Before “adulting” became an internet joke, psychologist Jeffrey Jensen Arnett spent five years interviewing about 300 young people across the United States about what adulthood felt like. One description kept recurring. They felt “in between.” They were no longer adolescents. They were making more of their own decisions, taking responsibility for themselves and separating from the routines that had structured childhood. Yet many still did not feel completely adult. Arnett turned those interviews into the idea of “emerging adulthood,” which he formally proposed in a 2000 paper in American Psychologist . His original theory focused mainly on the late teens through the mid-20s and described a period marked by identity exploration, instability, self-focus, possibilities and the feeling of being somewhere between adolescence and full adulthood. The concept has limits. Arnett himself argued that emerging adulthood is not universal and is most likely to appear in cultures where young peo...
In January 2026, Anthropic published a finding that complicates almost every clean prediction about AI and jobs. Looking across its accumulated Claude usage data, the company found that 49% of occupations showed AI use in at least a quarter of their tasks. That sounds dramatic. Then the researchers adjusted for how often workers actually perform those tasks, how long the tasks take, and whether Claude succeeds at them. The picture changed. Some occupations, including teaching and software development, looked less affected after that adjustment than their raw task counts suggested. Others, such as data-entry work, looked more exposed. That difference matters more than another prediction about which professions will disappear. A job is not a title. It is a collection of things somebody has to notice, produce, check, explain, decide, coordinate and eventually take responsibility for. AI does not have to replace all of those things to change the job. It only has to make some of them m...
In a 2026 study of self-silencing, a university student given the pseudonym Joy described herself very differently depending on who she was with. She spoke openly with her parents. Their relationship felt stable to her, and she associated it with trust. A close friendship had taught her something else. The friend could become aggressive and sometimes hit her when she did not comply. Joy later told the researcher that she had “lost [her] confidence” and no longer knew how to handle conflict in close relationships. What caught my attention was not simply that Joy became quieter. She had not lost the ability to speak her mind everywhere. She still did it at home. Her silence belonged to particular relationships. Christin Grothaus found the same broad pattern among the 53 young adults she studied at international universities in Thailand. Participants often expressed themselves comfortably in one relationship and held back in another. The strongest dividing line was whether the relati...
Atomic Habits and Man’s Search for Meaning are both routinely filed under personal growth. That should make us suspicious of the category. James Clear is interested in why a person who intends to exercise keeps finding themselves on the sofa. Viktor Frankl was writing about meaning after imprisonment in Nazi concentration camps. One book is partly about making useful behavior easier to repeat. The other asks what remains worth living for when circumstances have stripped away much of a person’s control. Calling both of them “self-improvement” tells us almost nothing about which one somebody should read. The question I would use instead is more practical: what keeps going wrong? If the answer is behavior, there are books for that. If it is time, anxiety, self-criticism or bad judgment, the shelf changes. And if the problem is that all the goals are working but no longer seem particularly meaningful, a productivity book may make things worse by helping somebody become more efficie...
In a 2025 experiment, researchers recruited 1,035 adults and asked them to name something they were currently procrastinating on. The people in the intervention group spent a few minutes answering seven questions about that task. They described what they were avoiding, considered why completing it might be useful, picked an easier part to begin with, estimated how long that piece would take and chose a small reward for afterward. The whole study took participants about six and a half minutes on average. When asked how likely they were to complete the task within the next 24 hours, the intervention group averaged 5.38 on a nine-point scale. The control group averaged 4.88. That is hardly the kind of result from which to build a productivity religion. The researchers also measured what people intended to do rather than following them home to see whether the tax return, difficult email or unfinished report actually got done. I still like the experiment. It starts with the task someone is ...