Giving robots self-awareness as they move through space - and maybe even providing them with gene-like methods for storing rules of behavior - could be important steps toward creating more intelligent machines.
Now roboticists are going a step further, training their creations to do even better: understand their own image in space and interact with humans like humans do. Today, there are already robot-teachers like KeeKo, robot-pets like Moffin, robot-babysitters like iPal, and robotic companions for the elderly like Pepper.
But even these reasonably intelligent creations still have huge limitations, some scientists think. “There are niche applications for the current generations of robots,” says professor Anthony Zador at Cold Spring Harbor Laboratory—but they are not “generalists” who can do varied tasks all on their own, as they mostly lack the abilities to improvise, make decisions based on a multitude of facts or emotions, and adjust to rapidly changing circumstances. “We don’t have general purpose robots that can interact with the world. We’re ages away from that.”
Robotic spatial self-awareness – the achievement by the team at Columbia – is an important step toward creating more intelligent machines. Hod Lipson, professor of mechanical engineering who runs the Columbia lab, says that future robots will need this ability to assist humans better. Knowing how you look and where in space your parts are, decreases the need for human oversight. It also helps the robot to detect and compensate for damage and keep up with its own wear-and-tear. And it allows robots to realize when something is wrong with them or their parts. “We want our robots to learn and continue to grow their minds and bodies on their own,” Chen says. That’s what Zador wants too—and on a much grander level. “I want a robot who can drive my car, take my dog for a walk and have a conversation with me.”
Columbia scientists have trained a robot to become aware of its own "body," so it can map the right path to touch a ball without running into an obstacle, in this case a square.
Jane Nisselson and Yinuo Qin/ Columbia Engineering
Today’s technological advances are making some of these leaps of progress possible. One of them is the so-called Deep Learning—a method that trains artificial intelligence systems to learn and use information similar to how humans do it. Described as a machine learning method based on neural network architectures with multiple layers of processing units, Deep Learning has been used to successfully teach machines to recognize images, understand speech and even write text.
Trained by Google, one of these language machine learning geniuses, BERT, can finish sentences. Another one called GPT3, designed by San Francisco-based company OpenAI, can write little stories. Yet, both of them still make funny mistakes in their linguistic exercises that even a child wouldn’t. According to a paper published by Stanford’s Center for Research on Foundational Models, BERT seems to not understand the word “not.” When asked to fill in the word after “A robin is a __” it correctly answers “bird.” But try inserting the word “not” into that sentence (“A robin is not a __”) and BERT still completes it the same way. Similarly, in one of its stories, GPT3 wrote that if you mix a spoonful of grape juice into your cranberry juice and drink the concoction, you die. It seems that robots, and artificial intelligence systems in general, are still missing some rudimentary facts of life that humans and animals grasp naturally and effortlessly.
How does one give robots a genome? Zador has an idea. We can’t really equip machines with real biological nucleotide-based genes, but we can mimic the neuronal blueprint those genes create.
It's not exactly the robots’ fault. Compared to humans, and all other organisms that have been around for thousands or millions of years, robots are very new. They are missing out on eons of evolutionary data-building. Animals and humans are born with the ability to do certain things because they are pre-wired in them. Flies know how to fly, fish knows how to swim, cats know how to meow, and babies know how to cry. Yet, flies don’t really learn to fly, fish doesn’t learn to swim, cats don’t learn to meow, and babies don’t learn to cry—they are born able to execute such behaviors because they’re preprogrammed to do so. All that happens thanks to the millions of years of evolutions wired into their respective genomes, which give rise to the brain’s neural networks responsible for these behaviors. Robots are the newbies, missing out on that trove of information, Zador argues.
A neuroscience professor who studies how brain circuitry generates various behaviors, Zador has a different approach to developing the robotic mind. Until their creators figure out a way to imbue the bots with that information, robots will remain quite limited in their abilities. Each model will only be able to do certain things it was programmed to do, but it will never go above and beyond its original code. So Zador argues that we have to start giving robots a genome.
How does one do that? Zador has an idea. We can’t really equip machines with real biological nucleotide-based genes, but we can mimic the neuronal blueprint those genes create. Genomes lay out rules for brain development. Specifically, the genome encodes blueprints for wiring up our nervous system—the details of which neurons are connected, the strength of those connections and other specs that will later hold the information learned throughout life. “Our genomes serve as blueprints for building our nervous system and these blueprints give rise to a human brain, which contains about 100 billion neurons,” Zador says.
If you think what a genome is, he explains, it is essentially a very compact and compressed form of information storage. Conceptually, genomes are similar to CliffsNotes and other study guides. When students read these short summaries, they know about what happened in a book, without actually reading that book. And that’s how we should be designing the next generation of robots if we ever want them to act like humans, Zador says. “We should give them a set of behavioral CliffsNotes, which they can then unwrap into brain-like structures.” Robots that have such brain-like structures will acquire a set of basic rules to generate basic behaviors and use them to learn more complex ones.
Currently Zador is in the process of developing algorithms that function like simple rules that generate such behaviors. “My algorithms would write these CliffsNotes, outlining how to solve a particular problem,” he explains. “And then, the neural networks will use these CliffsNotes to figure out which ones are useful and use them in their behaviors.” That’s how all living beings operate. They use the pre-programmed info from their genetics to adapt to their changing environments and learn what’s necessary to survive and thrive in these settings.
For example, a robot’s neural network could draw from CliffsNotes with “genetic” instructions for how to be aware of its own body or learn to adjust its movements. And other, different sets of CliffsNotes may imbue it with the basics of physical safety or the fundamentals of speech.
At the moment, Zador is working on algorithms that are trying to mimic neuronal blueprints for very simple organisms—such as earthworms, which have only 302 neurons and about 7000 synapses compared to the millions we have. That’s how evolution worked, too—expanding the brains from simple creatures to more complex to the Homo Sapiens. But if it took millions of years to arrive at modern humans, how long would it take scientists to forge a robot with human intelligence? That’s a billion-dollar question. Yet, Zador is optimistic. “My hypotheses is that if you can build simple organisms that can interact with the world, then the higher level functions will not be nearly as challenging as they currently are.”
This fall, Robert Reinhart of Boston University published a study finding that electrical stimulation can boost memory - and Reinhart was surprised to discover the effects lasted a full month.
Scientists believe brain circuits tend to uncouple as we age. Since brain neurons communicate by exchanging electrical impulses with each other, the breakdown of these links and associations could be what causes the “senior moment”—when you can’t remember the name of the movie you just watched.
In 2019, Boston University researchers led by Robert Reinhart, director of the Cognitive and Clinical Neuroscience Laboratory, showed that memory loss in healthy older adults is likely caused by these disconnected brain networks. When Reinhart and his team stimulated two key areas of the brain with mild electrical current, they were able to bring the brains of older adult subjects back into sync — enough so that their ability to remember small differences between two images matched that of much younger subjects for at least 50 minutes after the testing stopped.
Reinhart wowed the neuroscience community once again this fall. His newer study in Nature Neuroscience presented 150 healthy participants, ages 65 to 88, who were able to recall more words on a given list after 20 minutes of low-intensity electrical stimulation sessions over four consecutive days. This amounted to a 50 to 65 percent boost in their recall.
Even Reinhart was surprised to discover the enhanced performance of his subjects lasted a full month when they were tested again later. Those who benefited most were the participants who were the most forgetful at the start.
An older person participates in Robert Reinhart's research on brain stimulation.
Robert Reinhart
Reinhart’s subjects only suffered normal age-related memory deficits, but NIBS has great potential to help people with cognitive impairment and dementia, too, says Krista Lanctôt, the Bernick Chair of Geriatric Psychopharmacology at Sunnybrook Health Sciences Center in Toronto. Plus, “it is remarkably safe,” she says.
Lanctôt was the senior author on a meta-analysis of brain stimulation studies published last year on people with mild cognitive impairment or later stages of Alzheimer’s disease. The review concluded that magnetic stimulation to the brain significantly improved the research participants’ neuropsychiatric symptoms, such as apathy and depression. The stimulation also enhanced global cognition, which includes memory, attention, executive function and more.
This is the frontier of neuroscience.
The two main forms of NIBS – and many questions surrounding them
There are two types of NIBS. They differ based on whether electrical or magnetic stimulation is used to create the electric field, the type of device that delivers the electrical current and the strength of the current.
Transcranial Current Brain Stimulation (tES) is an umbrella term for a group of techniques using low-wattage electrical currents to manipulate activity in the brain. The current is delivered to the scalp or forehead via electrodes attached to a nylon elastic cap or rubber headband.
Variations include how the current is delivered—in an alternating pattern or in a constant, direct mode, for instance. Tweaking frequency, potency or target brain area can produce different effects as well. Reinhart’s 2022 study demonstrated that low or high frequencies and alternating currents were uniquely tied to either short-term or long-term memory improvements.
Sessions may be 20 minutes per day over the course of several days or two weeks. “[The subject] may feel a tingling, warming, poking or itching sensation,” says Reinhart, which typically goes away within a minute.
The other main approach to NIBS is Transcranial Magnetic Simulation (TMS). It involves the use of an electromagnetic coil that is held or placed against the forehead or scalp to activate nerve cells in the brain through short pulses. The stimulation is stronger than tES but similar to a magnetic resonance imaging (MRI) scan.
The subject may feel a slight knocking or tapping on the head during a 20-to-60-minute session. Scalp discomfort and headaches are reported by some; in very rare cases, a seizure can occur.
No head-to-head trials have been conducted yet to evaluate the differences and effectiveness between electrical and magnetic current stimulation, notes Lanctôt, who is also a professor of psychiatry and pharmacology at the University of Toronto. Although TMS was approved by the FDA in 2008 to treat major depression, both techniques are considered experimental for the purpose of cognitive enhancement.
“One attractive feature of tES is that it’s inexpensive—one-fifth the price of magnetic stimulation,” Reinhart notes.
Don’t confuse either of these procedures with the horrors of electroconvulsive therapy (ECT) in the 1950s and ‘60s. ECT is a more powerful, riskier procedure used only as a last resort in treating severe mental illness today.
Clinical studies on NIBS remain scarce. Standardized parameters and measures for testing have not been developed. The high heterogeneity among the many existing small NIBS studies makes it difficult to draw general conclusions. Few of the studies have been replicated and inconsistencies abound.
Scientists are still lacking so much fundamental knowledge about the brain and how it works, says Reinhart. “We don’t know how information is represented in the brain or how it’s carried forward in time. It’s more complex than physics.”
Lanctôt’s meta-analysis showed improvements in global cognition from delivering the magnetic form of the stimulation to people with Alzheimer’s, and this finding was replicated inan analysis in the Journal of Prevention of Alzheimer’s Disease this fall. Neither meta-analysis found clear evidence that applying the electrical currents, was helpful for Alzheimer’s subjects, but Lanctôt suggests this might be merely because the sample size for tES was smaller compared to the groups that received TMS.
At the same time, London neuroscientist Marco Sandrini, senior lecturer in psychology at the University of Roehampton, critically reviewed a series of studies on the effects of tES on episodic memory. Often declining with age, episodic memory relates to recalling a person’s own experiences from the past. Sandrini’s review concluded that delivering tES to the prefrontal or temporoparietal cortices of the brain might enhance episodic memory in older adults with Alzheimer’s disease and amnesiac mild cognitive impairment (the predementia phase of Alzheimer’s when people start to have symptoms).
Researchers readily tick off studies needed to explore, clarify and validate existing NIBS data. What is the optimal stimulus session frequency, spacing and duration? How intense should the stimulus be and where should it be targeted for what effect? How might genetics or degree of brain impairment affect responsiveness? Would adjunct medication or cognitive training boost positive results? Could administering the stimulus while someone sleeps expedite memory consolidation?
Using MRI or another brain scan along with computational modeling of the current flow, a clinician could create a treatment that is customized to each person’s brain.
While Sandrini’s review reported improvements induced by tES in the recall or recognition of words and images, there is no evidence it will translate into improvements in daily activities. This is another question that will require more research and testing, Sandrini notes.
Scientists are still lacking so much fundamental knowledge about the brain and how it works, says Reinhart. “We don’t know how information is represented in the brain or how it’s carried forward in time. It’s more complex than physics.”
Where the science is headed
Learning how to apply precision medicine to NIBS is the next focus in advancing this technology, says Shankar Tumati, a post-doctoral fellow working with Lanctôt.
There is great variability in each person’s brain anatomy—the thickness of the skull, the brain’s unique folds, the amount of cerebrospinal fluid. All of these structural differences impact how electrical or magnetic stimulation is distributed in the brain and ultimately the effects.
Using MRI or another brain scan along with computational modeling of the current flow, a clinician could create a treatment that is customized to each person’s brain, from where to put the electrodes to determining the exact dose and duration of stimulation needed to achieve lasting results, Sandrini says.
Above all, most neuroscientists say that largescale research studies over long periods of time are necessary to confirm the safety and durability of this therapy for the purpose of boosting memory. Short of that, there can be no FDA approval or medical regulation for this clinical use.
Lanctôt urges people to seek out clinical NIBS trials in their area if they want to see the science advance. “That is how we’ll find the answers,” she says, predicting it will be 5 to 10 years to develop each additional clinical application of NIBS. Ultimately, she predicts that reigning in Alzheimer’s disease and mild cognitive impairment will require a multi-pronged approach that includes lifestyle and medications, too.
Sandrini believes that scientific efforts should focus on preventing or delaying Alzheimer’s. “We need to start intervention earlier—as soon as people start to complain about forgetting things,” he says. “Changes in the brain start 10 years before [there is a problem]. Once Alzheimer’s develops, it is too late.”