Kategori: News in English

The economic cost of food waste in Australia is staggering. It’s estimated $36.6 billion is lost to the economy every year. Much of our fresh produce never even makes it to stores, rejected at the farm gate due to cosmetic reasons, such as its appearance, size or ripeness.

We’ve known about this problem for a long time, which has given rise to the “ugly” food movement. Once-rejected produce has been rebranded as “wonky” in the UK, “inglorious” in France, “naturally imperfect” in Canada or an “odd bunch” in Australia.

While the existence of these campaigns is commendable, there’s another major marketing challenge if we want to reduce food waste – acceptance of climate-affected produce.

Broadly speaking, this refers to produce affected by extreme or moderate weather events. Droughts are an example of such climate events, predicted to become more intense and frequent as a result of global climate change.

Climate-affected produce resembles “ugly” food as it is often smaller, misshapen or has surface imperfections.

Climate-affected produce often has a lot in common with ‘ugly’ fruit, but may also differ in taste and texture.
Alexey Borodin/Shutterstock

But in contrast to “ugly food”, the taste and texture of climate-affected produce can be quite different.

Under the effects of drought, apples may become sweeter and more granular, chillies hotter and onions more pungent. In the case of mild or moderate droughts, such produce is still edible.

Our recent research points to some uncomfortable truths. Many consumers prefer to avoid climate-affected produce altogether. And when price is a factor, they won’t choose it without a discount.

But our research also offers suggestions on how purchases of such produce could be encouraged – including marketing messages that highlight the “resilience” of climate-affected produce.

Our research

We carried out two discrete choice experiments with consumers who buy fresh fruit and vegetables. One sample was drawn from among Australian students, the other from members of the wider Australian population.

Participants were shown eight different apple options simulating a shopping environment, which were described with a range of different attributes including firmness, sweetness, appearance and size.

The apples were also labelled with a price tag and information on whether they were sold at a supermarket or farmers’ market. All climate-affected apples were presented with a “resilience” message: “resilient apple – survived the drought”.

We sought to examine how produce’s “organoleptic” properties – the way it impacts our different senses – as well as levels of empathy toward the farmers impact consumers’ willingness to choose climate-affected produce, and how much they’d pay for it.

Drought can make apples sweeter, smaller, and less firm.
The Conversation, Natthapol Siridech/Shutterstock, PickPik

A preference for perfect

We found when an apple’s firmness, size and aesthetics were important and empathy towards farmers was low, consumers tended to avoid climate-affected produce. They instead chose unaffected alternatives at higher prices (no such effect was observed for sweetness).

This finding might not be surprising, but it’s still cause for concern. If farmers cannot repurpose climate-affected produce into spreads, jams, smoothies or animal feed, it can’t enter supply chains and may end up as waste.

Previous campaigns for “ugly” fruit and vegetables may not offer much help with this problem, either. These campaigns emphasise the unaffected taste and texture of the produce. Marketing climate-affected produce needs a different approach.

Otherwise, we expect a discount

When price was important to consumers, they chose climate-affected produce, regardless of their levels of empathy toward farmers. But they were only willing to pay discounted prices for it.

That might seem like a more positive outcome. But consumer expectations that climate-affected produce will always be discounted may disadvantage farmers with lower profit margins and diminish its value as a still-usable resource.

Getting climate-affected (but still edible) produce into supply chains can help reduce food waste.
Ekaterina Pokrovsky/Shutterstock

The power of “resilience” messaging

Importantly, we found when the “resilience” message resonated with consumers, they were more inclined to consider climate-affected apples. This was true even when their empathy towards farmers was low.

This suggests that when empathy fails, leveraging marketing messages that highlight “resilience” could be another avenue worth exploring.

Our research team is now exploring what types of “resilience” messages can encourage purchases of climate-affected produce.

Australians have been conditioned for many years to expect only aesthetically pleasing fruit and vegetables.

Given extreme weather events are unlikely to become less frequent in the future, climate-affected produce is likely here to stay. If we want consumers to embrace it, we need to have uncomfortable conversations around its different taste and texture, and rethink what we’re willing to accept. Läs mer…

Opposition Leader Peter Dutton has doubled down on his commitment to sack thousands of public servants if he’s elected prime minister.

Dutton has again highlighted the “wasteful” 36,000 increase in public service jobs under Labor, which he says has made the Australian Public service “bloated and inefficient”.

While there is considerable political hyperbole and Trumpian allusions in Dutton’s statements, there are areas where legitimate savings could be made by whoever wins the coming election. That includes a second-term Albanese government, which would need to find efficiencies to offset promised wage increases.

Dutton’s commitment

Dutton unrealistically mentioned A$24 billion in potential savings over four years by reversing the growth in the number of new public service jobs since the last election.

Dutton’s claim of 36,000 extra bureaucrats under the Albanese government is broadly correct. The latest State of the Service Report shows the ongoing workforce increased from 133,976 in June 2021 to 170,186 in June last year. This was offset by a reduction of around 4,000 non-ongoing employees.

Labor has reduced the use of consultants and contractors, though at best those savings only partially offset the costs of the public service expansion.

Reversing the net increase in costs in the next term of Parliament, however, will not be easy and could not be done immediately.

Peter Dutton is flagging thousands of public service job cuts.
Steven Markham/AAP

In turn, Labor is hiring fewer consultants and contractors. Those numbers could rise again if permanent positions are axed under a Coalition government.

Dutton is careful not to make any specific commitments regarding the number of jobs that would go nor the dollar savings involved. However, he and his shadow ministers have repeatedly referred to the 36,000 new positions under Labor.

While the Coalition won’t be detailing any spending cuts until after the election, Dutton has alluded to US President Donald Trump’s playbook by targeting “culture, diversity and inclusion advisers”.

Dutton contends these roles add to costs while providing little public service:

Such positions, as I say, do nothing to improve the lives of everyday Australians.

Putting the public service growth into context

Despite Dutton’s combative language, the growth of the Australian Public Service is not nearly as dramatic as he claims, nor is it concentrated in Canberra.

The State of the Service Report shows the Australian Public Service headcount is lower now (0.68%) as a percentage of the Australian population than it was in 2008 (0.75%). It is also a smaller share of the overall Australian workforce (1.36% compared to 1.52%).

Despite Dutton’s often repeated claim that all of the additional public servants are based in Canberra, the proportion of the public service working in the capital has decreased to just 36.9%.

The numbers back up the government’s claim that the expanded bureaucracy has delivered improvements to critical public services such as the National Disability Insurance Scheme, Veterans’ Affairs and Centrelink outside of Canberra.

Labor has also committed to savings

Despite its defence of the public service, a re-elected Labor government would also need to find efficiencies.

The Australian Financial Review has drawn attention to the mid-year budget update, which forecast no growth in the public service wages bill from 2025–26 to 2027–28. This is despite an enterprise bargaining agreement to increase wages by 11.2% over the three years to March 2026.

Finance and Public Service Minister Katy Gallagher has dismissed the Coalition’s claims of a $7.4 billion black hole. She says Labor’s forecasting method is the same as the one the Liberals used in government

And the minister has restated Labor’s commitment to finding its own savings through the 1% efficiency dividend, which she says is “largely a good thing”.

In other words, the Albanese government is assuming pay increases will be offset by efficiency measures over the next three years. That will require some effort.

Public Service Minister Katy Gallagher is facing a wages ‘blackhole’ across the APS.
Mick Tsikas/AAP

Where savings could actually be made

Regardless of who forms the next government, there are savings to be made across the public service, which has become too top heavy.

Remuneration is a mess, with extraordinary variations in pay, particularly among the senior executive level.

A wholesale change in the membership of the Remuneration Tribunal, which sets public service pay levels, and a review of its methodology are much needed.

There should also be more emphasis on skills and capability, and less on diversity. A strong business case exists to maximise the talent pool the public service draws on, but care is needed to not compromise the merit principle in the pursuit of equity.

Dutton’s plan raises legitimate concerns

Dutton’s populist rhetoric about the public service raises legitimate concerns beyond the potential job cuts.

There’s a real risk the Coalition will resurrect its ideological preference for the private sector, with its associated extra costs and conflicts of interest.

Nor is there any clear commitment to avoiding a return to the politicisation of the bureaucracy evident under former prime minister Scott Morrison, which contributed to the Robodebt scandal.

The Albanese government has sadly dropped the ball by failing to legislate to promote merit-based appointments, leaving open opportunities for politically based hirings and firings.

With election day fast approaching, voters may reasonably be wary of both sides of politics when it comes to the independence and performance of the public service. Läs mer…

Until a few weeks ago, few people in the Western world had heard of a small Chinese artificial intelligence (AI) company known as DeepSeek. But on January 20, it captured global attention when it released a new AI model called R1.

R1 is a “reasoning” model, meaning it works through tasks step by step and details its working process to a user. It is a more advanced version of DeepSeek’s V3 model, which was released in December. DeepSeek’s new offering is almost as powerful as rival company OpenAI’s most advanced AI model o1, but at a fraction of the cost.

Within days, DeepSeek’s app surpassed ChatGPT in new downloads and set stock prices of tech companies in the United States tumbling. It also led OpenAI to claim that its Chinese rival had effectively pilfered some of the crown jewels from OpenAI’s models to build its own.

In a statement to the New York Times, the company said:

We are aware of and reviewing indications that DeepSeek may have inappropriately distilled our models, and will share information as we know more. We take aggressive, proactive countermeasures to protect our technology and will continue working closely with the US government to protect the most capable models being built here.

The Conversation approached DeepSeek for comment, but it did not respond.

But even if DeepSeek copied – or, in scientific parlance, “distilled” – at least some of ChatGPT to build R1, it’s worth remembering that OpenAI also stands accused of disrespecting intellectual property while developing its models.

What is distillation?

Model distillation is a common machine learning technique in which a smaller “student model” is trained on predictions of a larger and more complex “teacher model”.

When completed, the student may be nearly as good as the teacher but will represent the teacher’s knowledge more effectively and compactly.

To do so, it is not necessary to access the inner workings of the teacher. All one needs to pull off this trick is to ask the teacher model enough questions to train the student.

This is what OpenAI claims DeepSeek has done: queried OpenAI’s o1 at a massive scale and used the observed outputs to train DeepSeek’s own, more efficient models.

OpenAI claims that DeepSeek copied at least some of ChatGPT to build its own AI models.
Salvatore Di Nolfi/EPA

A fraction of the resources

DeepSeek claims that both the training and usage of R1 required only a fraction of the resources needed to develop their competitors’ best models.

There are reasons to be sceptical of some of the company’s marketing hype – for example, a new independent report suggests the hardware spend on R1 was as high as US$500 million. But even so, DeepSeek was still built very quickly and efficiently compared with rival models.

This might be because DeepSeek distilled OpenAI’s output. However, there is currently no method to prove this conclusively. One method that is in the early stages of development is watermarking AI outputs. This adds invisible patterns to the outputs, similar to those applied to copyrighted images. There are various ways to do this in theory, but none is effective or efficient enough to have made it into practice.

There are other reasons that help explain DeepSeek’s success, such as the company’s deep and challenging technical work.

The technical advances made by DeepSeek included taking advantage of less powerful but cheaper AI chips (also called graphical processing units, or GPUs).

DeepSeek had no choice but to adapt after the US has banned firms from exporting the most powerful AI chips to China.

While Western AI companies can buy these powerful units, the export ban forced Chinese companies to innovate to make the best use of cheaper alternatives.

The US has banned the export of the most powerful computer chips to China.
Nor Gal/Shutterstock

A series of lawsuits

OpenAI’s terms of use explicitly state nobody may use its AI models to develop competing products. However, its own models are trained on massive datasets scraped from the web. These datasets contained a substantial amount of copyrighted material, which OpenAI says it is entitled to use on the basis of “fair use”:

Training AI models using publicly available internet materials is fair use, as supported by long-standing and widely accepted precedents. We view this principle as fair to creators, necessary for innovators, and critical for US competitiveness.

This argument will be tested in court. Newspapers, musicians, authors and other creatives have filed a series of lawsuits against OpenAI on the grounds of copyright infringement.

Of course, this is quite distinct to what OpenAI accuses DeepSeek of doing. Nevertheless OpenAI isn’t attracting much sympathy for its claim that DeepSeek illegitimately harvested its model output.

The war of words and lawsuits is an artefact of how the rapid advance of AI has outpaced the development of clear legal rules for the industry. And while these recent events might reduce the power of AI incumbents, much hinges on the outcome of the various ongoing legal disputes.

Shaking up the global conversation

DeepSeek has shown it is possible to develop state-of-the-art models cheaply and efficiently. Whether they can compete with OpenAI on a level playing field remains to be seen.

Over the weekend, OpenAI attempted to demonstrate its supremacy by publicly releasing its most advanced consumer model, o3-mini.

OpenAI claims this model substantially outperforms even its own previous market-leading version, o1, and is the “most cost-efficient model in our reasoning series”.

These developments herald an era of increased choice for consumers, with a diversity of AI models on the market. This is good news for users: competitive pressures will make models cheaper to use.

And the benefits extend further.

Training and using these models places a massive strain on global energy consumption. As these models become more ubiquitous, we all benefit from improvements to their efficiency.

DeepSeek’s rise certainly marks new territory for building models more cheaply and efficiently. Perhaps it will also shake up the global conversation on how AI companies should collect and use their training data. Läs mer…

The website for the United States Agency for International Development (USAid), the world’s biggest aid donor, has gone dark.

Donald Trump’s new administration plans to place the autonomous agency under the control of the state department. The secretary of state, Marco Rubio, has now declared himself as head of the agency to “align” it with Trump’s priorities.

Several days ago, on January 26, Rubio said: “Every dollar we spend, every programme we fund, and every policy we pursue must be justified with the answer to three simple questions: Does it make America safer? Does it make America stronger? Does it make America more prosperous?”

But the decision to freeze USAid, which is part of Trump’s policy to put “America first”, places everyone at risk. Organisations that provide vital care for vulnerable people around the world are being forced to halt operations. The boss of one such organisation said: “People will die.”

Elon Musk, the world’s richest man and a close adviser to Trump, is playing an active role in the destruction of USAid. He has claimed – without providing any evidence – that the agency is “beyond repair”. “It needs to die,” Musk wrote on X.

Musk, who leads the newly formed Department of Government Efficiency (Doge), is gearing to cut trillions of dollars from the US budget. However, by seeing cuts to USAid as a solution, Trump and Musk are catering to an audience that has a fundamental misunderstanding about US foreign aid more generally.

Surveys demonstrate that Americans believe 25% of the federal budget is spent on foreign aid. In reality, the US gives about 0.2% of its gross national product (GNP), the total value of goods and services produced by a country, to foreign aid – or less than 1% of its federal budget. This is far below the UN target of 0.7% of GNP.

But, despite this, USAid provided 42% of all humanitarian aid globally in 2024. This included about US$72 billion (£58 billion) in aid in a wide range of areas, from helping people access clean water, sanitation, healthcare and energy, to providing disaster relief, shelter and food.

USAid also delivered programmes aimed at supporting democracy, civil society, economic development and landmine clearance in war zones, as well as working to prevent organised crime, terrorism and conflict. The gutting of USAid will have a profound impact on human security.

A protester holds a placard outside the USAid headquarters on February 3.
Will Oliver / EPA

The Trump administration has granted a waiver for the continuation of “life-saving humanitarian assistance”. This includes a programme that helps 20 million people living with HIV/Aids access anti-retroviral drugs. But there are questions about the future of US Aids organisation, the President’s Emergency Plan for Aids Relief (Pepfar).

To date, over 43 million people worldwide have died from Aids. But one of the biggest success stories of the George W. Bush administration was its launch of Pepfar in 2003. The World Health Organization says that Pepfar, working in partnership with USAid, has saved 26 million lives.

Pepfar employs more than 250,000 doctors, nurses and other staff across 55 countries. One of the functions that USAid performs is ordering and procuring the drugs used by Pepfar to keep the millions infected with HIV alive. It remains to be seen whether federal payments to USAid’s locally run partner organisations will be stopped.

We are, in any case, likely to see an uptick in other infectious diseases. USAid had been working to prevent current outbreaks of mpox and Marburg virus from spreading beyond Africa. It is not clear what the future is for these programmes.

And USAid’s work with malaria, a disease that kills about 450,000 children under the age of five each year, is facing uncertainty. From 2000 to 2021, USAid’s work helped to prevent 7.6 million deaths from malaria. Also in doubt is USAid’s work to develop and implement the malaria vaccine, which was considered a gamechanger for combating the disease.

At the same time, USAid responds to an average of 65 natural disasters each year. In 2024 alone, it responded to 84 separate crises across 66 different countries. The government is letting go all of the staff important for implementing these types of programmes.

Dozens of senior USAid officials have been placed on leave, while contractors working on the agency’s programmes have been furloughed. Up to 3,000 aid workers in Washington DC could reportedly be laid off this week.

What Trump’s team misunderstand is that the work of USAid is also vital for preserving American interests. China, which has poured more than US$1 trillion of assistance into infrastructure projects in Asia, Africa, Europe and Latin America since 2013, will now be given an opportunity to exert more influence around the world. The void in US aid is a gift for China in the battle for soft power.

White House press secretary, Karoline Leavitt, lists some of what she calls the ‘insane priorities’ that USAid has been spending money on.

Global aid sector in disarray

Foreign aid relies on certainty and transparency about the future of aid programmes. But the Trump administration has offered little clarity while US foreign aid programmes are all being reviewed. One aid organisation referred to the situation as an “absolute dumpster fire” due to the uncertainty.

There have already been reports of total confusion in health clinics previously supported by USAid, which were shut down without warning. Africa will probably be the region most negatively affected. Local workers in healthcare-related projects on the continent will lose their jobs, while nurses, doctors and healthcare workers across clinics will be unable to continue their vital work.

The Democrats have claimed that Trump does not have the legal authority to eradicate a congressionally funded independent agency. They have said court challenges are already in motion and have pledged to try to block approval of Trump’s state department nominations until the shutdown is reversed.

Trump did try to cut US foreign aid during his first term, but Congress refused. He then tried – and ultimately failed – to freeze the flow of aid appropriated by Congress. This time, Trump is not bothering to play by the rules. Läs mer…

Toxic substances used in flea and tick treatments pet owners give to their dogs and cats have been detected in birds’ nests, according to new UK research.

Fipronil and imidacloprid, two common insecticides distributed by vets in pipettes to kill or stunt fleas, were previously found in nearly all English rivers. These chemicals are known to harm aquatic insects, which has repercussions for the species that eat them. Birds are also known to ingest these insecticides in their food and water.

Our study now raises the risk of direct skin contact, as veterinary drugs were the most common insecticides colleagues and I found in bird nests. How do they get there – and what are they doing?

Many birds, including garden visitors such as blue tits and great tits, nest in tree hollows and nest boxes. To keep their eggs and chicks warm, these birds line their nests with soft materials such as fur. In fact, around 74% of European bird species use fur as nest insulation.

Across the UK, pet owners and wildlife enthusiasts leave brushed pet fur outside for birds to collect. But with around 80% of the country’s 22 million cats and dogs receiving regular flea and tick treatments, this well-meaning act can inadvertently expose birds to harm.

A previous study in the Netherlands found that insecticides used in flea treatment were appearing in birds’ nests. The study I led with colleagues is the first to identify the problem in the UK.

Banned on farms, used in homes

We examined 103 nests of blue tits and great tits and found the residue of 17 out of 20 insecticides commonly used as flea treatments in the UK. The most prevalent were fipronil, which we found in every single nest, imidacloprid and permethrin, which were both detected in 89% of nests.

All three of these chemicals are banned for use as pest control on EU farms due to their harmful effects on wildlife. Studies have shown that these insecticides can damage the nervous and reproductive systems of birds, and threaten their overall health. Yet they remain widely used in veterinary medicine.

We collected nests months after the breeding season, and so the concentrations of chemicals we found are likely to be lower than what was present in the nests during spring, when birds gather material for nests. This suggests that eggs and chicks were exposed during the whole breeding season.

The nests we found with higher concentrations of insecticides contained more unhatched eggs and dead chicks. Other factors could explain these deaths, such as predation. But the known dangers of these chemicals should make us question their wider impact on the environment. While more research is needed to fully understand their risks, the evidence already suggests that exposure could be harming nestlings, which are at a critical stage of development.

Flea and tick treatments either kill insects or halt their development.
Nick Alias/Shutterstock

Scientists and conservation groups are urging the UK government to conduct a more thorough environmental risk assessment of veterinary treatments, particularly those used on dogs and cats.

Public awareness will also be key to addressing the problem. Many pet owners do not know that their routine flea treatments affect wildlife. Small changes could help reduce this impact. For example, year-round flea treatment is not necessary, particularly in winter when fleas and ticks are less active.

If treatment is required then tablets could be a better choice as they do not involve direct skin contact for birds and would not wash away every time a pet swims or is bathed either. They may be excreted in faeces and contaminate the soil, however – that’s why a thorough environmental risk assessment is necessary.

Pet owners who enjoy helping birds can still leave out fur as nesting material, perhaps by saving the brushed fur from untreated pets during winter and putting it out the following spring.

As awareness of this issue grows, pet owners, scientists and policymakers can ensure that veterinary treatments do not come at the cost of the UK’s wildlife.

Don’t have time to read about climate change as much as you’d like?
Get a weekly roundup in your inbox instead. Every Wednesday, The Conversation’s environment editor writes Imagine, a short email that goes a little deeper into just one climate issue. Join the 40,000+ readers who’ve subscribed so far. Läs mer…

China’s new DeepSeek Large Language Model (LLM) has disrupted the US-dominated market, offering a relatively high-performance chatbot model at significantly lower cost.

The reduced cost of development and lower subscription prices compared with US AI tools contributed to American chip maker Nvidia losing US$600 billion (£480 billion) in market value over one day. Nvidia makes the computer chips used to train the majority of LLMs, the underlying technology used in ChatGPT and other AI chatbots. DeepSeek uses cheaper Nvidia H800 chips over the more expensive state-of-the-art versions.

ChatGPT developer OpenAI reportedly spent somewhere between US$100 million and US$1 billion on the development of a very recent version of its product called o1. In contrast, DeepSeek accomplished its training in just two months at a cost of US$5.6 million using a series of clever innovations.

But just how well does DeepSeek’s AI chatbot, R1, compare with other, similar AI tools on performance?

DeepSeek claims its models perform comparably to OpenAI’s offerings, even exceeding the o1 model in certain benchmark tests. However, benchmarks that use Massive Multitask Language Understanding (MMLU) tests evaluate knowledge across multiple subjects using multiple choice questions. Many LLMs are trained and optimised for such tests, making them unreliable as true indicators of real-world performance.

An alternative methodology for the objective evaluation of LLMs uses a set of tests developed by researchers at Cardiff Metropolitan, Bristol and Cardiff universities – known collectively as the Knowledge Observation Group (KOG). These tests probe LLMs’ ability to mimic human language and knowledge through questions that require implicit human understanding to answer. The core tests are kept secret, to avoid LLM companies training their models for these tests.

KOG deployed public tests inspired by work by Colin Fraser, a data scientist at Meta, to evaluate DeepSeek against other LLMs. The following results were observed:

LLM Performance test.

The tests used to produce this table are “adversarial” in nature. In other words, they are designed to be “hard” and to test LLMs in way that are not sympathetic to how they are designed. This means the performance of these models in this test is likely to be different to their performance in mainstream benchmarking tests.

DeepSeek scored 5.5 out of 6, outperforming OpenAI’s o1 – its advanced reasoning (known as “chain-of-thought”) model – as well as ChatGPT-4o, the free version of ChatGPT. But Deepseek was marginally outperformed by Anthropic’s ClaudeAI and OpenAI’s o1 mini, both of which scored a perfect 6/6. It’s interesting that o1 underperformed against its “smaller” counterpart, o1 mini.

DeepThink R1 – a chain-of-thought AI tool made by DeepSeek – underperformed in comparison to DeepSeek with a score of 3.5.

This result shows how competitive DeepSeek’s chatbot already is, beating OpenAI’s flagship models. It is likely to spur further development for DeepSeek, which now has a strong foundation to build upon. However, the Chinese tech company does have one serious problem the other LLMs do not: censorship.

Censorship challenges

Despite its strong performance and popularity, DeepSeek has faced criticism over its responses to politically sensitive topics in China. For instance, prompts related to Tiananmen Square, Taiwan, Uyghur Muslims and democratic movements are met with the response: “Sorry, that is beyond my current scope.”

But this issue is not necessarily unique to DeepSeek, and the potential for political influence and censorship in LLMs more generally is a growing concern. The announcement of Donald Trump’s US$500 billion Stargate LLM project, involving OpenAI, Nvidia, Oracle, Microsoft, and Arm, also raises fears of political influence.

Additionally, Meta’s recent decision to abandon fact-checking on Facebook and Instagram suggests an increasing trend toward populism over truthfulness.

DeepSeek’s arrival has caused serious disruption to the LLM market. US companies such as OpenAI and Anthropic will be forced to innovate their products to maintain relevance and match its performance and cost.

DeepSeek’s success is already challenging the status quo, demonstrating that high-performance LLM models can be developed without billion-dollar budgets. It also highlights the risks of LLM censorship, the spread of misinformation, and why independent evaluations matter.

As LLMs become more deeply embedded in global politics and business, transparency and accountability will be essential to ensure that the future of LLMs is safe, useful and trustworthy. Läs mer…

You would think that Jacques Audiard’s 13-time Oscar-nominated Emilia Pérez was the most watched film of the year given the discussion it has generated. The Mexican-set, French-made film’s opening weekend in Mexico tells a different story.

Emilia Pérez sees the eponymous antagonist-heroine experience a transformation, undergoing gender-affirming procedures in order to leave behind her former dangerous, violent life as a cartel leader in Mexico.

It came eighth at the box office in Mexico, which is hardly surprising. The effects of narco violence saw 613 murders and 626 disappearances between September and December 2024 in Sinaloa State in northwestern Mexico as its eponymous cartel’s factions fight for territory.

Considering the context in which it was released, little positive noise has been made about Emilia Pérez within Mexico given its sensationalist, reductive representations of violence. Internationally, its representation of trans experiences has been criticised.

Though well acted, it is thoughtless. The luxurious life Emilia lives as a trans woman is far detached from reality of most trans people in Mexico, where the average life expectancy for a trans person is 35.

Looking for something good? Cut through the noise with a carefully curated selection of the latest releases, live events and exhibitions, straight to your inbox every fortnight, on Fridays. Sign up here.

We follow Rita Mora Castro (Zoe Saldaña), an underappreciated lawyer who works hard only for men to take the credit. Rita is hired by cartel head Juan “Manitas” del Monte (Karla Sofía Gascón) to find a surgeon for her transition to start again as Emilia Pérez. After the transition, Emilia has Manitas declared dead, leaving behind her mourning wife, Jessi (Selena Gómez) and their two young sons who she has relocated to Switzerland for their safety.

After four years, Emilia tracks down Rita to have Jessi and the children moved back to Mexico, posing as Manitas’ distant relative. Emilia then works with Rita to launch a non-profit, “La Lucecita”, that helps the families of missing persons after Emilia becomes appalled by how many disappeared people there are in Mexico.

Emilia’s immediate reaction to such social injustice demonstrates a naivety on Audiard’s part. Despite Manitas having destroyed lives, Emilia wants to dignify them. We are asked to believe that she had no idea about these wretched, miserable souls. But thankfully, Emilia’s “La Lucecita” is here to rescue them. The NGO will find the remains of the disappeared, making them visible again. Good thing Emilia made all that (drug) money to fund the work…

Trailer for Emilia Pérez.

The sheer unbelievability of Pérez not knowing about the violent reverberations of her work aside, I was gratified to see the disappeared of Mexico centralised in the film. The stories of Argentina, Chile, Uruguay, Guatemala and Colombia usually dominate when it comes to the consequences of human rights abuses in the region.

Political prisoners, state terrorism, death flights and extrajudicial murders date back at least as far as the 1960s in Mexico, with the Indomitable Memory Museum in Mexico City doing fantastic work to highlight this history and dignify victims. In particular, the story of the Ayotzinapa 43, who were disappeared en route to Mexico City for an annual march against state corruption and human rights abuses in 2014.

But, considering its direction, Emilia Pérez takes on a white saviour narrative and our heroine simply throws (drug) money at the problem. Audiard’s (admitted) lack of serious thought given to violence ,wealth and power in this context is laughable. Ask “searcher” groups, who go looking for the remains of their disappeared loved ones, like Las Rastredoras de El Fuerte to conjure up money for their work at a fancy gala (and watch I Called for You in Silence, a heartbreaking documentary on their struggles) and see what the reaction is.

Emilia Pérez had the chance to add some nuance to the violence in Mexico today, to demonstrate that this does not exist in a vacuum. It had a chance to go beyond what the transfeminist philosopher Sayak Valencia and the expert in feminist visual culture Sonia Herrera Sánchez would term a kind of sensationalist, colonialist “pornomisery” to present gender fluidity and sexuality in a troubled and troubling context.

I was disappointed. I found it impossible to watch the film without seeing constant instances of what Sayak Valencia deems gore capitalism in action. “Death has become the most profitable business in existence,” according to Valencia.

She outlines that in the era of drug war Mexico (2006 to the present) power is the new capital in a moment where hyper-masculinity and levels of violence are out of control. The lifeless body signifies a capital of fear and power.

Rather than Emilia Pérez forming any coherent commentary on this, the film contributes to it – how much will Audiard make from a film about bodies, what is done to them and how they are destroyed by Mexico’s drug war? How many awards? How much (more) power gained?

Zoe Saldaña sings “El Mal” from Emilia Pérez.

Bodily transition – from living to dead; from male to female – is a motif in the film, and one used as a lazy plot device. Emilia is no longer Manitas; in fact, she’s Manitas’ antithesis, who, therefore, does good for society. This dichotomy between “giving woman” and “violent man” only serves to perpetuate outdated views of womanhood. Karla Sofía Gascón was strong in this role, though I must ask why a Mexican trans actress couldn’t have played Pérez. For instance, Nava Mau of Baby Reindeer.

We know that Emilia Pérez isn’t that bothered about nuance, being one reason the film has been so ripe for satire. It is a narco-telenovela-cum-queer musical from the perspective of a 72-year-old white French man.

If you are looking for a show or film that does what Emilia Pérez should have, I can only recommend the one-off series Somos, a thoughtful take on the 2011 Allende massacre to temper such thoughtless representation. Läs mer…

In 2018, a colleague and I, together with a team of Greenlandic research assistants, conducted one of the most comprehensive surveys to date on public opinion in Greenland. We travelled to 13 randomly selected towns and settlements across the island nation, conducting in-person interviews with a representative sample of adult residents.

The survey explored a wide range of topics. We asked for views on climate change, economic matters – and the prospect of independence from Denmark. Until recently, this was the latest poll on what the people of Greenland thought about this issue.

Greenland, a former Danish colony, is currently an autonomous territory within the Kingdom of Denmark. This political arrangement grants Greenland extensive self-rule, including control over most domestic affairs, as well as its own prime minister and parliament. However, Denmark retains authority over foreign policy, defence and monetary policy.

While our survey results were covered in Greenlandic and Danish media upon their release, they received scant international attention. This changed abruptly on January 15, when newly re-elected US president Donald Trump reposted an old news article about our results. The headline stated that two-thirds of Greenlandic citizens support independence.

Trump posting the 2018 poll in 2025.
Truth Social

Trump did not add a comment in the post but the insinuation was clear given his recent statements about annexing Greenland from Denmark: Greenlandic residents want independence from Denmark, and therefore, they might be open to other political or economic arrangements with the US.

“I think we’re going to have it,” Trump recently said after a phone call with the Danish prime minister, Mette Frederiksen, who told him the land was “not for sale”. Trump has in the past spoken of somehow “purchasing” Greenland but has since moved on towards speaking in more assertive terms about taking control of the territory.

Back in 2018, when we conducted the survey, Trump had not yet revealed any plans to annex the island nation. It was a scenario we could hardly even have imagined and therefore did not ask our participants about. As such, regardless of how Trump framed them, the survey results in no way indicated that the population harboured a desire to join the US.

In fact, a recent survey conducted by Sermitsiaq (a Greenlandic newspaper) and Berlingske (a Danish newspaper) directly addressed this question and found that only 6% of respondents wanted Greenland to leave Denmark and instead become part of the US.

In the study I published based on the 2018 data collection, I reported that a majority of the Greenlandic population aspired to independence. Two-thirds of the participants thought that “Greenland should become an independent country at some point in the future”.

Opinions were more divergent regarding the timing of independence. When asked how they would vote in an independence referendum if it were held today, respondents who stated a preference were evenly split between “yes” and “no” to independence.

‘I think we’re going to have it’.
EPA

The Act on Greenland Self-Government, passed in 2009, grants the Greenlandic government the legal authority to unilaterally call a referendum on separating from the political union with Denmark. According to the law, “the decision regarding Greenland’s independence shall be taken by the people of Greenland”.

During the 15 years since its passage, the option to call a referendum has not been exercised. This is likely due to the potential economic consequences of leaving the union with Denmark.

Each year, Denmark sends a block grant that covers approximately half of Greenland’s budget. This supports a welfare system that is more extensive than what is available to most Americans. In addition, Denmark administers many costly public services, including national defence.

This backdrop presents a dilemma for many Greenlanders who aspire to independence, as they weigh welfare concerns against political sovereignty. This was also evident from my study, which revealed that economic considerations influence independence preferences.

For many Greenlanders, the island nation’s rich natural resources present a potential bridge between economic self-sufficiency and full sovereignty. Foreign investments and the associated tax revenues from resource extraction are seen as key to reducing economic dependence on Denmark. Presumably, these natural resources, which include rare earths and other strategic minerals, also help explain Trump’s interest in Greenland.

As Greenland’s future is likely to remain at the centre of a geopolitical power struggle for some time, it is crucial to remember that only Greenlanders have the right to determine their own path. What scarce information is available on their views suggests that while many aspire to independence, it is not driven by a desire to join the US. Läs mer…

The new Channel 4 programme Go Back to Where You Came From is unsettling viewing, almost unbearable at times. It takes six British citizens – some staunchly anti-immigration, others more open – and drops them into lives shaped by conflict and displacement.

The premise is to cultivate understanding of the refugee experience, to make the unimaginable tangible. But in doing so, the show risks turning forced displacement into spectacle, reducing suffering to an immersive learning experience for those with the privilege of ignorance.

The show opens with participants offering their views – filmed in their homes or standing at the cliffs of Dover, where one man declares: “What I’d do is, I’d set landmines up, and then any boat that comes within 50m of this beach, they’d get blown up.”

Then, two teams, two journeys. One is sent to Somalia, the other to Syria.

In Mogadishu, Nathan, Jess and Matilda navigate a city carved up by checkpoints, escorted by an American security team. Nathan surveys the streets like a man assessing a lost cause: a “shithole”, he mutters. Jess, fiercely anti-immigration, feels exposed – her fear magnified by the weight of unfamiliar eyes, the choreography of a life not her own. She wants to leave.

At a camp for internally displaced people, women speak of gender-based violence, of female genital mutilation, of lives spent in spaces never built for them. Jess listens, nods and files their words neatly into the folder of convictions she brought with her. She does not question; she confirms. The mindsets of Somalian men, she concludes, are the problem.

In Raqqa, Bushra, Chloe and Dave pick their way through streets reduced to rubble. Chloe complains about the rubbish, as if it were neglect rather than obliteration. “They should stay and clean it up,” she says. The children sifting through debris do not register. In a bombed-out home, a father speaks of safety, the only thing he wants for his children. The children do not speak.

The violence of ‘refuge’

Watching the show, I thought of the conversations I’ve had with asylum seekers and refugees on the island of Ireland as part of my research. Many speak of the quiet violence of exclusion – how “welcome” is so often a hollow gesture, how refuge can feel like another form of exile.

Many recount racial hatred in the streets, the fear woven into daily movements, the gnawing sense that they are barely tolerated, not wanted. Some have told me, with devastating clarity, that had they known what awaited them here – homelessness, suspicion, a life in bureaucratic limbo – they might never have fled at all. Not because home was safe, but because this isn’t living either.

These experiences are not anomalies. They are built into the asylum systems in the UK and Ireland, where deterrence is policy. As of mid-2024, 122.6 million people have been forcibly displaced worldwide, yet the UK hosts just 1% of them.

And “hosting” often takes the form of offshore detention, indefinite waiting and policies designed to make seeking refuge as inhospitable as possible. In Ireland, the failure is just as insidious: asylum seekers sleeping rough, vulnerability assessments in name only, the quiet withdrawal of care until people simply disappear from view.

Read more:
’When you get status the struggle doesn’t end’: what it’s like to be a new refugee in the UK

After the first episode of the Channel 4 show, I am left wondering: what is the point of each participant’s journey? The documentary trades in empathy – tracking transformation by how much the participants feel, learn and change. But empathy, when it stops at the self, is just another performance. It asks: how have I been altered? Instead of: what must I do with what I now know?

This is the trap of a genre that packages suffering into something neatly consumable. As film researcher Pooja Rangan argues, humanitarian documentaries often render asylum seekers passive, their worth measured by how much sympathy they can elicit. Go Back to Where You Came From follows this script, focusing not on the agency of the displaced, but on the moral awakenings of those who continue to look away.

The real question is not whether the participants feel something, but whether feeling will ever translate into action – by them, or by us as viewers. To hold governments to account. To insist that refuge is a right, not a privilege. To refuse the quiet, grinding violence of neglect.

“Go back to where you came from” is a phrase hurled not just at refugees, but at anyone deemed out of place. The programme inverts it, sending its wielders on a reckoning. But in the end, they return. To safety, to comfort, to homes untouched by war or exile. Or, as one put it, back to the pub.

And yet, for those seeking refuge, the journey drags on – through border camps, detention centres, doorways, the freezing cold and the bureaucracy of the asylum system – while the world watches, then turns off their televisions. Läs mer…

Generative AI is often seen as the epitome of our times, and sometimes even as futuristic. We can use it to invent new art or technology, analyse emerging data, or simulate people, places and things. But interestingly, it is also having an impact on how we view the past.

AI imagery has already been used to illustrate popular articles, such as covering scientific discoveries about Neanderthals. It was employed to animate the Mesolithic period (from about 9,000 to 4,300 years ago) in a museum. TikTok users have adopted it to make realistic short videos about archaeology and history. It’s even been used in a TV documentary about Stonehenge.

AI-generated image.
Canva

Yet there are many issues with using AI imagery in archaeology – some of which are also found more broadly within generative AI use. These include its environmental impact and the violation of intellectual property (using training data created by humans).

But others are more specific to archaeology. As an academic who has worked extensively on “resurrecting” the past through digital technology, generative AI has both fascinating potential and enormous risk for archaeological misrepresentation.

Even before the use of AI, it was widely accepted within archaeology that visualisations of the past are highly fraught and should be treated with extreme caution. For example, archaeologist Stephanie Moser examined 550 reconstructions published in academic and popular texts on human evolution. Her review found highly biased depictions, such as only males hunting, making art and tools and performing rituals, while women were in more passive roles.

A similar study by Diane Gifford-Gonzalez revealed that “not one of 231 depictions of prehistoric males shows a man touching a child, woman, or an older person of either sex … no child is ever shown doing useful work.” These reconstructions do not reflect scientists’ nuanced understanding of the past. We know humans organised themselves in an incredible array of variety, with a multitude of gender roles and self-expression.

A recent DNA-based study, for example, showed that women were actually at the centre of societies in the iron age.

The stakes of representation in archaeology are high. For example, the hotly-debated, dark-skinned reconstruction of “Cheddar Man”, originally found in south-west England, was based on ancient DNA analysis. It made headlines for disrupting the perception that all human ancestors in the north were light-skinned.

Reconstructed head of the Cheddar Man based on the shape of his skull and DNA analysis, shown at the Natural History Museum in London.
wikipedia, CC BY-SA

This and similar controversies reveal the iconic power of reconstructions, their political implications, and their ability to shape our understanding of the past.

While the Cheddar Man reconstruction demonstrates that research is iterative, such reconstructions are sticky. They have profound visual legacies and are not easily supplanted when new data becomes available.

This is exacerbated as they are incorporated into generative AI data sets.
Beyond the use of outmoded data, generative AI visualisations of the past can be extremely poor.

Even when more plausible details are included, they can be seamlessly integrated with other highly inaccurate elements. For example, it is impossible for viewers to disentangle the data-led from the so-called hallucinations (mistakes) produced by AI.

Highlighting uncertainty is of central importance and concern among archaeologists. Archaeological illustrator Simon James noted that reconstruction artists have used strategically placed clouds of smoke to obscure unknown elements.

As a digital archaeologist, I have made virtual reconstructions of many different sites and subjects. I know there is often estimation and guesswork involved in making holistic representations.

Indeed, photo-realistic accuracy is not always the paramount consideration in visualisation – particularly when exploring different hypotheses or addressing young audiences. But knowing what is backed by archaeological data and what is more speculative is key for authentic visual communication.

Pseudoarchaeology

This is particularly important at a time when pseudoarchaeology is increasingly prevalent in popular media, such as the Ancient Apocalypse show on Netflix. The celebrity host and author Graham Hancock asserts there was a lost ice age civilisation of Atlantis, with advanced technology. But this claim has been thoroughly repudiated by archaeologists.

Arguably, hoaxes will be much easier to perpetuate using generative AI.
Beyond the high potential for misinformation about archaeology, the use of generative AI for archaeological visualisations can actually be harmful for archaeological knowledge production.

My research has shown that crafting reconstructions and illustrations in archaeology is incredibly important for understanding and interpreting the past. Creating visualisations based on science – and indeed soundscapes, smellscapes and other interpretations based on multiple senses – is very helpful for generating new questions.

Drawing allows archaeologists to create more detailed mental models and therefore a better understanding of archaeological remains. By delegating this creation to AI, archaeologists lose a powerful tool for knowledge generation. Moreover, my collaborative work with artists has demonstrated the intriguing possibilities that creative approaches open up to tell new stories about the past.

Even with all of these problems, I encourage an engaged, critical, applied approach to understanding the impact of digital technologies on our investigation of the past. And this includes exploring the uses of generative AI for archaeological visualisation.

Archaeologists and non-specialists are able to leverage generative AI to creatively produce interpretive media. Indeed, some archaeologists are already exploring AI to generate hypotheses about ancient life. And we are teaching critical uses of AI to our archaeology students.

But what remains imperative is that archaeologists engage with and critique all visualisations – both those created by generative AI and using other media. Läs mer…