Two very different things have been on many people's minds this autumn.

The first is the Facebook Files. Starting in mid-September, The Wall Street Journal published a series of articles based on internal Facebook documents, later revealed to have been shared by a former product manager, Frances Haugen. One slide from 2019, summarising the company's own research on Instagram, said: "We make body image issues worse for one in three teen girls." In October, Haugen testified before the US Senate. Later that month, Facebook renamed its parent company Meta and announced a large bet on the metaverse.

The second is a book. Oliver Burkeman's Four Thousand Weeks: Time Management for Mortals, published in August, starts from simple arithmetic: if you live to eighty, you get about four thousand weeks. Its argument is that most productivity advice is a way of avoiding that fact, and that a good life starts by accepting it.

These seem like separate stories, one about technology and regulation, the other about personal philosophy. They are really the same story. Our attention is limited, our time is limited, and a growing share of both is directed by systems designed to maximise engagement. This post looks at both sides: what the Facebook Files tell us about the ethics of engagement-optimised algorithms, and what philosophy has to say about spending a finite life well.

1. What the Facebook Files revealed

The Journal's series covered many topics. Three are most relevant here.

Topic What the documents suggested
Instagram and teenagers Internal research found that some teenage users, especially girls, linked Instagram to anxiety, depression and worse body image. "Teens blame Instagram for increases in the rate of anxiety and depression," one presentation said
Ranking for "meaningful social interactions" A 2018 change to the News Feed algorithm, meant to promote interactions between friends and family, also rewarded content that provoked strong reactions. Internal memos warned that it made outrage and divisive content more widespread
Known problems, slow fixes According to the documents, staff proposed changes to reduce some of these harms, and some were not adopted when they conflicted with engagement goals

Facebook disputes the framing. It has said the Journal mischaracterised its research, that many teenagers report Instagram makes them feel better about some issues, and that the existence of internal research shows it takes these problems seriously. Those points deserve a hearing. Internal research is often exploratory, and correlation is not causation.

But the central issue is not whether a particular slide was fair. It is a design question: what happens when a system that shapes the attention of billions of people is optimised mainly for engagement?

2. The problem with optimising for engagement

Recommender systems, the algorithms that rank feeds, suggest videos and decide what appears next, are among the most widely deployed forms of AI in the world. Most people interact with them many times a day without thinking of them as AI at all.

They are typically trained to predict and maximise signals such as clicks, likes, comments, shares and time spent. Those signals are easy to measure and closely linked to advertising revenue. The ethical problem is the gap between what is easy to measure and what people actually value.

What the system can measure What people may actually want
Time spent Time well spent
Clicks and reactions Information that is true and useful
Comments and shares Conversations that leave them better off
Return visits A sense that their life is going well

Strong emotions, especially anger and fear, reliably generate engagement. A system that learns from engagement will learn to show more of what triggers it, even if nobody designed it to. That is not a conspiracy; it is optimisation doing exactly what it was told.

Independent evidence points the same way. In October, Twitter published a study by its own researchers of algorithmic amplification in seven countries. In six of them, the home timeline algorithm amplified tweets from the mainstream political right more than those from the mainstream political left. The researchers said they did not yet know why. The striking part is not the political direction. It is that even the company running the system could not fully explain what it was amplifying.

3. The economics of attention is not new

In 1971, the economist and computer scientist Herbert Simon wrote that "a wealth of information creates a poverty of attention". What is new is the scale and precision with which attention can be captured. Machine learning allows each person's feed to be tuned to their individual reactions, continuously, in real time.

This is why many researchers argue that attention is an ethical issue, not just a matter of personal willpower. The philosopher James Williams, a former Google strategist, argues in his 2018 book Stand Out of Our Light that persuasive technology can undermine not only our moment-to-moment focus but our ability to pursue our own goals and reflect on what those goals should be. When the most sophisticated optimisation systems ever built are competing for your next minute, "just use your phone less" is not a complete answer.

4. Four thousand weeks: the philosophical side

Burkeman's book comes at the same problem from the other direction. Instead of asking what technology does to our attention, it asks what we should do with the little time we have.

His central claim is uncomfortable: you will never get everything done. There will always be more emails, more books, more opportunities and more obligations than time. Productivity techniques often make things worse, because becoming more efficient just means more things arrive to fill the space. The way out is not to optimise harder, but to accept limits and choose deliberately what to neglect.

This idea is very old. Around 49 AD, the Stoic philosopher Seneca wrote On the Shortness of Life:

"It is not that we have a short time to live, but that we waste a lot of it."

Seneca's targets were different, drunken banquets and political ambition rather than infinite scrolling, but the argument is the same. Most people do not lack time. They give it away, often to things they would not choose if they stopped to think. Seneca's advice was to treat time as the most valuable thing you own, because it is the only thing you can never get back.

Burkeman adds a modern point. Distraction is not only something done to us; it is also something we reach for. When work is hard, boring or anxious, the feed offers an escape. Understanding that impulse, and choosing to stay with the difficult thing, is part of the answer.

5. Practical steps for individuals

None of these are miracle cures. They are ways of taking back a little control.

  • Decide what deserves your attention, before the feed decides for you. Choose one or two things that matter most this week and protect time for them first.
  • Turn off non-essential notifications. Most apps send far more alerts than you need. Keep only messages from people.
  • Use chronological or "following" views where available. They give you more control over what you see.
  • Keep phones out of bedrooms and meals. Small physical barriers work better than willpower.
  • Accept that you will miss things. Most of what you skip will not matter. Burkeman calls this "the joy of missing out".
  • Notice what you reach for when something is hard. The urge to check your phone is often a signal that you are avoiding something worth doing.

6. What leaders and teams can do

Workplaces are attention systems too. Constant chat messages, endless meetings and "always on" expectations fragment attention just as social media does.

  • Limit priorities. A team with ten top priorities has none. Agree on a small number of things that matter most and say no to the rest.
  • Protect focus time. Block shared periods without meetings or chat.
  • Set response expectations. Make it clear that not every message needs an instant reply.
  • Measure outcomes, not activity. Being online is not the same as doing good work.
  • Respect limits. The Great Resignation shows that many people are reconsidering how much of their finite time work should take. US workers quit a record 4.4 million jobs in September alone. People are voting with their weeks.

7. What product builders can do

If you build products that compete for attention, you are part of this, and you have choices.

  • Choose better metrics. Measure whether users accomplish what they came to do, and whether they report being satisfied, not only how long they stay.
  • Test for harm, not just for growth. When an experiment increases engagement, ask what kind of engagement it increased and who might be hurt.
  • Give users real control. Let people choose how their feed is ranked, set limits and opt out of recommendations.
  • Be careful with young users. Teenagers are more vulnerable to social comparison. Design defaults for them with that in mind.
  • Publish and listen to your research. If your own studies find harm, the response should be a fix, not a reputational strategy.
  • Make ranking explainable. If you cannot explain what your system amplifies and why, you are not in control of it.

Why this matters

  • Attention is finite and valuable. What we pay attention to shapes what we know, how we feel and who we become.
  • Engagement is not wellbeing. Systems that optimise for engagement can work against the people they claim to serve.
  • Recommender systems are AI ethics. The most influential AI systems in daily life are not chatbots or robots; they are the algorithms ranking our feeds.
  • Regulators are paying attention. After this year's hearings, pressure for transparency and accountability for algorithms is likely to grow in the US and Europe.
  • Philosophy has practical value. Ancient ideas about finite time are a useful guide to modern problems of distraction.

Final Thoughts

It is easy to see this year's revelations as a story about one company. It is better seen as a story about a design choice made across much of the technology industry: to treat human attention as a resource to be maximised. Changing that will take regulation, better metrics and companies willing to accept slower growth in exchange for healthier products.

But there is also a personal question, and it does not depend on what any company does. Each of us has a limited number of weeks. The most important decision is not which productivity system to use, or even which apps to delete. It is what we choose to care about, and whether we give our attention to that or let something else choose for us. As the year ends, that seems like the right question to take into the next one.

Sources