Ewelina Kurtys quit scrolling. Working with AI brought her back to a notebook
Ewelina Kurtys deleted the Instagram app. She kept TikTok to publish her own videos, but stopped watching other people's content. A neuroscientist making that choice might be expected to explain it through an account of an overloaded brain. In her conversation with Krzysztof Kosman on EdTech Dots, she offers a more careful explanation: she decided she had better uses for her time.
That caution surprises the host. A brain researcher seems like precisely the person who should settle the question of what phones are doing to us. Kurtys refuses to turn a personal decision into a scientific verdict. She does not need neuroscience to justify stopping scrolling.
Later, she describes another choice. She uses AI extensively enough that, during complicated tasks, she needs a notebook to record what she is trying to do. One instruction produces an answer that opens up several new possibilities. It becomes easy to forget where she started.
These two experiences shape the conversation about attention and learning. Kurtys limits a stream of content that consumes her time while using tools that require her to make more decisions. In both cases, she asks what the technology is for and what she actually gets from using it.
Why she stopped watching Reels
Kurtys does not describe Instagram as devoid of value. She found inspiration and useful information there. She also recognised that its recommendations could adapt to her interests. The problem emerged when she compared the benefit with the time spent watching.
She describes spending twenty minutes watching Reels before finding something particularly useful. The other videos had taken her attention too, even when she had little reason to see them. Eventually, that way of finding information stopped making sense to her.
Deleting an app did not require answers to every question about the brain. It required an answer about her own day: was this how she wanted to spend it? For her, that meant cutting back on consumption while retaining the ability to publish.
This helps explain the apparent contradiction in her position. Someone can decide that scrolling is a poor use of an afternoon while remaining cautious about a claim that it permanently damages the brain. Those judgments require different kinds of justification.
She also leaves room to change her mind. She does not present removing apps as a rule everyone must follow. She describes a choice that fits her present objectives and experience.
When scientific language promises too much
Early in the episode, Kosman asks what scientists still fundamentally do not understand about the brain. Kurtys replies, with disarming frankness, that it is pretty much everything. She then qualifies the answer: researchers understand pieces of the mechanisms and can use some of that knowledge, but a complete account remains out of reach.
She distinguishes work on cells, animal research and studies involving people. As research moves closer to everyday human life, controlling the conditions and assigning an observed effect to a single cause becomes harder. A laboratory can reduce the number of variables. People's lives contain overlapping behaviours and circumstances that researchers do not determine.
This is why she keeps returning to correlation and causation. Two things occurring together do not, by themselves, explain why they occur together. She is not using that distinction to dismiss research on wellbeing. She acknowledges studies identifying problems associated with social media use. Her objection concerns the confidence with which those problems are sometimes explained through a particular account of brain chemistry.
She also separates different timescales. Distraction while working with notifications switched on is a different question from the ability to concentrate weeks later. She is reluctant to treat those situations as evidence of precisely the same effect.
She has a similar objection to the phrase “dopamine detox”, which she once used herself. After reading the literature, she became sceptical of that language. In her view, taking a break from a phone can be described as an attempt to feel calmer or improve wellbeing without attaching a promise of a biochemical reset.
She still limits her use of apps, although her reasons for defending that choice have changed. The practice remains; her certainty about the mechanism has diminished.
Access to knowledge creates a problem of choice
When the conversation turns to education, Kurtys is enthusiastic about opportunities she did not have as a student. She recalls learning languages without easy contact with native speakers. Speaking to them was a scarce opportunity, associated with options such as expensive language schools.
She contrasts that experience with online lectures and materials from respected universities. She wishes she had enjoyed that access earlier. Yet she also describes the difficulty that comes with abundant choice: deciding what to learn and how to spend a limited amount of time.
She encountered that problem when she began studying AI. Her background was in neuroscience, rather than engineering. She took introductory coding courses and paid for courses on Coursera to understand how tools in an unfamiliar field worked.
She could have searched for free resources. What she valued in the courses was their structure. In her experience, having access to information does not remove the need for a sequence, a plan and guidance through a subject.
Asked how to judge whether technology supports learning, she suggests looking at results. A student may be able to relate those results to an exam. In other circumstances, the question is whether an activity helps with work or improves wellbeing. Finding material engaging does not, on its own, establish why it deserves a place in the day.
One instruction becomes five more tasks
A question about AI removing cognitive effort leads to the episode's most concrete description of work. Kurtys recalls the early stages of building her own website with AI. She would request a task, receive a result and move to the next one. Eventually, she could struggle to recover her original intention.
One task branched into several more. The tool made progress possible, but it also produced additional threads that needed attention. A notebook became necessary to keep track of the goal and maintain order.
She compares this work to management. Executing a task yourself can be relatively linear. Coordinating other people or AI requires following several issues at once. As more tasks get done, more things also need supervision.
She does not present this as proof that every user's cognitive abilities improve. She describes her own experience: complicated work with AI makes her feel more challenged and gives her more to keep in mind. She does not feel that she has simply stopped thinking.
There is an important qualification within that account. Automation can amplify bad decisions as well as good ones. Completing an instruction faster does not establish that it was the right instruction to give. She therefore emphasises evaluating the work step by step.
When Kosman suggests that people may be reaching their limits, Kurtys narrows the claim. She is describing her skills at a particular moment and the experience of a first substantial project. She expects to improve with practice. That is a belief about learning from experience, rather than a measurement of the brain's limits.
The tutor can keep explaining. The learner gets tired
Kurtys also values AI as a patient partner for learning. A person can keep asking questions and request another explanation. The pace is easier to adjust than it is in a lecture.
As a student, she sometimes skipped lectures that were not compulsory and used a friend's notes. A lecturer could move too quickly for her or too slowly. Either way, she found herself distracted. Adjusting the pace responds to a difficulty she remembers from her own education.
The tool's patience has another side, however. During work, a person still has to absorb an answer, assess it and decide what happens next. Kurtys says she already limits the pace herself: AI could continue working while she is getting tired.
Kosman also raises responsibility. A teacher or university has an identifiable perspective and an established structure. How should a learner assess a system that gets to know them increasingly well and could influence their thinking?
Kurtys treats manipulation as a possibility, rather than an established account of what every model is doing. She does not offer a complete solution. She does say that, if she were twenty again, she would still go to university. She continues to learn from people and wants to maintain that contact.
She uses AI for technical questions and to organise her thinking. She does not give it responsibility for deciding what she should do with her life. The discussion of new tools ends with a concrete picture of her work: more answers arriving on a screen, a goal recorded in a notebook, and a person who needs time to assess the next step.
