Dr Jan Willem de Gee is giving a lecture in Barcelona for Deep Dive on Wednesday 13 January 2027. Deep Dive Lectures is now a reference lecture series in Barcelona: professors and experts give English-language talks in Barcelona bars. Someone who really knows their subject talks for about fifty minutes, and then everyone stays to discuss it over a drink. We are very happy to fly him into Barcelona for this incredible event, and a few tickets are still left, so come along.
De Gee is a neuroscientist and assistant professor in cognitive and systems neuroscience at the Swammerdam Institute for Life Sciences of the University of Amsterdam. His lab studies how the brain's neuromodulators, chemical messengers such as noradrenaline and dopamine, shape perception and decision-making. He works across species: in people he measures brain activity with fMRI and EEG and uses the size of the pupil as a window into how aroused the brain is, and in mice he records and controls brain cells directly. He has also worked in Tobias Donner's lab at the University Medical Center Hamburg-Eppendorf. In several of his studies, healthy volunteers took atomoxetine, a drug that raises noradrenaline and dopamine, or a placebo, without knowing which, and then made a long series of difficult visual decisions. His research is funded by a Veni grant from the Dutch Research Council.
His lecture, "Smart Drugs: What Can a Pill Really Do for Your Brain?", starts from a simple fact: millions of people take so-called smart drugs hoping for more focus, motivation and brainpower. But do they actually work? It is an honest, hype-free look at what the science really says. These five answers are a preview.
Five questions people ask about smart drugs, answered with his own published words. They come from peer-reviewed studies he wrote with colleagues, in English, lightly shortened and simplified for reading.
The noradrenaline systems of the brainstem project to large parts of the cerebral cortex, where the chemicals they release change how the target networks work. That puts them in an ideal position to shape brain-wide activity during decision-making in a coordinated way.
In our studies we used atomoxetine. It blocks the transporter that clears noradrenaline away after it is released, throughout the brain, and also slows the clearing of dopamine in specific regions such as the prefrontal cortex. At the dose we used, it raised heart rate and pupil size.
Pupil size is a measure of central arousal. Changes in pupil size that are not caused by light are causally related to the activity of control centres in the brainstem, including the locus coeruleus, the brain's main source of noradrenaline.
Sources: de Gee et al., PLoS Biology, 3 Sep 2025; Nuiten et al., eNeuro, Jul 2024. English originals, simplified.
It depends on the task. In one study, atomoxetine did not significantly affect objective task performance: it changed neither reaction time nor accuracy.
What it did change was bias. When people make decisions under uncertainty, they tend to systematically repeat, or alternate, their previous choices, even when what they see comes in a random order. These history biases were highly individual: some people leaned towards repeating, others towards alternating. Atomoxetine reduced these biases. In other words, raising these brain chemicals made people less likely to let their previous answers steer the next one.
In other studies, it did improve how well people could tell two images apart. But the improvement showed up in accuracy, not in speed, and only in some tasks.
Sources: de Gee et al., PLoS Biology, 3 Sep 2025; Beerendonk et al., PNAS, Jul 2025. English originals, simplified.
No. According to the Yerkes-Dodson law, task performance is best at moderate levels of arousal, and worse when arousal is very low or very high. Until recently, the evidence for this law in human decision-making was mixed. We showed that it holds for people across decision-making tasks and settings.
When we gave atomoxetine, it increased overall arousal, for example heart rate and blood pressure, but left the inverted U-shaped curve intact. Performance seems to adapt flexibly, so that people perform as well as they can in whatever general state of arousal they are in.
Sources: Beerendonk et al., PNAS, Jan 2024; Beerendonk et al., PNAS, Jul 2025. English originals, simplified.
Partly because people respond to drugs very differently. For example, the level of these chemicals someone starts with may change what a stimulant does to their performance, because performance is thought to be best at intermediate levels.
And it is notoriously difficult to work out where an individual starts on the Yerkes-Dodson curve. In one of our studies, the pattern suggested that participants were slightly under-aroused on placebo, which is common during long experiments, and closer to peak performance on atomoxetine.
Sources: Nuiten et al., eLife, 1 Dec 2023; Beerendonk et al., PNAS, Jul 2025. English originals, simplified.
Decisions usually come with a feeling of confidence. In our study, raising noradrenaline and dopamine improved people's perceptual decisions, but not their insight into those decisions. Their ability to tell when they were right or wrong did not improve. In fact, the drug weakened brain signals over the frontal cortex that are linked to that kind of self-evaluation.
Together with other research, this suggests that these chemicals tune the brain either towards getting the decision right or towards judging how good the decision was.
Sources: Nuiten et al., eNeuro, Jul 2024. English original, simplified.
Enhanced catecholaminergic neuromodulation thus improves perceptual but not metacognitive decision-making.
Doors 18:30 · Lecture 19:00 · Venue to be announced · In English