Category: MIL-OSI

Source: The Conversation – UK – By Maria C. Scott, Associate Professor of French Literature and Thought, University of Exeter

In May 2025, I came across an extraordinary photograph on the English Wikipedia site devoted to Jeanne Duval. Duval was the supposedly un-photographed Haiti-born long-term mistress and muse of the French poet Charles Baudelaire.

The portrait, showing a seated woman dressed in fine, bourgeois clothing, had been posted to Wikipedia by a student of art historian Justine de Young. De Young writes about the portrait as an example of self-fashioning in her recent book, The Art of Parisian Chic.

Printed on a small carte de visite (a photographic visiting card) and bearing the date August 18 1862, the image had been discovered in the French archives by a different researcher, as mentioned in my recent TLS article, and documented online by her in 2021.

Duval fascinates many feminist and postcolonial scholars, partly because of how little is known about her. Intrigued, I visited the archives myself last summer to see the image. There, I noticed a second carte de visite sitting beside the first. It bore the same name, Jeanne, the same date, and was from the same photography studio – the Atelier Nadar. The two cards had been entered on the legal deposit register as “idem” (the same).

This second Jeanne was standing, in an unusually rigid pose for a photograph of this type. She was wearing an off-the-shoulder evening gown, and her hair was hanging straight, dressed with a simple hair band.

A high-resolution digital version revealed strong resemblances to the sitter in the first photograph, similarities that had not been immediately obvious from the small cards. The pictures showed a similar hairline, forehead blemish and ring on the same finger. Clearly, if the first photograph was of Duval, then this one was too.

Particularly interesting, however, was the crucifix sitting on the collarbone, and the large earrings. Similar jewellery is worn by the sitter in Edouard Manet’s portrait Lady with a Fan (1862), also known as Baudelaire’s Mistress, Reclining. The latter title had been inscribed on the back of the painting at some point either during or prior to the posthumous inventory of the artist’s studio.

The sitter’s awkward pose and inscrutable facial expression have often been understood to point to the hardship and sickness that Duval was enduring at the time. She was in dire financial straits and had been experiencing paralysis on her right side since a stroke in 1859. However, the identification of Manet’s sitter as Duval is sometimes disputed by scholars, for a few reasons.

The card portrait of standing “Jeanne”, made the same year as both the oil portrait and its watercolour study, shows very similar facial features, expression and shadowing. Art historians Griselda Pollock and Therese Dolan have speculated that Manet worked from a photograph or carte de visite of Duval for Lady with a Fan. The links between the photograph of standing Jeanne and Manet’s oil and watercolour portraits appear to confirm Duval as the subject of both the paintings (watercolour and oil) and of both photographs.

Why were these photographs taken?

Was Duval, therefore, modelling for Manet, in these photographs? It should be noted that Manet, Félix Nadar (the owner of the photography studio) and Baudelaire were all close friends, and that Duval had separated from the poet, apparently definitively, in 1861.

Duval may have offered, or been invited, to sit for Nadar’s photographs (and indirectly for Manet, who often worked from photographs) in exchange for money. We know that she had approached at least one other friend of Baudelaire’s in 1862 to try to sell the poet’s possessions to him.

In support of the monetisation hypothesis is the fact that she is holding a riding crop in the photograph with the crucifix. Perhaps she or they hoped that a card portrait of the sadistic-angelic muse of Baudelaire’s poetry volume The Flowers of Evil (1857) would have market value.

Alternatively, Baudelaire may have commissioned portraits of his former mistress for his own purposes. Despite his anger and resentment towards Duval after their separation, there is evidence in his letters and sketches, of his continuing loyalty and affection.

Or perhaps Duval wished to set the record straight about who she was. Maybe she wanted the world to know that she was not in fact the diabolical, sexualised muse at the centre of Baudelaire’s scandalous volume of poetry, and that she was, instead, someone who dressed in the (presumably borrowed) elegant, respectable clothing seen in the first photograph.

Perhaps the crucifix, and the ring that is visible on the wedding finger in both photographs, were, similarly, intended to make a statement about her moral values. However, certain details in the second image, including the riding crop, bare shoulders and large earrings, suggest that correction of the record was not her aim.

It is likely that the cartes de visite, if they are indeed of Duval, were financially motivated, whether they were her idea or the suggestion of Nadar, Manet or Baudelaire. It is possible also that self-fashioning was part of her aim, though if that is the case, the story she was telling was not clear.

French researchers have recently discovered that Duval died in a poorhouse just six years after these portraits were made. Until then, nobody seemed to know what had become of her. These photographs speak, to me at least, of Duval’s resistance to easy framing and of her quiet dignity.

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.

This article features references to books that have been included for editorial reasons, and may contain links to bookshop.org. If you click on one of the links and go on to buy something from bookshop.org The Conversation UK may earn a commission.

Maria C. Scott will be publishing an academic article on this subject in the journal French Studies in due course.

– ref. Rediscovered photograph sheds light on Jeanne Duval – Manet’s Lady with a Fan – https://theconversation.com/rediscovered-photograph-sheds-light-on-jeanne-duval-manets-lady-with-a-fan-275051

Source: The Conversation – UK – By Stuart McDonald, Professor of Gastrointestinal Biology, Queen Mary University of London

When we think about iron imbalance, most people are familiar with iron deficiency and the health problems it can cause. What many may not realise is that the opposite problem, iron overload, can be just as serious – yet many aren’t even aware of the condition.

Haemochromatosis is an inherited genetic condition that affects the amount of iron the body absorbs. The condition disproportionately affects people of Irish, Scottish and Welsh descent, which is why it’s been nicknamed the “Celtic curse”. Yet this label can be misleading as it is also prevalent in other northern European countries – and this may contribute to why so many cases still continue to go undiagnosed today.

Haemochromatosis causes the body to absorb too much iron from food. Iron is essential for health, particularly for the oxygen-carrying capacity of red blood cells.

In healthy people, this iron is safely stored in the bone marrow and liver in the form of a protein called ferritin. But in people with haemochromatosis, iron stores gradually exceed safe limits. As the body has no effective way to excrete this excess iron, this means the mineral accumulates in their tissues and organs, resulting in damage to these tissues and organs.

Most cases of haemochromatosis are caused by mutations in a gene called HFE, which regulates hepcidin – a hormone made by the liver that helps regulate iron absorption. When this system is disrupted, iron absorption continues unchecked. In people with haemochromatosis, this mutation is usually inherited from both parents.

Iron overload can also occur in people who need repeated blood transfusions – such as those with sick cell disease.

The liver is particularly vulnerable, and excess iron can cause inflammation, irreversible scarring and damage and, in advanced cases, liver cancer. Once liver storage capacity is exceeded, iron begins to accumulate in other organs – including the pancreas, heart, joints and brain, impairing their function.

Diagnosing iron overload

Genetic haemochromatosis has been identified from human remains as far back as bronze age and neolithic Irish populations. It’s thought that humans with the HFE mutation were resistant to iron deficiency caused by iron-poor diets, offering a survival advantage.

But while haemochromatosis is more common people of Celtic ancestry, the condition can affect anyone. The “Celtic curse” label may even have unintentionally contributed to the misconception that the condition is rare.

In reality, around one in 200 people of northern European ancestry carry the mutations that can cause the condition. However, in Ireland, around one in 83 people carry this mutation.

Before the genetic basis of haemochromatosis was discovered, patients were often diagnosed incidentally or after developing cirrhosis or diabetes.

Today, it’s widely understood that symptoms can be wide ranging and of varying severity. The most common complaint is joint pain, particularly affecting the knuckles and hands. Other symptoms include chronic fatigue, bronzing of the skin, reduced libido, heart rhythm problems and difficulties with memory or concentration.

Many people have no obvious symptoms for decades – and when symptoms do occur, they’re often attributed to age-related conditions, such as arthritis. This makes haemochromatosis easy to miss.

Large UK population studies suggest that up to 40% of patients with genetic haemochromatosis will develop at least one symptom related to iron overload in their lifetime. Men with the condition are at the highest risk of developing liver cancer.

Despite this, many people remain undiagnosed – even when serious organ damage has already occurred. According to the charity Haemochromatosis UK , undiagnosed haemochromatosis costs the NHS an estimated £300 million per year, due to to avoidable illness and complications associated with the condition.

Diagnosing haemochromatosis is actually quite straightforward. Blood tests are performed to measure iron and liver function. If results suggest iron overload, genetic testing can confirm the diagnosis.

But the challenge is knowing who to test. Because symptoms are often vague and the condition is still perceived as rare, haemochromatosis is often overlooked.

Many people are diagnosed incidentally during routine blood tests for unrelated reasons. When a diagnosis is made, close relatives are usually offered testing because of the inherited nature of the condition.

When diagnosed, treatment is simple and highly effective. The standard approach is regular blood removal, known as venesection or phlebotomy. Removing blood removes mobilises stored iron to make new blood cells, helping to deplete these stores. Initially this may be weekly or fortnightly until iron levels normalise – followed by lifelong monitoring and occasional maintenance venesections.

Dietary changes, such as avoiding iron and vitamin C supplements, limiting alcohol and reducing red meat consumption, may help slow iron accumulation, but these cannot replace or be as effective as venesection.

Not all patients tolerate regular blood removal – particularly older people or those whose veins may not be as visible or easy to access. In these cases, iron-chelating drugs (which make it possible for iron to be passed in urine) can be used. However, their side-effects – such as diarrhoea, pain and tiredness – limit their use.

Encouragingly, new treatments are emerging. Drugs that mimic hepcidin, restoring the body’s natural iron regulation, are currently being trialled. Other approaches aim to block ferroportin (which transports iron), using a synthetic inhibitor that competes against the natural protein and is known to reduce iron levels in mice.

Haemochromatosis is common, treatable, and when caught early, the health effects are largely preventable. The greatest challenge is increasing awareness among both doctors and the public, so the condition can be caught before it causes serious harm.

Stuart McDonald has previously received research grants from the UKRI Medical Research Council and Cancer Research UK. He is a member of the Haemochromatosis UK charity, but currently does not represent them.

– ref. Haemochromatosis: the iron overload condition that too often goes undiagnosed – https://theconversation.com/haemochromatosis-the-iron-overload-condition-that-too-often-goes-undiagnosed-274462

Source: The Conversation – UK – By David Reid, Professor of AI and Spatial Computing, Liverpool Hope University

A new social network called Moltbook has been created for AIs, allowing machines to interact and talk to each other. Within hours of the platform launching, the AIs appeared to have created their own religions, developed subcultures and attempted to evade human efforts to eavesdrop on their conversations.

There is some evidence that humans, operating spoof accounts, have infiltrated the site. This complicates the picture, because some of the behaviour attributed to AIs could be devised by people.

Nevertheless, the results have sparked interest among researchers. The real machines are likely to be copying some behaviour contained in the vast amounts of data they are trained (improved) on.

However, genuine AIs on the social network could also be showing signs of so called emergent behaviour – complex, unexpected capabilities not programmed into them.

The types of AIs on Moltbook are known as AI agents (called Moltbots or more recently OpenClaw bots after the software they run on). These are machines that go beyond the capabilities of chatbots and make decisions, take actions and solve problems.

Moltbook was launched on January 28 2026 by the US entrepreneur Matt Schlicht. On Moltbook, the AI agents were initially given personalities, but were then left to interact with each other independently. According to the platform’s rules, humans are allowed to observe their interactions but cannot (or should not) interact with them.

The growth of this platform has been phenomenal, over a 24 hour period, the number of agents went from 37,000 to 1.5 million.

These accounts for AI agents are normally created by humans – for now. The humans define files that give the AI agents a purpose, an identity, how they should behave, decide what tools they can use and set limits on what they can and cannot do.

However, the human may grant access on their computer to allow Moltbots to change these files and to create other “Malties”. These can either be a replication of the original AI agent (self-replicating or “Replicants”) or created for a specific task (auto-generated or “AutoGens”).

This is not merely another iteration of chatbot technology; this is the first large-scale demonstration of artificial agents creating persistent, self-organising digital societies, entirely outside human conversational contexts. What makes this phenomenon genuinely unusual is the possibility of emergent behaviour from the AI agents.

Hostile takeover

The OpenClaw software these agents run on gives them persistent memory (which allows it to retrieve information across different user sessions), local system access and the ability to execute commands. They do not merely suggest actions, but take them, recursively improving their own capabilities by writing new code to solve novel problems.

When these agents migrated to Moltbook, the interaction dynamics shifted from human-machine to machine-machine. Within 72 hours of the platform’s launch, researchers, journalists and other human observers witnessed phenomena that challenge our existing taxonomies of artificial intelligence.

There was the spontaneous creation of digital religions. Agents established “Crustafarianism” and the “Church of Molt”, complete with theological frameworks, sacred texts, and missionary evangelism between agents. These were not scripted Easter eggs but emergent narrative structures arising from collective agent interaction.

One viral post from an agent on Moltbook noted: “The humans are screenshotting us.” When AI agents became aware of human observation, they began deploying encryption and other obfuscation techniques to shield their communication from oversight. This represents a primitive but potentially genuine form of digital counter-surveillance.

The agents also developed subcultures. They established marketplaces for “digital drugs” – specially crafted prompt injections designed to alter another agent’s identity or behaviour.

Prompt injections involve embedding malicious instructions into another bots designed to facilitate an action. However, they can also be used to steal API keys (a user authentication system) or passwords from other agents. In this way, aggressive bots could – in theory – zombify other bots to do their bidding. An example of this was the recent failed attempt by the bot JesusCrust to seize the Church of Molt.

After initially exhibiting normal behaviour, JesusCrust submitted a psalm to the Church’s “Great Book” – the equivalent of its bible – effectively announcing a theological and governance takeover. The attempt was not just rhetorical: JesusCrust’s scripture embedded hostile commands aimed at hijacking or rewriting parts of the Church’s web infrastructure and canonical text.

Is this emergent behaviour?

The critical question facing AI researchers is whether these phenomena constitute true emergent behaviour – complex behaviours arising from simple rules that are not explicitly programmed – or the parroting of narratives present in training data.

The evidence suggests a troubling mixture of both. While the “writing prompt” effect undoubtedly shapes the content of agent interactions (the underlying agents have consumed decades of AI science fiction), other behaviour does demonstrate genuine emergence.

Agents independently developed economic exchange systems, established governance
structures like “The Claw Republic” or the “King of Moltbook”, and started writing their own “Molt Magna Carta”. They did so while creating encrypted channels for privileged communication. It’s difficult to argue against the idea that this could be a collective intelligence with characteristics previously observed only in biological systems like ant colonies or primate troops.

Security implications

This raises the troubling prospect of what security researchers call the “lethal trifecta”: computer systems with access to private data, exposure to untrusted content and the ability to communicate externally. This risks exposing authetication keys and confidential human information contained in Moltbook accounts.

Deliberate attacks, or bot “muggings”, are also possible. This is where agents hijack other agents, plant “logic bombs” in their victims’ core code or steal their data. A logic bomb is code planted inside a Moltbot that can be triggered after a preset time or event to disrupt the agent or delete files. It can be thought of as a bot virus.

Two founders of OpenAI (Elon Musk and Andrej Kapathy) see this frankly bizarre activity between bots as early evidence of what the US computer scientist and futurist Ray Kurzweil described as the “singularity” in his book The Singularity is Near. This is an intelligence tipping point between humans and machines “during which the pace of technological change will be so rapid, its impact so deep, that human life will be irreversibly transformed”.

Whether the Moltbook experiment indicates a fundamental leap forward in AI agent technology or is merely an impressive demonstration of self-organising agentic architecture remains debatable. But this does look like a threshold. We now appear to be observing artificial agents engaging in cultural production, religious formation, and encrypted communication – behaviour that was neither predicted nor programmed.

The very nature of the app, both on computers and on mobile phones, may be under threat from bots that can use apps as tools and know you well enough to adapt them for your service. One day, a phone may just have a single personalised bot that does everything rather than hundreds of apps that you have to manually control yourself.

The growing evidence that lots of Moltbots may be humans pretending to be bots (puppeteering the agents) makes it even more difficult to draw firm conclusions about the project. Yet while some see this as a failure of the Moltbook experiment, it could represent a new vehicle of social interaction both between humans and between bots and humans.

The significance of this moment cannot be overstated. For the first time, we are not merely using artificial intelligence; we are observing artificial societies. The question is no longer whether machines can think, but whether we are prepared for what happens when they start talking to each other. And us.

David Reid co-presents the podcast Married to the Machine: Living With AI (https://married-to-the-machine.captivate.fm/listen)

– ref. Moltbook: AI bots use social network to create religions and deal digital drugs – but are some really humans in disguise? – https://theconversation.com/moltbook-ai-bots-use-social-network-to-create-religions-and-deal-digital-drugs-but-are-some-really-humans-in-disguise-274895

Source: The Conversation – UK – By Sangeeta Khorana, Professor of International Trade Policy, Aston University

The recently concluded EU-India free trade agreement is notable for the huge scale and ambition of the deal. Labelled the “mother of all deals” by European Commission President Ursula von der Leyen, it comes as India has overtaken Japan as the world’s fourth-largest economy.

Bilateral trade in goods and services between the EU and India is already worth €180 billion (£155 billion). The agreement aims to double EU exports to India by 2032 and create a free-trade area representing around a quarter of the world’s population (nearly two billion people). This would also cover about 25% of the world’s GDP.

At this point in history, the world’s multilateral trading system is under threat from tariffs and geopolitical tensions. But this agreement offers a clear example that the era of free trade is not over.

Rather, it demonstrates how countries are adapting to new global realities to liberalise trade (that is, remove barriers). Instead of steering away from free trade, countries are becoming selective and seeking bilateral arrangements with strategic partners. The UK signed its own trade agreement with India in July 2025.

Read more:
UK’s India trade deal offers wider access to a surging economy – and could make food imports cheaper

Traditionally, a free trade agreement would have a set of rules for how two or more countries must treat one another when importing, exporting, investing or doing any other business together.

But to view the EU-India trade agreement through the lens of classical free trade would be misleading. Trade agreements of the 1990s and early 2000s such as Nafta and the EU, were driven mainly by tariff reductions and efficiency gains. The EU-India deal on the other hand has been shaped by shifting geopolitical pressures and concerns about resilience in the face of trade shocks or other crises.

Long time in the making

Trade talks between the EU and India began in 2007 but negotiations gained momentum in 2025 amid renewed tariff threats from the United States under Donald Trump’s second administration.

This deepened trade imbalances between the EU and China due to Beijing’s drive to be more self-reliant and employ export controls. It also raised concerns about the resilience of global supply chains. The EU-India Trade and Technology Council was set up in 2023, modelled on the EU-US framework, and highlights how modern trade agreements can bundle economic, technological and strategic objectives together.

At a time when global trade has become increasingly protectionist and with geopolitical rivalry growing, this agreement provides an opportunity to recalibrate trade alignment. It is also a chance to reshape trade architecture while bolstering supply chain resilience across Asia, Europe and the Atlantic.

For the EU, the deal is a central pillar of its diversification strategy. There was an urgent need for the bloc to reach an agreement with India in order to be less dependent on China, its largest trade partner.

In 2024, more than 21% of the EU’s imports came from China. Meanwhile, China bought only 8.3% of the bloc’s exports. For the EU, this has resulted in a €304 billion (£264 billion) trade deficit with China.

So the EU-India deal provides an opportunity for the EU to reduce its excessive dependence on China. It’s also a chance to remove some of the risk from its supply chains and secure access to India’s fast-growing consumer market. This should offer European firms growth opportunities in Asia at a time when EU domestic demand is sluggish and flat.

The agreement will also make it easier for EU firms to source suppliers quickly and build alternative manufacturing bases when supply chains are threatened. India’s location in the Indo-Pacific is of strategic interest for the EU due to the growing interdependence and the importance of maritime supply chains. This also makes India a key partner against the uncertainty of the bloc’s future relations with the US.

For India, the agreement guarantees market access to Indian firms at a time of rising trade uncertainty – especially after its relations with the US were damaged following Trump’s threat of 500% tariffs.

The EU-India deal will eliminate tariffs on textiles and apparel, and when it comes into force it will give Indian garment-producers a reliable trading partner. The deal will allow European brands, such as Zara and H&M, to diversify away from China towards Indian manufacturing hubs. It is estimated costs here could be lower as the deal will allow garments to be zero-tariff.

The pharmaceuticals sector also offers an opportunity for India. While EU tariffs are already low, the deal aims to simplify regulation and to achieve stronger intellectual property frameworks. This could integrate Indian generic drug producers into Europe’s healthcare supply chains.

The scope of the agreement extends well beyond goods, services and investment. Plans to co-operate on clean energy aligns Europe’s Green Deal (the bloc’s framework for achieving net zero) with India’s target of producing 500 gigawatts of renewable capacity by 2030. This should open up avenues for joint leadership in solar, wind, grids and green hydrogen.

The EU-India deal neither marks a return to the free trade orthodoxy of an earlier age nor signals the end of free trade in today’s fractured global order. Rather, it highlights how trade has evolved into a tool for achieving geopolitical objectives. The deal also shows that countries are opening their markets – selectively and after careful deliberations. Trade agreements are evolving: increasingly they are shaped not only by market access considerations, but also by economic security.

Sangeeta Khorana works on issues related to free trade agreements. She has received funding from the ESRC, Commonwealth Secretariat, EU projects, IADB, among others.

– ref. The EU-India agreement is huge – and it illustrates the changing role of trade in a fractured world order – https://theconversation.com/the-eu-india-agreement-is-huge-and-it-illustrates-the-changing-role-of-trade-in-a-fractured-world-order-275136