Deep learning, which has in recent years become the dominant technique for creating new AIs, uses enormous amounts of data and computing power to fuel complex, accurate models. These resources are more accessible for researchers at large companies and elite universities. A study from Western University suggests there has been a “de-democratization” in AI: the number of researchers able to contribute to cutting-edge developments is shrinking. This may contribute to some of the ethical challenges facing AI development, including privacy invasion, bias and the environmental impact of large models.

To combat these problems, researchers are trying to figure out how to do more with less. One such recent advance is called “less than one”-shot learning (LO-shot learning). The principle behind LO-shot learning is that it should be possible for an AI to learn about objects in the world without being fed an example of each one. This has been a major hurdle for contemporary AI systems, which often require thousands of examples to learn to distinguish objects. Humans, on the other hand, are often able to abstract away from existing examples in order to recognize new never-before-seen items.

Allowing AIs to learn with considerably less data is important for several reasons. First, by building in abstractions that capture the relationships between objects, this technique reduces the potential for bias. Currently, deep-learning systems fall prey to bias arising from irrelevant features in the data they use to train. A well-known example of this problem is that AI classifies dogs as wolves when shown images of dogs in a snowy environment—because most images of wolves feature them near snow. Being able to zero in on relevant aspects of the image would help prevent these mistakes. Reducing data needs thus makes these systems less liable to this sort of bias.

Next, the less extensive the data one needs to use, the less incentive exists to surveil people to build better algorithms. For example, soft distillation techniques have already impacted medical AI research, which trains its models using sensitive health information. In one recent paper, researchers used soft distillation in diagnostic x-ray imagery based on a small, privacy-preserving data set.

Finally, allowing AIs to learn with less plentiful data helps to democratize the field of artificial intelligence. With smaller AIs, academia can remain relevant and avoid the risk of professors being poached by industry Not only does LO-shot learning make the barriers to entry lower by reducing training costs and lowering data requirements, but it also provides more flexibility for users to create novel data sets and experiment with new approaches. By reducing the time spent on data and architecture engineering, researchers looking to leverage AI can spend more time focusing on the practical problems they are aiming to solve.

A study presented at the just-finished American Society of Human Genetics Annual Meeting in San Diego reported a way to estimate whether an individual can expect to live longer or shorter than average.

An international research group studied the effect of genetic variations on lifespan across the human genome, which could improve our understanding of the diseases and cellular pathways involved in aging.

In the largest ever genome-wide association study of lifespan, the researchers paired genetic data from more than 500,000 participants in the United Kingdom Biobank and other cohorts with data on the lifespan of each participant’s parents.

They didn’t study the effects of one or more selected genes on lifespan, but looked across the whole genome to answer the question in a more open-ended way. The paper’s first author Paul Timmers from the University of Edinburgh said that because the effect of any given gene is so small, the large sample size was necessary to identify genes relevant to lifespan.

They confirmed six previously identified associations between genes and aging, such as the APOE gene, and they also discovered 21 new genomic regions that influence lifespan. Using their results to develop a polygenic risk score for lifespan, they developed a single, personalized genomic score that estimates a person’s genetic likelihood of a longer life.

“Using a person’s genetic information alone, we can identify the 10 percent of people with the most protective genes, who will live an average of five years longer than the least protected 10 percent,” said Timmers.

Also, they wanted to know whether genetic variants were affecting the aging process directly or affecting risk of individual diseases that could lead to death. They found that among common variants, those found in at least one in 200 people that are associated with Alzheimer’s disease, heart disease, and smoking-related conditions were linked to overall lifespan.

However, they did not find lifespan associations for other cancers, suggesting thatsusceptibilityto death caused by other cancers is due to rarer genetic variants or the environment.

That some people make weird associations between the senses has been acknowledged for over a century. The condition has even been given a name:synaesthesia. Odd as it may seem to those not so gifted, synaesthetes insist that spoken sounds and the symbols which represent them give rise to specific colours or that individual musical notes have their own hues.

Yet there may be a little of this cross-modal association in everyone. Most people agree that loud sounds are brighter than soft ones. Likewise, low-pitched sounds are reminiscent of large objects and high-pitched ones evoke smallness. Anne-Sylvie Crisinel and Charles Spence of Oxford University think something similar is true between sound and smell.

Ms. Crisinel and Dr. Spence wanted to know whether an odour sniffed from a bottle could be linked to a specific pitch. To find out, they asked 30 people to inhale 20 smells. After giving each sample a good sniff, volunteers had to click their way through 52 sounds of varying pitches, and identify which best matched the smell. The results of this study are intriguing.

The researchers’ first finding was that the volunteers did not think their request utterly ridiculous. It rather made sense, they told them afterwards. The second was that there was significant agreement between volunteers. Sweet and sour smells were rated as higher-pitched, smoky and woody ones as lower-pitched.

It is not immediately clear why people employ their musical senses in this way to help their assessment of a smell. But gone are the days when science assumed each sense worked in isolation. People live, say Dr. Spence and Ms. Crisinel, in a multisensory world and their brains tirelessly combine information from all sources to make sense, as it were, of what is going on around them.

Taste, too, seems linked to hearing. Ms. Crisinel and Dr. Spence have previously established that sweet and sour tastes, like smells, are linked to high pitch, while bitter tastes bring lower pitches to mind. Now they have gone further. In a study that will be published later this year they and their colleagues show how altering the pitch and instruments used in background music can alter the way food tastes. Volunteers rated the toffee eaten during low-pitched music as more bitter than that consumed during the high-pitched performance. The toffee was, of course, identical. It was the sound that tasted different.

Yawning can be a problem at the office for Lindsay Eierman, which makes her embarrassed. “I’ve explained, ’I’m sorry, I didn’t get much sleep last night,’” says Ms Eierman, a 26-year-old social worker from Durham, North Carolina. But a lack of sleep may not be the problem.

Researchers are starting to unravel the mystery surrounding the yawn, one of the most common and often embarrassing behaviours. Yawning, they have discovered, is much more complicated than previously thought. Although all yawns look the same, they appear to have many different causes and to serve a variety of functions.

Yawning is believed to be a means to keep our brains alert in times of stress. Contagious yawning appears to have evolved in many animal species as a way to protect family and friends, by keeping everyone in the group vigilant. Changes in brain chemistry trigger yawns, which typically last about six seconds and often occur in clusters.

To unravel the mystery of yawning, scientists built upon early, observed clues. Yawning tends to occur more in summer. Most people yawn upon seeing someone else do it, but infants and people with autism or schizophrenia aren’t so affected by thiscontagioneffect. And certain people yawn at surprising times, like parachutists who are about to jump out of a plane or Olympic athletes getting ready to compete.

A leading hypothesis is that yawning plays an important role in keeping the brain at its cool, optimal working temperature. The brain is particularly sensitive to overheating, according to Andrew Gallup, an assistant professor of psychology at the State University of New York at Oneonta. Reaction times slow and memory wanes when the brain’s temperature varies even less than a degree from the ideal 98.6 degrees Fahrenheit.

There are some practical applications. Dr. Gallup said managers might want to keep in mind the brain-cooling role of yawning when a meeting is long and boring. “One way to diminish yawning frequency in an office would be to keep it air-conditioned. If it’s very cold in the room, yawning rates are going to be quite low,” Dr. Gallup said.

Lonely people, it seems, are at greater risk than the gregarious of developing illnesses associated with chronic inflammation, such as heart disease and certain cancers. A paper published last year in the Public Library of Science, Medicine, shows the effect on mortality of loneliness is comparable with that of smoking and drinking after examining the results of 148 previous studies and controlled for factors such as age and pre-existing illness.

Steven Cole of the University of California, Los Angeles, thinks he may know why this is so. He told the American Association for the Advancement of Science meeting in Washington, D.C., about his work studying the expression of genes in lonely people. Dr. Cole harvested samples of white blood cells from both lonely and gregarious people. He then analysed the activity of their genes, as measured by the production of a substance called messenger RNA. This molecule carries instructions from the genes telling a cell which proteins to make. The level of messenger RNA from most genes was the same in both types of people. There were several dozen genes, however, that were less active in the lonely, and several dozen others that were more active. Moreover, both the less active and the more active gene types came from a small number of functional groups.

Broadly speaking, the genes less active in the lonely were those involved in staving off viral infections. Those that were more active were involved in protecting against bacteria. Dr. Cole suspects this could help explain not only why the lonely are iller, but how, in evolutionary terms, this odd state of affairs has come about.

The crucial bit of the puzzle is that viruses have to be caught from another infected individual and they are usually species-specific. Bacteria, in contrast, often just lurk in the environment, and may thrive on many hosts. The gregarious are therefore at greater risk than the lonely of catching viruses, and Dr. Cole thus suggests that past evolution has created a mechanism which causes white cells to respond appropriately. Conversely, the lonely are better off ramping up their protection against bacterial infection, which is a bigger relative risk to them.

What Dr. Cole seems to have revealed, then, is a mechanism by which social environment reaches inside a person’s body and tweaks its genome so that it responds appropriately. It is not that the lonely and the gregarious are genetically different from each other. Rather, their genes are regulated differently, according to how sociable an individual is. Dr. Cole thinks this regulation is part of a wider mechanism that tunes individuals to the circumstances they find themselves in.