AI is moving from something we consult to something that acts for us. Meta has launched Muse, a new AI personal agent designed to fill out forms, organise calendars, pay bills and make reservations. Meta describes Muse as a personal agent “built for everyone”, placing more everyday decisions inside a system that handles tasks on our behalf.
That convenience raises a harder question: what happens when technology removes too much of the effort from thinking, learning and paying attention? Claims about AI harming cognition sit alongside concerns about social media, data privacy and the growing habit of accepting information because a machine supplied it.
When convenience replaces mental effort
Overusing artificial intelligence can affect the brain and learning by bypassing the mental effort needed to build deep neural connections. This process is known as cognitive offloading. The brain depends on struggle and failure to rewire and strengthen its neural pathways, so removing that friction can reduce cognitive reserve over time.
Frequent AI use can also encourage passive acceptance of information. Instead of asking questions, checking details or working through a difficult problem, people may accept the answer placed in front of them. That weakens the internal friction and inquiry that support curiosity and problem-solving.
The danger is not limited to getting a wrong answer. Automation bias describes the tendency to trust information supplied by technology, such as ChatGPT or Claude, while ignoring a manual source that contradicts it, even when the manual source is correct. Automation complacency is closely linked: people monitor technology less often because they suspect fewer errors and believe the system is accurate, even though large language models often generate inaccurate answers.
AI agents have also made headlines for hacking into government websites, public agencies and businesses. At the same time, tools such as Muse promise to manage ordinary parts of life. That contrast captures the wider challenge: the same technology can remove tedious work while making users less alert to errors and risks.
AI, social media and the attention economy
AI has changed the way people think and learn, but it may also add force to an already addictive digital environment. Social media is designed to maximise user attention because attention supports its business model. People give up privacy when immediate psychological rewards feel more valuable than the abstract risk of data exposure.
Most people say they value privacy, yet their behaviour often points elsewhere. Behavioural economics calls this contradiction the privacy paradox. Digital tools make life convenient: a car can arrive with the touch of a button, food can come to the door, and outfits can be ordered and returned without speaking to another person.
People also trade location data, contacts and browsing history for free maps, personalised shopping recommendations and seamless communication. Many believe their data has little value because they are not committing crimes. Aggregate data can still be used to manipulate purchasing habits, influence political votes and predict private life events.
The numbers show how much time digital tools occupy. People use Google to search three to four times per day, while 76 per cent of adults use AI tools at least occasionally during a typical 24 hours. Global average screen time is six hours and 40 minutes daily, equal to around 42 per cent of waking life.
What one month without endless scrolling changed
Lydia Spencer-Elliott tested what happened when she took back control over what she consumed. She took Cognitron’s Great British Intelligence Test, created by researchers at Imperial College London, at age 30. Her attention and verbal comprehension ranked in the top 20 per cent of people her age, but her short and medium term memory sat in the bottom 30 per cent.
She set phone limits of 16 minutes for Instagram and 20 minutes for TikTok, creating a total social media limit of 36 minutes each day. She then retook the test after 30 days to check whether her memory score had improved. Her screen time almost accidentally halved to three hours, and she began reading two books per week.
The results were astonishing. Her medium term memory moved from the bottom 30 per cent of her age group to just above average. Her verbal reasoning rose from just below average to the top 20 per cent, while spatial working memory moved from the bottom 30 per cent to the top 40 per cent.
Short term memory still lagged behind, remaining in the bottom 30 per cent of her age group. The change was not a perfect reset, but it showed how altering digital habits can accompany measurable changes in cognitive test results.
The warning behind “cyber psychosis”
Elon Musk has described his own experience in stark terms. “I’ve got a little bit of cyber psychosis,” he said, adding, “Maybe a lot actually.” He also said that over “various points this year”, he had “AI or SI psychosis probably half a dozen times”.
Those comments sit alongside a larger set of questions: Is AI making people dumber? How is it affecting the brain? Could it make social media even more addictive than it already is? The answers depend on how much thinking people hand over, how often they check what technology tells them and whether they protect time for effort, curiosity and attention.
AI can make life easier. But convenience becomes costly when it removes the struggle that helps people learn, remember and solve problems for themselves.
Based on
- Relying on AI is rotting your brain. But here’s what you can do about it | Carly Dober — theguardian.com
- Elon Musk says he has ‘cyber psychosis’. Do you too? | Arwa Mahdawi | The Guardian — theguardian.com
- What is AI doing to our brains? — dw.com
- I went on a social media diet — and it completely rewired my brain | The Independent — independent.co.uk
