Source: The Conversation (Au and NZ) – By Heather Douglas, Professor of Law and Deputy Director of the Centre of Excellence for the Elimination of Violence Against Women (CEVAW), The University of Melbourne

Every year across Australia, more than 100,000 people obtain a family violence protection order.

For some victim-survivors, protection orders provide a much-needed safety net. They can help prevent further violence from occurring.

Protection orders are future looking. They aim to stop a specific person (the respondent) using family violence against the victim-survivor in the future.

While they can be helpful, the system is far from perfect. Here’s how these orders work and what could be done to improve them.

What is a protection order?

When a victim-survivor has experienced, or is experiencing family violence, they may apply for a protection order. Any adult, and in some places children, can apply.

Evidence shows most orders are made to protect women against their current or previous male intimate partner. In some places, it’s common for police to apply for the order to protect the victim-survivor.

Generally, a magistrate can make a final protection order when they are satisfied the respondent has committed family violence against the victim-survivor and is likely to continue to do so, or to do so again.

A protection order includes conditions. These may include that the respondent must not commit family violence, that they stay a certain distance away from the victim-survivor’s home, work, or school or that they do not not contact the victim-survivor except in a specific way (such as through a lawyer).

Civil or criminal?

The protection order system is described as a hybrid civil/criminal system.

The process of obtaining a protection order is a civil process, and the magistrate must be satisfied on the balance of probabilities that the order is appropriate.

Police are expected to enforce protection orders. If a condition is breached, the respondent can be charged with a criminal offence of breaching the order.

If the magistrate is satisfied beyond a reasonable doubt that a condition has been breached, the respondent will be found guilty. Beyond a reasonable doubt is a higher standard of proof than the civil standard of balance of probabilities.

Depending on the situation, a breach of a protection order can result in a prison sentence. The threat of criminal sanction provides an incentive to some respondents not to be violent.

A court quagmire

The protection order system in each state and territory is slightly different. The orders have different names, the definitions of family violence that underpin them is different, the orders made can have different durations from months to years, and they apply to different relationships.

Protection orders are part of state and territory law. The family law system is part of the federal law system. This state/federal divide can be a problem.

State and territory magistrates have the power to include children as protected people on protection orders across Australia. Magistrates can also vary family court orders, where they think people are unsafe.

However, victim-survivors report some magistrates are reluctant to include children on protection orders. It’s also rare for magistrates to vary orders coming from the family court. This may be because some magistrates see the family law system as responsible for making orders about children.

This state/federal divide often requires victim-survivors to navigate two separate court systems to seek protection and resolve parenting or property disputes.

The disconnect between systems also facilitates systems abuse with respondents playing off systems against each other, delaying legal cases, forcing ongoing contact and further abusing victim-survivors.

More work to do

Despite these challenges, protection orders have been associated with reduced domestic violence. Research from the Australian Institute of Criminology found the orders seem to be more effective where the victim-survivor can be more independent and has fewer ties to the respondent.

In 2017 laws were changed so that a protection order made in one state or territory can be enforced by police in another state or territory. This ensures victim-survivors do not need to apply for a new protection order when they move interstate.

However, the presence of a protection order does not guarantee safety for victim-survivors. In 40% of cases where a woman was killed by a current or former partner, she had a protection order.

In some cases, police misidentify victim-survivors as the violent person and take out a protection order against the wrong party.

Read more: South Australia’s domestic and sexual violence royal commission recommendations should be embraced across the country

The institutions and services that are responsible for keeping victim-survivors safe – including police and courts – have more work to do. This includes better enforcement of breaches of orders and taking allegations of family violence seriously (including non-physical abuse).

Culturally-responsive approaches must be embedded so that First Nations people are not over-criminalised.

Victim-survivors need better support to obtain protection orders through accessible information, trauma-informed practices and greater connections between the different systems.

Protection orders have the potential to improve safety for people experiencing family violence. But, the message from victim-survivors is clear: to save lives, these orders have to be policed properly and taken seriously.

– ref. Family violence protection orders can be a lifeline, but the system needs reforming – https://theconversation.com/family-violence-protection-orders-can-be-a-lifeline-but-the-system-needs-reforming-278544

Source: The Conversation (Au and NZ) – By Dylan A Mordaunt, Research Fellow, Faculty of Education, Health, and Psychological Sciences, Te Herenga Waka — Victoria University of Wellington; Flinders University; The University of Melbourne

Across child protection services, frontline staff are often making decisions in the hardest possible conditions: under time pressure, with incomplete information and high stakes on every side.

Get it wrong and the consequences are serious. A child may remain in danger. Or a family may be disrupted unnecessarily, with harms of its own.

There is also a triage problem. Some families need urgent intervention. Some need support. Some need monitoring. And some need less intrusion, not more.

In practice, those judgements already rely on reading signals from fragmented information and, in effect, making predictions about risk.

Predictive modelling aims to make that process more systematic. By analysing patterns in large administrative datasets, it can help identify which children may be most at risk of future harm.

With New Zealand’s social workers under more strain than ever, what are the opportunities of using these tools more actively – and what are the potential dangers?

NZ and predictive analytics

New Zealand is no stranger to predictive modelling, nor debate surrounding it.

More than a decade ago, it was among the first countries to seriously explore how predictive modelling could be applied to child protection.

Work led by Professor Rhema Vaithianathan and colleagues at the Auckland University of Technology showed that integrated administrative data could identify newborn children at elevated risk of later maltreatment.

Still, agencies have been deliberately cautious in framing how these models might be used.

The Ministry of Social Development has said they should enhance intake decisions, support rather than replace professional judgement and first be tested in a simulated setting. A Statistics New Zealand peer review echoed that point: a model should trigger closer assessment, not automatic intervention.

Steps to move from research to practice have nonetheless proved contentious.

A proposed 2015 observational study – which would have assigned risk scores to newborns and tracked outcomes – was ultimately halted amid concerns about privacy, bias and the role of the state.

While these concerns have not disappeared, neither has pressure on the system. Oranga Tamariki received more than 55,000 reports of concern in the second half of 2024 – a sharp increase on the previous year.

Recent internal surveys of the agency’s frontline staff meanwhile highlight how cases are becoming more complex and that decisions are being made under uncertain conditions.

Predictive modelling tools, however, are still not used by those workers. To date, testing of the technology has been carefully limited to historical, anonymised data – and carried out alongside extensive ethical, privacy and Māori-led reviews.

Promise and pitfalls

Where predictive modelling has been piloted in the United States, post evaluations have suggested it can help if used carefully.

In Pennsylvania’s Allegheny County, for instance, one pilot programme resulted in fewer children being removed from their homes. In another in Los Angeles, cases where children suffered life-threatening harm was observed to fall by 23%.

This suggests that models can add more precision to interventions. But it hasn’t always been the case.

Authorities in Illinois abandoned one system after it produced too many alerts. It was also criticised for missing cases that resulted in tragedy, despite the children already known to child welfare agencies.

This demonstrated that if a model overwhelms workers with data it can simply add clutter instead of reducing harm.

Another risk facing frontline workers is what are called “false negatives”, such as missed cases, and “false positives”, such as wrongful accusations.

The former can mean a child remains unsafe. The latter can mean a child is removed unnecessarily, with serious and lasting consequences.

This challenges the logic of workers “erring on the side of caution” in their decision-making.

If caution means reflexive removal, it can create a different form of damage. Here, the case for predictive analytics is arguably strong.

Should ‘do nothing’ stay an option?

In New Zealand, there are obvious sociological factors that make this issue more complex. One is the risk that existing patterns of inequality are reproduced, because Māori are disproportionately represented in child protection pathways.

That pattern is not unique to Aotearoa: in Australia, Aboriginal and Torres Strait Islander children are around 11 times as likely as non-Indigenous children to be in out-of-home care. That is why Indigenous data sovereignty cannot be an afterthought in any moves to use predictive modelling.

Nor is it enough to simply say a model is “evidence-based”. Agencies need to be clear about what data is being used, what it is trying to optimise, how decisions can be overridden, how bias is monitored and who can challenge it.

It may seem safer to reject these tools on perceived moral grounds. Often, it is simply the more familiar choice.

But doing so does not create a neutral system – it means relying on inconsistent judgements made under pressure, with uneven information and little ability to test whether decisions are improving.

Predictive analytics will not fix deeper system failures. But, if carefully governed, it can help prioritise urgency, target support and make decisions more transparent and informed.

– ref. Child protection workers are under pressure in NZ. Can predictive modelling help? – https://theconversation.com/child-protection-workers-are-under-pressure-in-nz-can-predictive-modelling-help-278298

Source: The Conversation (Au and NZ) – By Kevin Veale, Senior Lecturer in Media Studies, School of Humanities, Media and Creative Communication, Te Kunenga ki Pūrehuroa – Massey University

During the first world war, the British government was looking for ways to help people stretch their limited food supplies. It found pamphlets from a noted 19th-century herbalist who said rhubarb leaves could be used as a vegetable along with the stalks.

The government duly printed its own pamphlets advising people to eat rhubarb leaves as a salad rather than throwing them out. There was one problem: rhubarb leaves can be poisonous. People reportedly died or became ill.

The advice was corrected and the pamphlets pulled from circulation. But during the second world war, the government was again looking for ways to stretch food supplies.

It found a stockpile of old resources from the previous war that explained unorthodox sources of food, including rhubarb leaves. Reusing the pamphlets seemed an efficient thing to do, so they were sent out to the public. Once again, people reportedly died or became ill.

Those pamphlets were misinformation, but the public had no reason to suspect them either time. They were official resources developed by the government – why wouldn’t they be safe?

That is how misinformation can cause problems even after the initial error is corrected. And the moral of the story still reverberates in the age of generative artificial intelligence (AI).

Chatbots are not search engines

Generative AI is used to generate text and images (and other forms of data) based on original information it has ingested. But it can also be an engine for churning out misinformation faster than people can produce safe information, let alone fact-check and correct it.

And as the rhubarb story illustrates, corrections can’t always properly remove the original contamination.

AI platforms such as ChatGPT and Claude don’t work like a conventional search engine. But people use them as one because they seem to summarise complex topics quickly and require fewer clicks than conventional internet searches.

Search engines rely on articles and text about a given topic, and then weigh how reliable those articles are. Generative AI instead relies on huge bodies of text, from which it measures the odds of words appearing next to each other.

These “large language models” are purely looking to generate reasonable-looking sentences, rather than accurate ones.

For example, if “green eggs and ham” appeared frequently enough in its huge pile of words, it is more likely to describe “eggs and ham” as green if someone asks.

‘Plausible yet incorrect’

OpenAI, which developed ChatGPT, has admitted (based on its own study) there’s no way to stop false information being presented as truth due to the way generative AI works. Explaining why large language models “hallucinate”, the researchers wrote:

Like students facing hard exam questions, large language models sometimes guess when uncertain, producing plausible yet incorrect statements instead of admitting uncertainty.

This can have real-world consequences. One recent study showed ChatGPT failed to recognise a medical emergency in more than half of cases. This can be exacerbated by already existing errors in medical records, which a UK inquiry in 2025 found affected up to one in four patients.

While a doctor might order more tests to confirm a diagnosis, one researcher explained that generative AI “delivers the wrong answer with the exact same confidence as the right one”.

The problem, as another scientist noted, is that generative AI “finds and mimics patterns of words”. Being right or wrong is not really the point: “It was supposed to make a sentence and it did.”

Research has shown generative AI tools misrepresent the news 45% of the time, no matter the language or geographic region. And there is now genuine concern about AI risking lives by generating non-existent hiking routes.

It’s easy to make fun of generative AI when it advises people to eat rocks or hold toppings on a pizza base with glue.

But other examples aren’t so amusing – such as the supermarket meal planner that suggested a recipe that would produce chlorine gas, or the dietary advice that left someone with chronic toxic exposure to bromide.

Look for older information

Education and establishing good rules around the appropriate and cautious use of generative AI will be essential, especially as it makes inroads into governments, bureaucracies and complex organisations.

Politicians are already using generative AI in their everyday work, including for policy research. And hospital emergency departments are using AI tools to record patient notes to save time.

One safeguard is to try to source more reliable information produced before AI-contaminated text and imagery infiltrated the internet.

There are even tools available to help simplify that process, including one created by Australian artist Tega Brain “that will only return content created before ChatGPT’s first public release on November 30 2022”.

Finally, if your instinct is to fact-check the story at the start of this article, good old-fashioned books might be your best bet: references to how the British government twice encouraged rhubarb poisoning can be found in the The Poison Garden’s A-Z of Poisonous Plants and Botanical Curses and Poisons: The Shadow Lives of Plants.

– ref. Using your AI chatbot as a search engine? Be careful what you believe – https://theconversation.com/using-your-ai-chatbot-as-a-search-engine-be-careful-what-you-believe-277616

Source: The Conversation (Au and NZ) – By Nikhil Datta, Assistant Professor, Economics, University of Warwick

The US-Israel strikes on Iran in late February caused an immediate spike in oil prices, and volatility has only increased since then. It quickly led to fears among motorists of “price-gouging” – petrol retailers raising their prices to take advantage of consumer panic.

In the UK, Chancellor Rachel Reeves asked the Competition and Markets Authority (CMA) to remain on “high alert” for profiteering by petrol retailers. Trade body the Petrol Retailers Association quickly hit back, saying her language was “incorrect and inflammatory”.

But what does the economic evidence suggest about retailers’ behaviour at times when oil prices are fluctuating wildly? As part of our yet-to-be-published research into UK petrol retailers and large oil price shocks, we examined Russia’s full-scale invasion of Ukraine in February 2022.

The invasion led to a large and sudden increase in global oil prices, providing a valuable context in which to determine how shocks to crude oil supply filter through to prices at the pump.

The first striking pattern we found was that wholesale unleaded and diesel price changes closely tracked crude oil price changes. When oil prices rose, wholesale fuel prices increased almost immediately. Our estimates suggest that roughly 80% of changes in oil prices are reflected in wholesale fuel prices within a few days.

Read more: What oil, stocks and bonds are telling us about the Iran conflict and how long it might last

Retail prices, however, react quite differently. Prices at the pump adjusted more slowly and were considerably smoother than wholesale prices. In periods where wholesale prices increased sharply, retail prices typically rose by less and with a delay.

At the immediate peak of the shock in the weeks following the invasion, wholesale diesel prices rose by about 39 pence per litre, while pump prices increased by only about 16 pence per litre.

The implication is that retailer margins compressed during price spikes as the gap between retail and wholesale prices narrowed temporarily. In other words, although consumers experienced higher petrol prices, the evidence does not suggest that retailers increased their markups during these periods.

But why would retailers reduce their margins when prices spike? One explanation is that consumers become more aware of petrol prices at these times. Using data from price comparison site PetrolPrices.com, we found that when average petrol prices rose above £1.50 per litre during 2022, search activity increased dramatically. The growing number of daily searches indicated that consumers were actively seeking out cheaper filling stations when prices increased.

The crossing of the £1.50 threshold also attracted media attention, increasing people’s awareness and encouraging consumers to compare prices. By using geographically granular data on search activity, combined with daily petrol price data from nearly all petrol stations in the UK, we can causally link this increase in consumer attention with intensifying price competition.

As prices began to stabilise, we found that search intensity on the price comparison site dropped. Search activity itself did not return to pre-shock levels, but instead dropped and plateaued at a higher level than before, consistent with predictions from well-established economic models.

Correspondingly, price impacts narrow over time. At the peak of increased search activity following the Russian invasion of Ukraine, a 10 percentage point increase in search activity was associated with roughly a 2% reduction in local area petrol prices. We then found that this was driven primarily by stations that already had higher prices in January 2022. These higher-priced petrol stations cut their prices the most as consumers became more price-sensitive.

The research suggests that when oil prices increase and there is lots of media attention, consumers make more effort to search for better prices. Competition then increases and this puts downward pressure on retail prices. So retailers may actually experience falling margins when oil prices spike.

Rockets and feathers

It seems that it is not the level of prices that drives consumer attention, but whether those prices are rising rapidly. As price increases slow or reverse, consumers search price-comparison sites less intensively, reducing the sense of competition between petrol stations.

But then a clear asymmetry emerges: retail prices rise more quickly following cost increases than they fall following cost decreases. This pattern is known as the “rockets and feathers” effect: prices rise like rockets but fall like feathers.

In our study, we examined the transmission from wholesale to retail prices over a period of more than ten years. As expected, when wholesale costs fell, pump prices dropped more slowly. This temporarily increased the gap between wholesale and retail prices – meaning retailers’ profits grew.

This pattern means if wholesale prices go up by ten pence per litre and then come back down, over the entire adjustment time motorists end up paying about a penny more per litre than they would if prices adjusted evenly.

But this varied across petrol stations. For some, there was very little additional cost to consumers. For others, it was up to five times larger, meaning that the same increase and subsequent decrease would cost consumers up to five pence per litre more.

Taken together, our findings point to a clear conclusion. Petrol retailers do not appear to profiteer during periods when oil prices are rising rapidly. If anything, their margins tend to be squeezed. If concerns about excess profits are warranted, the evidence suggests that it is more likely to occur when oil prices are falling than when they’re spiking.

– ref. Do petrol retailers really ‘price-gouge’ during oil price spikes? – https://theconversation.com/do-petrol-retailers-really-price-gouge-during-oil-price-spikes-278843

Source: The Conversation (Au and NZ) – By Monica Duffy Toft, Professor of International Politics and Director of the Center for Strategic Studies, The Fletcher School, Tufts University

Wars are rarely lost first on the battlefield. They are lost in leaders’ minds − when leaders misread what they and their adversaries can do, when their confidence substitutes for comprehension, and when the last war is mistaken for the next one.

The Trump administration’s miscalculation of Iran is not an anomaly. It is the latest entry in one of the oldest and most lethal traditions in international politics: the catastrophic gap between what leaders believe going in and what war actually delivers.

I’m a scholar of international security, civil wars and U.S. foreign policy, and author of the book “Dying by the Sword,” which examines why the United States repeatedly reaches for military solutions and why such interventions rarely produce durable peace. The deeper problem with the U.S. war in Iran, as I see it, was overconfidence bred by recent success.

Dismissed concerns

Before the conflict involving Iran, Israel and the U.S. escalated, Energy Secretary Chris Wright dismissed concerns about oil market disruption, noting that prices had barely moved during the 12-day war in June 2025 between Israel and Iran. Other senior officials agreed.

What followed was significant: Iranian-aimed missile and drone barrages against U.S. bases, Arab capitals and Israeli population centers. Then Iran effectively closed the Strait of Hormuz, through which roughly 20% of the world’s oil supply passes daily − not with a naval blockade, not with mines or massed anti-ship missiles, but with cheap drones.

A few strikes in the vicinity of the strait were enough. Insurers and shipping companies decided the transit was unsafe. Tanker traffic dropped to zero, although the occasional ship has made it through recently. Analysts are calling it the biggest energy crisis since the 1970s oil embargo.

Iran’s new supreme leader, Mojtaba Khamenei, has since vowed to keep the strait closed. U.S. Sen. Chris Murphy, a Connecticut Democrat, reported after a closed-door briefing that the administration had no plan for the strait and did not know how to get it safely back open.

With no embassy in Tehran since 1979, the U.S. relies heavily for intelligence on CIA networks of questionable quality and Israeli assets who have their own country’s interests in mind. So the U.S. did not anticipate that Iran had rebuilt and dispersed significant military capacity since June 2025, nor that it would strike neighbors across the region, including Azerbaijan, widening the conflict well beyond the Persian Gulf.

The war has since reached the Indian Ocean, where a U.S. submarine sank an Iranian frigate 2,000 miles from the theater of war, off the coast of Sri Lanka – just days after the ship had participated in Indian navy exercises alongside 74 nations, including the U.S.

The diplomatic damage to Washington’s relationships with India and Sri Lanka, two countries whose cooperation is increasingly important as the United States seeks partners to manage and mitigate Iran’s blockade, was entirely foreseeable. Washington has put them in a difficult position, with India choosing diplomacy with Iran to secure passage for its vessels and Sri Lanka opting to retain its neutrality, underscoring its vulnerable position.

But U.S. planners didn’t foresee any of this.

The wrong lesson from Venezuela

The swift military intervention by the U.S. in Venezuela in January 2026 produced rapid results with minimal blowback − appearing to validate the administration’s faith in coercive action.

But clean victories are dangerous teachers.

They inflate what I call in my teaching the “hubris/humility index” − the more a leadership overestimates its own abilities, underestimates the adversary’s and dismisses uncertainty, the higher the score and the more likely disaster will ensue. Clean victories inflate the index precisely when skepticism is most needed, because they suggest the next adversary will be as manageable as the last.

Political scientist Robert Jervis demonstrated decades ago that misperceptions in international relations are not random but follow patterns. Leaders tend to project their own cost-benefit logic onto opponents who do not share it. They also fall into “availability bias,” allowing the most recent operation to stand in for the next.

The higher the hubris/humility index, the less likely there is to be the kind of strategic empathy that might ask: How does Tehran see this? What does a regime that believes its survival is at stake actually do? History shows that such a regime escalates, improvises and takes risks that appear irrational from an outside perspective but are entirely rational from within.

Recent cases reveal this unmistakable pattern.

The United States in Vietnam, 1965–1968

American war planners believed material superiority would force the communists in Hanoi to surrender.

It didn’t.

American firepower alone didn’t lead to military defeat, much less political control. The Tet Offensive in 1968 – when North Vietnamese and Viet Cong forces launched coordinated attacks across South Vietnam – shattered the official U.S. narrative that the war was nearly won and that there was “light at the end of the tunnel.”

Athough the U.S. and South Vietnamese forces ultimately repelled the attacks, their scale and surprise caused the public not to trust official statements, accelerating the erosion of public trust and decisively turning American opinion against the war.

The U.S. loss in Vietnam didn’t occur on a single battlefield, but through strategic and political unraveling. Despite overwhelming superiority, Washington was incapable of building a stable, legitimate South Vietnamese government or recognizing the grit and resilience of the North Vietnamese forces. Eventually, with mounting casualties and large-scale protests at home, U.S. forces withdrew, ceding control of Saigon to North Vietnamese forces in 1975.

The U.S. failure was conceptual and cultural, not informational. American analysts simply couldn’t picture the war from their opponent’s perspective.

Afghanistan: Deadly assumptions

The Soviet Union in Afghanistan in 1979 and the United States in Afghanistan after 2001 conducted two different wars but held the same deadly assumption: that external military force can quickly impose political order in a fractured society strongly resistant to foreign control.

In both cases, great powers believed their abilities would outweigh local complexities. In both cases, the war evolved faster − and lasted far longer − than their strategies could adapt.

Russia, Ukraine and the Strait of Hormuz

This is the case that should most haunt Washington.

Ukraine demonstrated that a materially weaker defender can impose huge costs on a stronger attacker through battlefield innovation: cheap drones, decentralized adaptation, real-time intelligence, and the creative use of terrain and chokepoints to find asymmetrical advantages. The U.S. watched it all unfold in real time for four years and helped pay for it.

Iran was also watching − and the Strait of Hormuz is the proof.

Iran didn’t need a navy to close the world’s most important energy chokepoint. It needed drones, the same cheap, asymmetric technology Ukraine has used to blunt Russia’s onslaught, deployed not on a land front but against the insurance calculus of the global shipping industry.

Washington, which had underwritten much of that playbook in Ukraine, apparently never asked the obvious question: What happens when the other side has been taking notes? That is not a failure of U.S. intelligence. It is a failure of strategic imagination − exactly what the hubris/humility index is designed to highlight.

Iran does not need to defeat the U.S. conventionally. It needs only to raise costs, exploit chokepoints and wait for a fracture among U.S. allies and domestic political opposition to force a fake U.S. declaration of victory or a genuine U.S. withdrawal.

Notably, Iran has kept the strait selectively open to Turkish, Indian and Saudi vessels, rewarding neutral countries and punishing U.S. allies, driving wedges through the coalition.

Historian Geoffrey Blainey famously argued that wars start when both sides hold incompatible beliefs about power and only end when reality forces those beliefs to align.

That alignment is now happening, at great cost, in the Persian Gulf and beyond. The Trump administration scored high on the hubris index at exactly the moment when it most needed humility.

– ref. Overconfidence is how wars are lost − lessons from Vietnam, Afghanistan and Ukraine for the war in Iran were ignored – https://theconversation.com/overconfidence-is-how-wars-are-lost-lessons-from-vietnam-afghanistan-and-ukraine-for-the-war-in-iran-were-ignored-278604

Source: The Conversation (Au and NZ) – By Renee Goreham, Associate Professor, Physics, University of Newcastle

I remember sitting very close to the television as a child and seeing the image was made up of tiny coloured dots, each of which broke down into miniature vertical strips of red, green and blue when I looked even closer.

Back then, a television was a bulky box that sat on its own stand. Today, screens are so thin they hang flat on the wall.

At the same time, picture quality seems to improve every few years. Manufacturers promise sharper resolution, brighter images and richer colours, with labels such as HD, 4K, 8K, OLED and QLED.

But can television images really keep improving forever? Or are we approaching the limits of what our eyes can actually see?

From bulky boxes to ultra-thin screens

Early televisions used cathode ray tube (CRT) technology. Inside the screen, beams of electrons swept rapidly across a phosphor coating, lighting up tiny points that formed the image.

The process happened so quickly that our eyes perceived a continuous picture. These televisions were bulky and deep, but for decades they were the standard way of watching TV.

In the early 2000s, flat-panel displays began replacing bulky CRT televisions. Liquid crystal displays (LCDs) allowed screens to become much thinner and lighter, and they made higher-resolution displays easier to produce. However, early LCD TVs often struggled with contrast, particularly when trying to display deep blacks.

A major step forward came with the development of efficient light-emitting diodes (LEDs) that produced blue light, a breakthrough recognised with the 2014 Nobel prize in physics.

Blue LEDs made it possible to create bright white LED light sources – by combining red, green and blue LEDs – which are used widely to backlight liquid LCDs. This allows more control of the amount of light passing through each pixel.

More recently, technologies such as OLEDs (organic light-emitting diodes) have improved picture quality further.

Unlike LCD screens, where a backlight shines through liquid crystals to create the image, OLED displays allow each pixel to produce its own light. Because individual pixels can be switched off completely, OLED screens can achieve deeper blacks, higher contrast and more vivid images.

The race for more pixels

Much of the marketing around televisions focuses on resolution or the number of pixels that make up the image. Standard definition television contained only a few hundred lines of pixels.

High definition (HD) increased this dramatically. Then came 4K, which contains roughly four times as many pixels as HD.

Now manufacturers are promoting 8K displays with even more detail. But resolution alone does not determine how good a picture looks.

At typical viewing distances in a living room, human eyesight limits our ability to distinguish individual pixels. For many of us, the difference between a 4K and 8K television may be difficult (or even impossible) to notice unless the screen is extremely large or viewed very closely.

Instead, other factors such as contrast, brightness, colour accuracy and motion handling often have a bigger impact on how realistic an image appears.

Tiny particles, better colours

Some of the biggest improvements in modern displays come from advances in materials science.

One example is quantum dots: tiny semiconductor particles only a few nanometres across. When light hits them, they emit very specific colours that depend on their size.

Smaller dots produce bluer light, while larger dots produce redder light. This size-dependent behaviour allows the colours to be tuned very precisely, improving the brightness and colour range of modern televisions.

Interestingly, the same materials are also used in scientific research. In my own work, we use quantum dots not to improve televisions, but to help detect biological targets.

Because these nanoparticles emit very bright and precise colours, they can act as tiny fluorescent labels that highlight disease markers or pathogens. The same nanoscale properties that make TV colours more vibrant can also help scientists see biological processes more clearly.

Are there limits to how good screens can get?

Even with these advances, displays cannot improve indefinitely. Human vision places some limits. Our eyes can only perceive a certain range of colours and brightness levels and only resolve a certain level of detail at a given distance.

One study found the average human eye can distinguish 94 pixels per degree of the visual field. In practice, that means you need to be less than two metres away from a 65-inch TV to detect any difference between a 4K screen and an 8K one.

Physics also plays a role. Screens cannot become infinitely bright without becoming uncomfortable, perhaps even unsafe, to watch.

And reproducing every colour the human eye can perceive is an enormous technical challenge. This means that while television technology will continue to improve, the most noticeable gains may no longer come from simply adding more pixels.

Instead, future advances may focus on better contrast, wider colour ranges, improved motion, and more immersive viewing experiences.

The future of television

Television displays have come a long way from the bulky CRT sets some of us remember. Advances in materials, nanotechnology and electronics have transformed how images are produced. But as screens approach the limits of what human vision can perceive, the race for ever-higher resolution may begin to slow.

The next big improvements in television may not come from adding more pixels, but from making the ones we already have look even more lifelike. After all, most of us no longer sit close enough to the screen to see those tiny red, green and blue lines anymore.

– ref. TVs keep getting more pixels – but we are approaching the limits of what our eyes can actually see – https://theconversation.com/tvs-keep-getting-more-pixels-but-we-are-approaching-the-limits-of-what-our-eyes-can-actually-see-277836

Source: The Conversation (Au and NZ) – By Sergey Alexeev, Senior research fellow, University of Sydney; UNSW Sydney

It’s a Friday evening and you pour a glass of wine while your teenager sits at the kitchen bench scrolling their phone. They barely look up. But they notice more than you think.

My new study found the drinking habits parents model at home carry over to their children.

The influence is strongest during a specific window: when children are aged 15 to 17. This is the stage when teens begin navigating social situations with alcohol and start deciding what “normal” drinking looks like.

It doesn’t mean you have to give up alcohol altogether. But there are behaviours you can tweak to improve the chance your children will have a healthy relationship with alcohol as they grow up.

Tracking influence over 23 years

My study used 23 years of nationally representative Australian data from the Household, Income and Labour Dynamics in Australia (HILDA) survey. This tracked more than 6,600 people over time, drawing on more than 43,000 observations.

To estimate parental influence, I linked each person’s drinking at a given age to their mother’s and father’s average drinking when that person was aged 12–18. I then compared how strongly those links showed up at different stages of life.

I found parental influence is strongest when children are aged 15–17, declines through the twenties, and rebounds at 28–37 for those who have become parents.

The effect runs mostly along same-sex lines. Mothers influence daughters most clearly, and fathers influence sons. There is no detectable father-to-daughter effect.

There is some crossover from mothers to sons, particularly during adolescence and again in the late twenties and thirties.

When adult children become parents themselves, they appear to revisit the drinking habits they grew up with. Daughters draw on their mothers’ examples; sons who become fathers begin to follow paternal patterns they had not previously adopted.

Genetics vs household norms

The evidence points more toward household norms than genetics. When I compared birth parents with non-birth parents – a broad category that includes step, adoptive, foster and other non-biological caregivers – the mother-to-daughter link held firm regardless of biological connection.

That suggests daughters are learning behaviour, not inheriting a fixed trait. For sons, the picture is more mixed, but the overall message is the same: what children observe matters.

None of this means a single glass of wine in front of your teen will do damage. The study measures repeated patterns of drinking over years, not one-off moments.

What appears to matter is the background signal: how often alcohol appears, how much, and what role it seems to play in everyday life. Is it the centrepiece of every celebration? The first response to a bad day? Or something that shows up occasionally, without fanfare?

How teens’ ideas about alcohol are shaped

My findings fit with broader evidence on how parents shape children’s drinking. A review of long-term (longitudinal) studies found parental modelling, limiting adolescents’ access to alcohol, monitoring, relationship quality and clear communication were all linked to lower levels of drinking in adulthood.

Another Australian study found parents’ heavy drinking episodes were associated with a higher likelihood teenagers had drunk alcohol. Children seem to learn not just whether adults drink, but what place alcohol has in ordinary family life.

Australian longitudinal research has also found parental supply of alcohol to teenagers – even with good intentions – is linked to heavier drinking and more alcohol-related problems down the track, rather than teaching children to drink responsibly.

The good news is broader trends are moving in the right direction. Far fewer Australian teens drink now than two decades ago. In 2001, about 70% of 14-to-17-year-olds had drunk alcohol in the previous year. By 2022–23, that figure was around 30%.

Similar declines have been documented across many high-income countries. The possible reasons include changing cultural attitudes, better education about risk and, as my study suggests, shifts in parental behaviour that flow down through families.

So what can parents actually do?

The practical goal is not perfection. It’s harm minimisation – shaping household norms so that alcohol is less central, less emotionally loaded and less available.

The evidence supports:

• keeping your own drinking moderate and low-key. Australian guidelines recommend no more than ten standard drinks a week for adults, and not drinking at all is the safest option for under-18s

• not supplying alcohol to teenagers, even with good intentions. Australian research suggests parental supply is linked to heavier drinking and more alcohol-related problems later on

• setting clear rules and having calm, consistent conversations about alcohol. In one longitudinal study, teens drank least when strict rules were paired with good-quality, regular communication

• being especially deliberate with your alcohol choices when your children are 15 to 17, because that is when family influence appears to bite hardest.

If your children are already adults, your example may still matter. My study found parental influence re-emerges when adult children start families of their own — particularly for daughters. The habits you modelled years ago can resurface when your grown-up children are deciding what kind of household they want to run.

Parents don’t control everything. Friends, stress and the broader social environment matter too. But what parents can shape is the background – the slow, steady signal about what alcohol is for and how much of it is normal.

– ref. Is it OK to drink in front of your kids? New research shows the age they’re most influenced – https://theconversation.com/is-it-ok-to-drink-in-front-of-your-kids-new-research-shows-the-age-theyre-most-influenced-278667

Source: The Conversation (Au and NZ) – By Angel Zhong, Professor of Finance, RMIT University

Every few months, someone in the superannuation industry declares that Australians now “need” around A$1 million to retire comfortably. It’s a big, scary number.

But consumer advocates say most people can retire with far less.

Independent estimates suggest something closer to $322,000 is enough for many retirees who own their own home. So who’s right – and what assumptions drive these wildly different targets?

It’s easy to put off thinking about superannuation when retirement is years away. In this five-part series, we ask top experts to explain how to sort your super in a few simple steps, avoid greenwashing, and set goals for retirement.

What the two key benchmarks say

Two key organisations publish retirement benchmarks in Australia, and they paint very different pictures.

The Association of Superannuation Funds of Australia (ASFA), the lobby group for the super industry, publishes two lifestyle options in its 2026 Retirement Standard. This was recently updated to reflect a higher cost of living:

Modest retirement: Covers the basics – a budget car, basic private health insurance, one domestic holiday a year. This costs around $35,503 a year for a single homeowner, and the age pension (the regular government payment available to eligible retirees aged 67+) covers most of it. You’d only need around $110,000 in super.

Comfortable retirement: Includes top-level private health insurance, a newer car, regular dining out, and overseas travel. ASFA puts this at around $54,240 a year for a single homeowner, requiring roughly $630,000 in super. For couples, it’s about $77,375 a year, needing around $730,000.

These are significant sums – but well below $1 million.

Then there’s Super Consumers Australia, an independent consumer group that recommends a substantially lower amount.

Rather than imagining a lifestyle, the consumer group uses actual Australian Bureau of Statistics data on what retirees really spend. Its headline finding: a typical single retiree spending at the middle level out of three options needs just $322,000 in super.

Remember, retirees don’t have work-related expenses and they also enjoy a range of discounts on things such as council rates, electricity and medicines, which can really add up.

Part of the difference is the industry body, ASFA, has an interest in encouraging people to contribute more to their super. Its “comfortable” standard is higher than most Australians’ standard of living while working.

Why the numbers differ

The gap comes down to what each benchmark is measuring.

ASFA describes an aspirational lifestyle. Super Consumers describes what real retirees actually spend.

The age pension does a lot of the heavy lifting either way. At Super Consumers’ medium spending level, about 67% of retirement income comes from the age pension, and the remainder from your super balance.

But here’s a crucial new factor: the age pension isn’t keeping up with what retirees actually spend money on.

While the pension is indexed to inflation, retirees’ major expenses – insurance, rates, utilities, health care and food – have been rising faster than general consumer prices.

That means retirees who rely heavily on the pension are seeing more financial pressure than the headline inflation numbers suggest.

There’s a housing catch

Here’s the crucial fine print: every one of these benchmarks assumes you own your home outright when you retire.

That assumption is becoming shaky. Research shows the share of Australians aged 55–64 still carrying mortgage debt has tripled since 1990, and the average debt for that age group now exceeds $230,000. More than one in three Millennials expect to retire with a mortgage still running.

The ASFA budgets are built on the assumption of full home ownership. That means they do not include rent, mortgage repayments or major housing costs.

If you’re renting or carrying a mortgage into retirement, the required super balance can rise dramatically. ASFA estimates renters need $340,000–385,000 for a modest lifestyle – more than a homeowner needs for a comfortable one.

Super Consumers Australia presents a similar gap, estimating that a renter requires about $659,000 in superannuation, compared with only $322,000 for a homeowner.

With more people retiring with mortgage debt today than previous generations, both key benchmarks may underestimate housing-related stress for future retirees.

The gender gap in retirement

Retirement targets are often discussed as if everyone starts from the same position. They don’t.

Australian women retire with about 25% less super than men. The gender pay gap (currently around 21%) compounds over a working life into a much larger retirement savings gap. Women also live longer on average, meaning their money needs to stretch further.

The government began paying super on parental leave in July 2025 – a meaningful step forward. But the gap remains significant.

What this means for you

There’s no single right number. But ask yourself these questions before chasing any benchmark:

• will you own your home outright?
• do you want to travel or are you a homebody?
• are you planning for one income or two?

The gap between ASFA comfortable and Super Consumers medium is $8,497 a year in spending – but nearly $308,000 in required super. That difference is almost entirely lifestyle choice.

For a personalised estimate, the free MoneySmart Retirement Planner is a good starting point, or call the government’s free Financial Information Service on 132 300.

The $1 million figure isn’t evidence-based for most Australians. But the lower benchmarks all carry the same caveat: they assume you’re a homeowner. As more people retire with debt or as renters, even those more modest numbers may understate what you actually need.

Disclaimer: This article provides general information only and is not intended as financial advice.

– ref. How much do you really need to retire? It’s probably a lot less than $1 million – https://theconversation.com/how-much-do-you-really-need-to-retire-its-probably-a-lot-less-than-1-million-276375

Source: The Conversation (Au and NZ) – By Paula Lorgelly, Professor of Health Economics, University of Auckland, Waipapa Taumata Rau

This time six years ago, as officials prepared to move New Zealand into lockdown, the public was suddenly introduced to the complex and somewhat bewildering world of pandemic modelling.

These highly mathematical models mapped out how COVID-19 might spread, projecting potential infections, hospitalisations and deaths in stark scenarios, sometimes making for alarming news headlines.

But the models told us much less about the wider economic and social consequences of the decisions being made in those crisis weeks and months.

As the recently released Phase Two report of the Royal Commission on COVID-19 concluded, New Zealand’s pandemic response was ultimately effective.

Yet it also acknowledged the significant social and economic impacts of both the virus and the response – including strains on trust and social cohesion – and singled out areas for improvement.

That included the need for better modelling, data and frameworks to help decision-makers weigh those society-wide impacts more effectively.

Notably, the inquiry recommended a new strategic function – based in either the Treasury or the Department of the Prime Minister and Cabinet – to help guide New Zealand through the next pandemic.

It might be asked how such a function could sit outside the Ministry of Health, which was central to the initial COVID-19 response.

But in these emergencies, decision-makers shouldn’t be choosing between health and the economy. Pandemics affect both – and understanding the full impacts requires all perspectives.

If the inquiry made one thing resoundingly clear, it is that in the next pandemic – and there will be one – leaders must be able to make faster, better informed calls.

So, what if, next time, New Zealand had the modelling tools it needed to weigh up all the potential impacts at once?

A new pandemic model?

The British statistician George Box famously said that “all models are wrong, but some are useful”.

In essence, models may never perfectly capture the complexities of the real world, but they remain valuable tools for decision-making and prediction.

Pandemic responses are largely informed by epidemiological models, which help us understand how disease spreads through populations, and which simulate the impact of different control measures. But they also have limitations, including the difficulty of capturing human behaviour.

During COVID-19 and earlier events, these models were typically developed separately from economic modelling, each serving distinct purposes. Increasingly, however, there is recognition that integrating these approaches can provide a more complete picture.

Reflecting this shift, the UK-based Institute for Government has examined the benefits and challenges of so-called “epi-econ” models, which combine epidemiological and economic analysis.

Rather than simply assessing cost-effectiveness, these more sophisticated models aim to capture both the public health and broader economic impacts of policies such as lockdowns, border closures, vaccination programmes and wage subsidies.

They can simulate how such measures may affect infections, hospitalisations and deaths, as well as their possible impacts on inflation, employment and economic growth.

Read more: Second COVID inquiry: why being politically prepared for the next pandemic is crucial

They can also take both short and long-term perspectives, capturing the complex, dynamic interactions between public health crises, the interventions used to manage them and their broader macroeconomic consequences.

Such modelling can take several forms. Computable general equilibrium (CGE) models, for instance, simulate how households, firms, government and markets interact across the economy.

They have been recently used to assess the impacts of tariffs introduced by US President Donald Trump; earlier, they helped New Zealand’s government model the economic impacts of fuel outage scenarios.

To be useful in a pandemic, however, these models need to be “dynamic” – or able to track how conditions and intersections within the model evolve over time rather than offering a static snapshot.

Dynamic CGE models have been used to assess impacts on sectors such as food security and tourism, as well as wider economic effects across countries.

They are increasingly incorporating factors such as behaviour and resilience, helping understand the impacts of businesses closing and reopening, how illness affects the workforce, and how fiscal stimulus can assist recovery.

Building better tools for the next crisis

The COVID-19 response showed us that when epidemiological and economic models are developed separately – and often with different assumptions – decision-makers are left weighing conflicting advice.

An integrated epi-econ approach makes those trade-offs explicit and better informs the choices policymakers must make.

They can also reveal hidden non-linear effects – for example, how shutting non-essential activity can ripple through to workers in essential services. They can incorporate feedback loops too, showing how economic support measures during downturns might influence the spread of infections.

In the UK, early pandemic policies showed how such incentives could unintentionally accelerate spread.

Importantly, the inquiry also recommended that future governments ensure greater transparency and communication around decisions, including the science and evidence underpinning them.

The UK Institute for Government argues that combined epi-econ models should be accessible across ministries and departments, ensuring decision-makers are working from a shared evidence base.

Risk and uncertainty are inherent in these models. If New Zealand adopts them, ministers need to be able to understand the models’ inbuilt uncertainties and clearly communicate these to the public.

The Ministry of Health is leading the response to the Royal Commission’s final report and will advise the government on how to act on its recommendations.

In doing so, there is an opportunity to look beyond traditional epidemiological models and consider the added value of integrated approaches.

New Zealand will need stronger collaboration across government, academia and industry so that when the next crisis comes decisions are better informed from the outset.

– ref. In the next pandemic, NZ doesn’t need to choose between health and the economy – https://theconversation.com/in-the-next-pandemic-nz-doesnt-need-to-choose-between-health-and-the-economy-277951