calm, balanced material. If the system learns this, it may gradually push users towards more extreme videos and opinions.

4 A related worry is the so-called "filter bubble". If you are mostly shown content similar to what you have already enjoyed, you may rarely meet ideas that challenge your views. Over time, critics argue, this could make people less tolerant of those who think differently. The evidence for filter bubbles is in fact mixed; several studies have found that many people still encounter a wide range of views online, partly because they use several different sources. But the concern has been strong enough to influence public debate and, in some places, new laws.

5 The companies themselves stress that their systems are constantly being adjusted and that users have more control than they realise. Most platforms allow you to remove items from your viewing history, mark content as "not interested", or turn off personalised recommendations altogether. Few people use these options, however, either because they don't know they exist or because the default experience is simply easier.

6 Some experts argue that the answer lies not only in regulation but in changing our own habits. Deliberately searching for something new, following people you disagree with, or choosing what to watch before opening the app are small steps that put the user, rather than the algorithm, back in charge. The technology may be complex, but the question it raises is simple: who is deciding what you see?