What Brain Training Can and Can't Do

In January 2016, the Federal Trade Commission announced a $2 million settlement with the company behind Lumosity, then the most recognizable name in brain games. The charge was deceptive advertising. Lumosity had told customers its games could help them perform better at work and school and stave off age-related decline, and the FTC found the company simply did not have the science to back that up. "Lumosity preyed on consumers' fears about age-related cognitive decline," said Jessica Rich, then director of the FTC's Bureau of Consumer Protection.

That settlement is a strange thing for a brain-training platform to lead with. We're doing it anyway, because it marks the line every honest company in this category has to respect. The science of cognitive training is genuinely interesting. It is also genuinely limited. If you understand both halves, you can decide for yourself whether training is worth your time, which is the only basis we want you here on.

The best study we have

Start with ACTIVE, the Advanced Cognitive Training for Independent and Vital Elderly trial. Funded by the National Institutes of Health and launched in the late 1990s, it enrolled 2,832 adults between 65 and 94 and randomized them into four groups: memory-strategy training, reasoning training, speed-of-processing training, or a no-contact control. Training was modest by any athlete's standard. Ten sessions of about an hour, spread over five or six weeks, with brief booster sessions for some participants later on.

The results, published across a series of papers ending with a ten-year follow-up in 2014, are the clearest picture we have of what this kind of training does.

First: it works, on what it trains. Each group improved on its trained ability, and the improvements were durable. Ten years after those ten sessions, the reasoning and speed groups still outperformed controls on the trained tasks. For an intervention measured in hours, that persistence is remarkable. It is the same phenomenon a skier knows from technique work: a skill practiced deliberately, even briefly, does not simply evaporate.

Second: transfer is modest. Training memory strategies did not make people faster processors. Speed training did not improve reasoning. The abilities behaved like separate muscle groups. At the ten-year mark, all three training groups did report less difficulty with everyday tasks like managing medications and finances than controls did, which is encouraging, but that finding rests on self-report, and researchers treat it with appropriate care.

Third, and this is where discipline matters: a later analysis of ACTIVE data, published in 2017, reported that participants who received speed-of-processing training had lower rates of dementia over the following decade. That result got enormous press. It is also exactly the kind of finding that demands restraint. Dementia was not the trial's primary endpoint. The analysis was exploratory, the number of cases was small, and the result has not been replicated in a trial designed to test it. It is a genuinely interesting lead for science. It is not a fact about any product, including ours, and we will not sell it to you as one.

The scientists' argument

In 2014, the Stanford Center on Longevity and the Max Planck Institute for Human Development published a consensus statement signed by roughly seventy researchers. Its message was blunt: the brain-game industry's claims had run far ahead of the evidence. Practicing a game makes you better at that game, the signatories noted, but "there is no compelling scientific evidence" that this translates into the broad real-world benefits the ads implied.

A rebuttal followed within months, signed by a different large group of scientists who argued the first statement had underplayed legitimate findings, ACTIVE among them. Then in 2016, a team led by psychologist Daniel Simons published an exhaustive review of the industry's cited studies in Psychological Science in the Public Interest. Their summary has become the field's most quoted sentence, and it is worth reading slowly: there is extensive evidence that brain-training interventions improve performance on the trained tasks, less evidence that they improve closely related tasks, and little evidence that they improve distantly related tasks or everyday performance.

That is not a debunking. It is a map. Trained tasks: strong. Near transfer: some. Far transfer: unproven. Any brain-training product describing itself honestly has to locate its claims on that map.

So what is training actually for?

Here is where we land, stated as plainly as we can.

Training your memory, attention, and processing speed makes you measurably better at the skills you practice, and those gains can last for years. If you train recalling names, you get better at recalling names. If you train rapid visual attention, you get faster at rapid visual attention. Some of those trained skills are useful in themselves. Remembering names is not a laboratory abstraction; it is Tuesday.

What training has not been shown to do is change your underlying cognitive hardware or protect you from disease. We do not claim that our workouts prevent, delay, or reduce the risk of dementia, Alzheimer's disease, or cognitive decline, because no one has earned the right to that claim. If the evidence changes, through replicated, well-designed trials, we will report it here with the same caveats we would demand of anyone else.

There is also an honest case for the practice itself, apart from any test score. A daily session is a structured act of attention in a distracted age. Athletes understand training as a relationship with effort, not just a ledger of outcomes, and ten focused minutes on a hard mental task has a value you can feel without a neuroscientist's sign-off. We think that is worth something. We just won't dress it up as medicine.

How to read this whole category

When you evaluate any brain-training product, including this one, ask three questions. What tasks does it train, specifically? What evidence connects those tasks to anything you care about outside the app? And do its claims stay inside that evidence, or wander past it?

The Lumosity settlement happened because a company answered that last question badly, at scale, to people who deserved better. The lesson we take from it is not that brain training is a scam. The research says it isn't; trained skills improve and endure. The lesson is that the moment a company promises you protection from what you fear most, it has left the science behind. Read the fine print, here and everywhere. Then decide, the way you'd evaluate any training plan: by what it demonstrably does, not by what you wish it did.