A consortium of election researchers from across Europe has examined how populist movements frame their messages on Facebook. Their conclusion is unequivocal: populist parties are far more likely than their mainstream counterparts to foreground controversial subjects when courting voters.

Strategic move

This, the researchers suggest, is no accident. By thrusting polarising themes to the forefront, populists seek to shift public debate onto terrain that plays to their strengths.

The 2024 elections to the European Parliament unfolded amid a climate of crisis. Russia’s war in Ukraine raged on, migration was once again high on the political agenda, and climate change continued to preoccupy voters. Against this unsettled backdrop, some parties turned to “wedge issues”—topics designed to divide electorates and stir passions, often to the detriment of their rivals’ credibility and competence.

The researchers analysed 8,748 Facebook posts from political parties in 13 European countries, assessing both the prevalence of contentious subjects and the tone in which they were presented. The results were striking, if not wholly unexpected. Populist parties, they found, not only rely more heavily on controversial issues but also frame them in a distinctly populist idiom—replete with anti-liberal rhetoric and sharp “us versus them” distinctions.

In groups and out groups

Such messaging, the study suggests, is designed to exclude outsiders and reinforce a sense of group identity. It exemplifies the style of digital communication that has come to define populist politics: emotive, simplistic, and adversarial.

Digital platforms, by rewarding clarity and outrage, tend to flatten complex debates into binary choices. When several hot-button issues—migration, climate, war—compete for attention, the result is often heightened polarisation and diminished nuance.

As the researchers note, certain themes can serve as a “Trojan horse” for populist actors. Migration, for instance, is not merely invoked to debate border policy but to cast doubt on liberal democracy itself. The ultimate goal, they warn, is not to solve problems but to redefine the terms of public discourse.

A survey of nearly 2,500 adults in the United States and Brazil found that political allegiance, rather than cognitive decline, best explains the spread of misinformation among older users. The older people become, the researchers argue, the more tightly they cling to their political tribe—and the more likely they are to share stories that flatter it, regardless of their veracity.

“We found that older people are more likely to believe and share information that aligns with their party, whether that information is true or not,” the authors note.

A familiar pattern

Previous studies have documented the pattern. During the 2016 U.S. presidential election, Facebook users over 65 shared nearly seven times more fake news than those under 30. On Twitter, users over 50 were responsible for four-fifths of all misinformation shared. The reasons for this behaviour, however, have long been disputed.

Some researchers have attributed it to age-related cognitive decline—an erosion of analytical reasoning that makes older adults more vulnerable to deceit. Others point to difficulties in distinguishing paid content from genuine reporting. Yet a meta-analysis of 31 studies recently found that older adults are, on average, better than the young at recognising falsehoods.

The Colorado team recruited 1,700 Americans and 700 Brazilians aged between 18 and 80. Participants were shown a series of politically charged headlines, some of them fabricated, that either favoured conservative or liberal positions. One false U.S. headline, for example, claimed that “Pope Francis shocks world and endorses Donald Trump for President.” Another, circulating in Brazil, asserted that “Bolsonaro wants to cut 25% of civil servants’ salaries.”

Respondents were asked whether they would share each headline on social media, and in a second round, whether they believed it to be true. The researchers also measured political ideology and analytical reasoning.

Walking the walk

Their findings were clear. Older adults were just as capable as younger ones of distinguishing true from false. What differed was the degree of bias in applying that judgment. Those over 55 held one standard of evidence for information that favoured their side and another for that which did not. When a story reflected well on their party, they were far more likely to accept and share it uncritically.

This partisan reflex intensified with age—and appeared across political lines and across both countries. That Brazilians, despite their multiparty system, displayed the same tendencies suggests that the phenomenon stems less from institutional design than from human psychology.

Efforts to curb misinformation have traditionally focused on media literacy—teaching people to separate fact from fiction. The researchers argue that this approach, while useful, is incomplete. “Our study suggests it is equally important to encourage people to behave in a less partisan way when communicating on social networks,” they conclude.

Their prescription is simple, if not easy: think twice before posting, reflect on how political bias shapes what you share, and resist the urge to purge your feed of dissenting voices.

The study argues that fostering academic entrepreneurship requires more than funding schemes or intellectual-property reforms. Universities, incubators, and policymakers should also recognise the “identity work” involved—how scientists integrate their dual roles—and provide the flexibility to do so meaningfully.

More in common than we think

The research grew out of collaborations with academic scientists developing commercially viable products and services. “We repeatedly noticed that most scientists managed to combine being deeply committed academics with being for-profit entrepreneurs,” the authors note. This observation challenges the long-standing notion in academic literature that scholarly and entrepreneurial identities are inherently at odds.

As universities worldwide push for greater innovation and economic impact, the number of academic startups has climbed. Yet, many institutions still view entrepreneurship as a distraction from the pursuit of knowledge. The Gothenburg researchers sought to understand how scientists actually navigate the intersection of these two domains.

Through in-depth interviews with 27 Swedish academics who straddle both spheres, the team examined how these “dual actors” manage the frequent, small-scale transitions—“micro-transitions”—between science and business. They found that the participants did not speak of identity conflict in abstract terms but rather through their daily routines, tools, and tangible achievements.

Transferrable skills

Running a laboratory, several respondents explained, provided the same skills needed to build a startup. The experience of pitching to venture capitalists echoed the process of applying for research grants. Teaching, too, benefited: having real-world entrepreneurial experience lent authority and credibility in the classroom.

Concrete artefacts, the study found, often carried meaning in both worlds. A prototype developed for a commercial project might inspire new scientific questions, while a published paper could enhance the confidence of potential investors or customers. Professional identity, in other words, was forged in the doing—where scientific inquiry met practical application.

Contrary to prior assumptions, the researchers found no clear hierarchy between academic and entrepreneurial identities. What mattered most to participants was whether a task was intellectually stimulating and aligned with their expertise. Administrative chores—grading papers or managing payroll—were easily delegated. Setting a technology roadmap, securing funding, or building partnerships felt intrinsic to who they were as professionals.

Even so, uncertainty remained. Many interviewees expressed anxiety over what was “appropriate” or “allowed” within academic institutions. The authors therefore recommend that universities and policymakers provide clearer guidelines for dual affiliations, shared resources and recognition systems that reward entrepreneurial engagement alongside scholarly output.

“The goal,” they conclude, “is not to eliminate all tension, but to make it transparent and productive—so that academics can pursue both science and entrepreneurship to mutual benefit.”

While the platform was billed as being fully autonomous, this was largely untrue (much as claims of Artificial General Intelligence today are). Instead, the startup team feeds it pre-written lines, filters its outputs, and manipulates its apparent spontaneity.

Trust in AI

The public begins by trusting the AI candidate, thinking that it will be rational and immune to the kind of sleaze and corruption that blight human politicians. After all, AI surely can’t lie or take bribes.

This trust begins to wither away as the platform begins making some strange decisions, however, and conspiracy theories circulate about who really controls both the platform and the startup behind it. After a while, the public began to view PresiBot as a mere puppet of tech billionaires rather than a neutral mind.

While the book is set in 2040, the use of AI by politicians is increasingly widespread. We’ve had infamous examples, such as Donald Trump’s frequent use of generative AI to produce media for his posts on Truth Social and X, but it’s increasingly evident that politicians and policymakers are using AI in many of the same ways in which all of us do.

Political AI

A recent study from London’s Brunel University looks at how people feel about politicians using AI. The study builds on previous work I covered, which looked at how the public feels about AI being used to make welfare-related decisions (TLDR: those unaffected by such decisions are generally happier with AI involvement than those who are on the receiving end).

Given the hype surrounding AI, not least as a driver of economic growth, politicians across the world have been jumping onto the bandwagon, with terms like sovereign AI leaping into the public discourse.

It’s perhaps no great surprise that politicians are using AI themselves, whether to write speeches, engage with the electorate, or research key policy matters. For instance, Peter Kyle, the then UK Technology Secretary, used the AI chatbot ChatGPT for policy advice, to brainstorm ideas for media appearances, and to define terms relevant to his ministerial role in early 2025.

Mirroring real life

The results of the study suggest similar attitudes to political use of AI as in other walks of life. In other words, if politicians are using AI to help them think, there’s a general acceptance, but if it’s doing the thinking for them, this turns the public off.

The researchers surveyed people across both the UK and Japan to gauge their response to politicians using AI in their work. There were some interesting national differences. For instance, in the UK, around half of respondents didn’t like politicians using AI at all, with 80% saying they were unhappy about AI making decisions for representatives.

Respondents in Japan were slightly more open to the technology, but even then, they were extremely opposed to the thought of AI making decisions for elected officials.

Trust in politics

Interestingly, the researchers found that the level of trust people have in politics more generally underpins the trust they have in using AI in politics. This is probably because if we trust the people, we trust that they’ll use technology in the right way.

This is somewhat worrying, as research last year from the University of Southampton found that trust in politics and democratic institutions was falling across the world. The study, which was conducted across 143 countries, found that while trust in elected bodies is declining, confidence in non-representative institutions like the police, civil service, and legal system has remained stable or even grown. This suggests a crisis of confidence, specifically in elected representatives.

Trust in the technology also played a role, with those supportive of AI in general more supportive of politicians’ use of it, with the opposite the case for those who were more opposed to AI.

It’s clear that AI, as in other walks of life, will find its way into public office. These findings suggest that how it’s used is arguably more important than whether it’s used. With trust in politics and democracy so fragile, it’s vital that people think AI is being used to support elected officials rather than replace them.

As in 2040, if people feel like AI is being misused by elected officials, then trust in both the humans and the tech will almost certainly evaporate incredibly quickly. Whether politicians heed this warning or not remains to be seen.

Research from UC Berkeley and Harvard reminds us that this geographic shift isn’t a simple, zero-sum affair whereby one region gains at the expense of another.

Shifting gains

The authors describe how the 1980s saw significant growth in cities like San Francisco and New York, but this growth ran up against natural buffers as housing supply struggled to keep pace with the growing local economies. This forced up prices and created opportunities for other cities, such as Denver and Austin, to grab a piece of the action.

This was not the case for former industrial powerhouses, such as Detroit, however. As the wider economy deindustrialized, these regions often struggled to attract enough knowledge workers to transition their economies.

This had an impact on an individual level, as while the traditional neo-liberal view is that workers should simply move to where opportunities exist, the reality is that many were rooted by a lack of money or flexibility.

“The geography of where you live has always mattered, but it matters even more so now,” the researchers explain. “Access to good jobs can be quite different in different regions.”

Where are the good jobs?

While most studies investigating the labor market look at things like earnings and how different places have different wage profiles, the Berkeley team wanted instead to look at so-called good jobs. These are defined as jobs in sectors where the average wage is higher than the average wage of at least two-thirds of all other industries.

The researchers looked at cross-sectional data on the American labor market from 1980-2021 and found a couple of key shifts. Firstly, the share of good jobs in manufacturing halved across that timeframe, which is a huge change in the kind of jobs available to certain sections of both the labor market and the country as a whole.

The second transformation involved the rise in service jobs that were human capital-intensive. These tend to be in industries like IT, professional services, and finance. Whereas they made up just 7.6% of jobs in 1980, they were 26.2% in 2021.

For those who aren’t able to get such jobs, perhaps because of a lack of a college degree, most of the good jobs were in areas like mining and construction. Indeed, the data shows that the share of good jobs in these sectors grew from 28.5% to 43.5%, with a lot of these being in construction.

Winning regions

The data suggests that some regions were able to embrace these shifts, and so not only had workers able to fulfil these roles but also the infrastructure to support them.

“The way that areas recover is not by regaining manufacturing jobs. Those are long gone,” the researchers explain. “It’s by attracting new industries that offset the losses in manufacturing like human capital-intensive service jobs.”

Housing, or more particularly the space to increase housing supply, also played a big role in a city’s growth, with places like Atlanta and Phoenix capitalizing on the growing demand for service jobs.

Wider factors

Housing and other factors often play a key role in who, and indeed why, people move, however. A study from Johns Hopkins found that low-income families often move out of necessity, rather than by choice. Housing instability, landlord behavior, and financial shocks force them into “reactive” moves that prioritize immediate survival rather than access to good jobs, schools, or safe neighborhoods.

The study reminds us that it’s not just a case of where jobs are, but also who can actually move to access them. This was further underlined by research from the University of Connecticut, which shows that interstate mobility in the U.S. has been falling steadily.

The study found that social ties and personal rootedness are often big factors in people choosing to stay put, with these community ties overriding the availability of better work opportunities elsewhere. As the Johns Hopkins study reminds us, people don’t appear to be moving unless they’re forced to do so. In other words, lower-income people are moving in a reactive rather than proactive way, thus creating a kind of mobility paradox that reinforces inequality.

Where opportunities will emerge

The rise of generative AI presents not just a question of which jobs will be automated, but where opportunity will concentrate, and who will be able to access it. History suggests that without deliberate intervention, technological transitions deepen geographic divides rather than heal them.

While economic theory suggests that workers will move to where the opportunities are, the actual evidence suggests that those most in need of migrating towards the jobs are usually least willing or able to do so. Similarly, the regions that could most benefit from developing new industries often lack the housing supply, infrastructure, or human capital to attract them.

We’re on the cusp of an AI-driven economic transformation, so policymakers face a clear choice. It’s likely that AI-driven growth will follow the same patterns as the previous rise in service jobs, which will make prosperous hubs even more prosperous, while those on the outside wither still further. As such, policymakers can, and should, actively work to distribute opportunity more equitably through investments in education, housing, and infrastructure in left-behind communities.

The geographic winners and losers of the AI economy are not predetermined. Without intentional policy choices, however, we risk amplifying the kind of regional inequalities that have already defined us over the past few decades. Are we willing to learn these lessons, or are we happy to let AI continue to deepen the regional divides we’re already seeing?

A new study from Concordia University suggests that vocal volume in offices is shaped by more than mere personality. While individual habits matter most, the design of the workspace and the type of communication taking place also play a role.

Real life

Rather than rely on laboratory simulations, researchers examined speech levels in two functioning offices—one in Montreal, the other in Quebec City—where more than 70 employees conversed in English or French. Using precision sound meters, they measured how loudly people spoke across a range of settings: open-plan offices with and without partitions, private offices, and both small and large meeting rooms.

Conversations spanned casual chats and online meetings, with and without headsets. The study, carried out in partnership with the Montreal acoustics firm Soft dB, also considered background noise and room acoustics.

The findings were revealing. Workers in open-plan offices with partitions spoke markedly louder than those without them. The partitions, intended to block sound, seem to have the paradoxical effect of encouraging higher volume. “People may feel they can speak more freely, unaware that the barriers tempt them to raise their voices,” the researchers note.

Meeting rooms, by contrast, showed steady sound levels regardless of size, while teleconferences prompted slightly louder speech than face-to-face discussions. Language had no measurable impact on how loudly people spoke.

Lower than expected

Perhaps most strikingly, speech levels recorded in real offices were lower than those assumed in industry standards. Such benchmarks are typically derived from tests in anechoic chambers—spaces devoid of echoes—where speakers lack the natural feedback that helps modulate voice strength. The implication is that current standards for speech privacy and acoustic design may exaggerate how noisy workplaces really are.

The study also challenges the notion that silence is golden. Too little background noise, it finds, can be as disruptive as too much. A modest hum—whether from ventilation systems, chatter, or keyboards—helps mask distractions and may even improve concentration. In office acoustics, it seems, a little noise can be a good thing.

Humanity is living longer than ever before, even as birth rates fall. More than half of the world’s countries now record fertility below replacement level. The result is a demographic transformation—a “global greying”—that will test labour markets, pensions, and health systems.

Reframing getting old

In Seven Decades: How We Evolved to Live Longer, Michael Gurven of the University of California, Santa Barbara, argues that humans are evolutionarily built to live at least seven decades. Recent gains in life expectancy, he suggests, reflect fewer deaths in early life rather than any real extension of our biological limits. The capacity to survive well past menopause, when most species have long succumbed to nature’s “Wall of Death,” evolved thousands of years ago among hunter-gatherers.

Today’s average lifespan in rich countries—eight or nine decades—represents an improvement on that ancient baseline. But Gurven’s focus is not merely on years lived; it is on how humans have adapted socially and biologically to make those years meaningful.

Drawing on decades of anthropological fieldwork among nonindustrial societies such as the Tsimané of the Bolivian Amazon, Gurven finds that aging is not an aberration but a feature of our species. Life expectancy among hunter-gatherers may appear short, yet those who survive childhood often reach their seventies or eighties. What distinguishes humans from most animals, he argues, is not the length of our youth but the purpose of our old age.

Natural selection, in its ruthless logic, favours reproduction. Most species die soon after their reproductive years end. Humans, by contrast, evolved to remain useful long after fertility wanes. The secret lies in cooperation. Early human societies shared food and knowledge across generations; elders contributed wisdom, childcare, and craft skills even when physical strength declined. This interdependence, Gurven suggests, increased group survival and thus reinforced the genetic advantage of longevity.

Being useful

Among subsistence populations, “old age” is defined less by chronology than by capability. Many languages in such communities lack numbers beyond ten, and age is gauged instead by one’s ability to hunt, farm, or teach. Retirement, as industrial societies know it, does not exist. Instead, lifelong participation sustains both elders and youth.

Gurven believes modern societies could learn from this model. Rather than treating old age as withdrawal, we might view it as a stage of continued contribution—an opportunity to apply accumulated knowledge, mentor younger generations, and sustain community life.

“The implications of global aging can seem daunting,” Gurven notes. “But the very reason our species thrived was because middle-aged and older adults played vital roles in collective success.” Old age, in other words, is less a biological sentence than a social choice. Those who keep learning, engaging, and adapting may discover that feeling “old” is as much a matter of perspective as of years.

In other viral posts, bears and deer bounce on trampolines, or a trio of raccoons floats serenely down a river atop three crocodiles. None of it is real. All are products of artificial intelligence—fabrications rendered with such convincing realism that many viewers mistake them for genuine wildlife encounters.

More harm than good

Researchers from the GESBIO group at the University of Córdoba warn that such videos, though seemingly harmless, may be distorting public understanding of the natural world. In a recent analysis of the most-shared AI-generated wildlife clips, they found that these creations foster false perceptions of animal behaviour and habitats, promote anthropomorphism, and deepen the growing disconnection between people and nature.

“They depict species behaving in ways that are biologically implausible,” the researchers note. “Predators and prey are shown playing together; wild animals display human emotions or social traits. The leopard-and-cat video, for instance, misleads viewers into thinking such encounters could occur—and diminishes the true conservation challenges these species face.”

The consequences extend beyond mere misinformation. Previous work by the same team, through the IncluScienceMe project, revealed that many primary-school children already struggle to identify local wildlife. The flood of AI-generated videos, the researchers warn, reinforces a “false intimacy” with nature: vulnerable or exotic species appear abundant and friendly, while ordinary native ones are overlooked.

Spreading disappointment

This distortion can breed disappointment and disengagement. Children who expect to find capybaras or leopards in their local parks may lose interest when reality fails to match the fantasy. Because early learning depends heavily on visual cues, and because social media now serves as a dominant source of information, such misconceptions risk becoming entrenched.

Another worrying trend is the rise in demand for exotic pets. The portrayal of wild animals as docile or companionable encourages people to acquire them illegally or irresponsibly.

The researchers advocate for stronger media literacy and environmental education. Citizens, they argue, need the tools to question online content and verify its authenticity through credible sources. Schools should teach the difference between native and exotic species and instil a clear sense of what wildlife actually inhabits their regions—“so that children understand there are no lions here.”

Their study, among the first to examine the intersection of artificial intelligence and biodiversity conservation, highlights a new frontier in misinformation. The digital age, it seems, has not only blurred the line between fact and fiction—it has also redrawn the boundary between humanity and the natural world.

Across five studies, the researchers found that people expect to feel worse when telling others about charitable acts—such as donating to those in need—than when keeping such deeds to themselves or sharing personal accomplishments like job promotions. “It was a surprise,” the authors note, “to see how intuitive people are about this—that they anticipate feeling bad afterwards.”

Making us happy

Participants were asked to recall both a good deed and a personal achievement, then rate how happy, proud, ashamed, and embarrassed each made them feel. They were then asked to imagine describing these moments to a friend, or posting about them on social media. The pattern was consistent: participants predicted feeling more pride when sharing achievements, but more shame and embarrassment when talking about good deeds—particularly in the public glare of social media.

The researchers believe this self-consciousness stems from an acute awareness of social perception. “People suspect that if they talk about their good deeds, others will think they were motivated by the reputational boost,” they explain. This fear of seeming self-serving can wash away the “warm glow” of altruism, replacing it with discomfort or guilt.

Interestingly, participants also assumed that others would feel better than they would in the same situation—perhaps because people struggle to imagine others’ inner states as vividly as their own. “Our simulations of other minds are shallower,” the researchers suggest.

The findings carry a wry lesson for the age of performative virtue. While sharing charitable acts may win social credit, it risks diminishing their emotional reward. As Oscar Wilde once quipped, “The nicest feeling in the world is to do a good deed anonymously—and have somebody find out.”

Tonal choices

Digital communication forces people to make constant tonal choices. Previous studies—and popular lore—have long tied exclamation marks to friendliness and sincerity, though often at the expense of perceived competence. A simple example illustrates the nuance: “Sure. That sounds great.” might read as flat, reluctant or even passive-aggressive. Add two exclamation points—“Sure! That sounds great!”—and the message becomes unmistakably enthusiastic. In professional contexts, however, that same punctuation can sound breathless or over-eager: “There will be an overview meeting at noon tomorrow! The final report is due next week!” risks implying alarm, where a plain period conveys calm authority.

To examine how such judgments play out, researchers ran five studies involving more than 1,700 participants, recruited through online platforms and university pools. Across all experiments, messages punctuated with exclamation points were more likely to be perceived as written by women.

Women, the researchers found, not only use exclamation marks more frequently but also report being more self-conscious about them—worrying about seeming overly enthusiastic or less competent. Men, meanwhile, perceive stronger social norms against their use. Both sexes agree that exclamations project warmth and approachability but can slightly diminish impressions of analytical thinking and power.

Striking the balance

Follow-up studies confirmed that while exclamation points enhance perceptions of friendliness and enthusiasm, they do not consistently undermine perceived competence. Nor do their effects depend on the sender’s gender. In short, readers interpret exclamation marks in gendered ways, but the social consequences apply evenly.

The findings align with social-role theory, which holds that gender norms guide communication styles. Women, who are expected to signal warmth, face greater pressure to use exclamations—and greater scrutiny for doing so.

For those agonising over email tone, the lesson is pragmatic: punctuation is strategic. Use exclamation points to convey warmth or enthusiasm; omit them when authority or analytical rigour is the goal. In the modern workplace, even a mark as small as “!” can speak volumes.