An expert on sleep and problem-solving responds to Hal Y. Zhang’s “Dream Soft, Dream Big.”
In 1921, Nobel Prize winner Otto Loewi had a problem. For 17 years he had suspected that nerve cells communicated with one another through chemical transmission rather than electrical waves. But he did not have a way to prove his theory. Then, one night, he awoke with an idea for an experiment and quickly wrote it down. Unfortunately, as happens to many of us, in the morning he had forgotten the idea, and worse, his nocturnal notes were completely illegible. Fortunately, the next night he again awoke with the idea, and instead of trusting his notes and memory, he immediately went to the lab and performed the experiment. Not only did the experiment support his theory, it sparked his research program on neurotransmitters that won him the Nobel Prize.
Anecdotes like Loewi’s suggest a tantalizing possibility—that we may be able to strategically use sleep to solve our, and the world’s, problems.
Hal Y. Zhang’s story “Dream Soft, Dream Big” imagines a near-future where this is possible. In the story, Katia and other aspiring Dreamers study cryptic pictures before they sleep and report their dreams the following morning. The pictures convey an intractable problem, such as factoring large numbers or predicting how a protein will fold. Being a profitable Dreamer means having dreams that solve the problems. But to have these Utility Dreams, you first have to dream about the topic—and this turns out to be more difficult that you might think.
Some information and experiences are more likely to be processed during sleep than others. For example, stressful events are more likely to be incorporated into dreams. Unsurprisingly, people are currently reporting a large amount of COVID-related dreams involving disease, wearing masks, and staying home. We also dream more about things that are important to us and that occur close in time to when we fall asleep. For example, athletes are more likely than psychology students to dream about sports. But while our dreams often incorporate recent experiences, they very rarely replay events exactly as they occurred.
Beyond dreams, other parts of sleep may be equally or more involved in overnight processing. People remember information better if they study it at night before bed and try to recall it the following morning than if they study in the morning and recall it later the same day. This sleep benefit for memory is primarily associated with deep sleep, when fewer dreams occur. Just as in dream incorporation, important information is most likely to benefit from sleep. Like Katia, we can influence what is processed during sleep by focusing on its importance, incorporating it into our daily lives, and thinking about it immediately before sleep. However, this process is not foolproof, as anyone who has ever received a C after cramming the night before an exam knows.
To understand how sleep enhances our memory and problem-solving, scientists have developed a technique to increase the likelihood that certain information is processed. This technique is called targeted memory reactivation because it selectively targets a specific awake experience and then attempts to prompt the reprocessing (reactivation) of the memory of that experience during sleep. Reactivation occurs naturally during sleep—targeted memory reactivation just influences what gets priority. Here’s how it works. Before sleep, you pair information with a sound or smell. For example, you might learn the location of two objects, a dog and a bell, on a grid. When you learn the location of the dog, the sound of a dog barking plays, and when you learn the location of the bell, a bell ringing. Then, during sleep, the dog’s bark is presented without waking you up, but not the bell’s ring. In the morning, when you try to recall the objects’ locations, you are more likely to remember the dog’s location than the bell’s.
Targeted memory reactivation may sound a little like now-debunked mid-1900s experiments that purported to teach people a new language during sleep. But it’s something else entirely. The sounds in targeted memory reactivation studies elicit processing of what you already know. They do not implant new information in your head. This may seem at odds with problem-solving, since a solution is the creation of something new. However, when reflecting on his nocturnal discovery, Loewi credited sleep with making the association between his theory from 1903 and an experimental technique he had recently used to answer a different question. In other words, he already had all the information—sleep helped him put it together.
In a recent study, my collaborators and I asked whether targeted memory reactivation could improve problem-solving. In evening sessions, participants attempted to solve brain teasers, each paired with a different music clip. Then, we presented some of the music clips while participants slept. In the morning, participants reattempted the same brain teasers they failed to solve the prior night. We were excited to find that participants solved more of the brain teasers that were cued overnight. Interestingly, unlike for Katia, the solutions did not come to them in a dream. And unlike Loewi, they did not awaken with the solution in the middle of the night. Instead, participants solved the brain teasers when they actively worked on them again.
There are many directions this research is headed. We’re making progress on new technology to monitor sleep at home paired with apps that present targeted memory reactivation cues. This will allow large-scale studies at home, using natural sleeping conditions and processing real-world problems. In addition, like Katia reviewing the problem before sleep, the transitional state from awake to asleep may be important for prioritizing sleep processing. The Media Lab at MIT is developing a device to specifically target this hypnogogic state. And, if dreams are useful for problem-solving, an open question is whether attempting to solve a problem while consciously aware of being in a dream shows the same benefit.
Katia ultimately abandons her Dreamer career and the company folds largely because the Dreamers do not successfully produce solutions. Maybe one of the challenges was convincing the sleeping brain to process the problems. Our ability to guide what is processed during sleep is still nascent, and fully controlling this prioritization may be ill-advised. But, in the meantime, try thinking about a problem before sleep and taking a moment to reflect on it the next morning. While people often compare sleep to death, our sleeping brain is not truly offline. Your next Aha! moment may be one sleep away, even if it is 17 years in the making.