Feed aggregator

The European Union (EU) and southern Medi-terranean partners launched the Pact for the Mediter-ranean in November 2025 to reset relations with the EU’s “Southern Neighbourhood” in an increasingly challenging regional context. The Pact comes 30 years after the 1995 Barcelona Process promised to foster economic – and to a lesser degree political – integration in the Mediterranean Basin. The Pact’s declared objective is to “achieve deeper integration within the common Mediterranean space” (EC & HR, 2025). This policy brief discusses the Pact’s prospects for achieving this goal, which previous efforts have failed to reach. For long-time observers of Euro-Mediterranean rela-tions, the Pact appears to be a “back to the future” approach. Its three substantive “pillars” (people, econo-mies and security) echo the three “baskets” (political/ security, economic and socio-cultural) of the original Euro-Mediterranean Partnership. Structurally, it relies on the same mix of differentiated bilateral agreements (now termed “comprehensive partnerships”) within a multilateral regional framework. The Pact’s success depends on whether the EU and Mediterranean partner countries can resolve four core dilemmas that have long challenged their relations:
• The “autocracy dilemma”: balancing the need to work with authoritarian governments with European interests in supporting democracy.
• The “migration dilemma”: securing borders while respecting human rights.
• The “rentierism dilemma”: finding solutions to immediate economic, social and environmental challenges while making necessary reforms to rentier political economies.
• The “regionalism dilemma”: cutting bilateral deals while trying to build regional structures to address collective action problems.
The term “pact” is normally used to describe an agree-ment between two partners, setting out agreed objec-tives and actions for both sides. The Pact for the Mediterranean is an EU policy framework that, at most, represents a tacit agreement with southern Mediter-ranean governments, without committing either side to policy changes or reforms that might have long-term implications. The Pact for the Mediterranean has potential to strengthen sectoral cooperation, for example on renew-able energy, connectivity infrastructure and labour mobility. If accompanied by sufficient resources and mutual trust-building, this functional cooperation may create incentives for deeper integration. This, in turn, will still depend on whether the EU and southern Mediterranean governments can move beyond trans-actionalism and invest in partnerships between their societies: support for democratic movements and institutions, investment in public goods, protection of the natural environment and investment in collective regionalism. Thus far, there is little indication that the EU and southern Mediterranean governments will take advantage of this opportunity.

Également dans l'édition de jeudi : l'Iran, Frontex, UE-Chine, la Hongrie, du café, l'OTAN

The post Orbán et la fin de l’histoire appeared first on Euractiv FR.

La Turquie se prépare à accueillir le prochain sommet de l'Otan en juillet et souhaite renforcer son rôle au sein de l'Alliance atlantique. Cet engagement croissant ne fait pas consensus dans la société. La gauche appelle même à la quitter.

- Articles / Turquie, Courrier des Balkans, Défense, police et justice, Guerre Moyen Orient, Une - Diaporama

- Agenda / Agenda - Diaporama, Région parisienne

Sortie de l'album : Cosmogonie (Rue Des Balkans - 2026)
Originaire de Sarajevo, carrefour entre l'Orient et l'Occident, Srdjan Ivanovic est un batteur au parcours des plus atypiques.
Ayant fui la guerre en Bosnie lorsqu'il était enfant, Srdjan s'est retrouvé en Grèce, où il s'est plongé dans un mélange éclectique de musiques, allant du chant byzantin et du folk grec à la pop italienne, entre autres. Ajoutant à cela sa découverte du jazz, il a développé son propre style musical, à la fois (…)

- Agenda / Région parisienne

L’entreprise Melouki Santé participe à la 28e édition du Salon International de la Santé (SIMEM), organisée du 8 au 11 avril 2026 au Centre de […]

L’article SIMEM 2026 à Oran : Melouki Santé dévoile ses nouveautés en diagnostic et équipements de laboratoire est apparu en premier sur .

Le ministère de l’Énergie et des Énergies renouvelables a annoncé une nouvelle mesure visant à mieux protéger les consommateurs face aux coupures d’électricité. Désormais, les […]

L’article Factures d’éléctricité : l’État fixe de nouvelles règles pour les coupures en cas d’impayés est apparu en premier sur .

Le président de la République, Abdelmadjid Tebboune, a décidé de mettre fin aux fonctions du ministre de l’Hydraulique, Taha Derbal, a indiqué un communiqué de […]

L’article Tebboune mets fin aux fonctions du ministre de l’Hydraulique est apparu en premier sur .

Ex-President Yahya Jammeh's 22-year rule was known for enforced disappearances and extrajudicial killings.

By James Alix Michel
VICTORIA, Seychelles, Apr 8 2026 (IPS)

We live in a century of extraordinary achievement.

Humanity has split the atom, mapped the genome, and sent astronauts to the Moon, with plans now underway to reach Mars. Our knowledge has expanded, our tools have become more powerful, and our capacity to shape the world around us exceeds anything previous generations could have imagined. We communicate instantaneously across continents, diagnose diseases earlier, monitor climate patterns in real time, and design artificial intelligences that can aid in everything from medicine to climate modelling.

James Alix Michel

And yet, for all this advancement, we are caught in a troubling paradox.

We possess the means to protect our planet, restore degraded ecosystems, and build a future that is regenerative and sustainable. The Earth still holds enough resources to feed, shelter, and nourish every person on it.

The science is clear, the solutions are known, and the pathways are increasingly understood. We know how to phase out the most damaging fossil fuels, how to design circular economies, and how to restore forests and oceans on a large scale. The question is not whether we can heal, but whether we choose to.

Instead of using this knowledge to nurture life, we spend trillions on weapons, war, and systems of domination. We continue to refine instruments of destruction with the same ingenuity that once helped us survive as hunter gatherers.

From spears and arrows to missiles and nuclear arsenals, technology has evolved far faster than our moral imagination. The same species that can design satellites and decode life itself is also capable of perfecting the means to erase itself. We have turned our curiosity into a danger when it is not paired with humility.
War has become normalised. We export violence beyond our borders, fuel conflicts in distant lands, and justify the dehumanisation of others in the name of power, ideology, or fear.

In doing so, we risk losing sight of what it means to be human: to care, to share, to protect, and to build together. Our intelligence has grown, but our ethics have often lagged behind. We have impressive control over external environments, yet we struggle to govern our own impulses—greed, resentment, the desire for domination over cooperation.

We still behave as if survival depends on conquest, as though strength is measured by the capacity to destroy rather than by the courage to cooperate.

In that sense, humanity is trapped between two identities: one capable of profound creativity and compassion, and another still governed by ancient instincts of greed, lust for power, and tribal dominance.

We have evolved in technology, but not always in spirit. We built institutions meant to protect rights and distribute justice, yet those very institutions are often weaponised or hollowed out by self interest.

The Earth is still rich enough to nourish us all. The ocean still teems with life, the land can still grow food, and the air can still be cleansed. We have the tools to live in balance, instead of in excess. We can choose renewable energy systems that do not poison our skies, farming practices that restore soil instead of depleting it, and urban designs that integrate nature instead of paving it over.

The problem is not scarcity, but choices—choices that prioritise short term gain over long term survival, accumulation over equity, and fear over trust.

If humanity is to truly evolve, it must move beyond the old logic of domination and embrace a new ethic of stewardship. This is not a soft or sentimental vision. It is a hard, practical necessity if we want civilisation to continue.

Stewardship means recognising that power is not only the ability to control, but the responsibility to protect. It means designing economies that reward regeneration, not extraction; diplomacy that favours mediation over militarisation; and education systems that nurture empathy as much as efficiency.

Progress cannot be measured only by how far we can reach into space, or how fast we can compute. It must be measured by how well we can care for the planet and for one another. It must be measured by how peacefully we resolve our differences, how fairly we share resources, and how seriously we protect the rights of future generations.

True progress is the transition from a species that merely adapts to its environment, to one that consciously shapes it for the benefit of all life, not just a privileged few.

We have not lost our humanity. We have only forgotten it.
The challenge now is to rediscover it—not as a romantic ideal, but as a practical imperative.

In a world capable of such beauty, creativity, and connection, the only true insanity is the choice to destroy rather than to heal, to dominate rather than to share, and to fear rather than to love.

After all, the moon and the stars will remain, no matter how we choose; what is at stake is whether we will still be worthy of the Earth we were given.

That is the real test of our century. And it is one we must pass together.

IPS UN Bureau

 

Follow @IPSNewsUNBureau
!function(d,s,id){var js,fjs=d.getElementsByTagName(s)[0],p=/^http:/.test(d.location)?'http':'https';if(!d.getElementById(id)){js=d.createElement(s);js.id=id;js.src=p+'://platform.twitter.com/widgets.js';fjs.parentNode.insertBefore(js,fjs);}}(document, 'script', 'twitter-wjs');  

Le procès de l’ancien ministre de l’Industrie et de la Production pharmaceutique, Ali Aoun, s’est officiellement ouvert dans la soirée du lundi 6 avril, au […]

L’article Scandale d’IMETAL : lourd réquisitoire contre l’ex-ministre Ali Aoun et un ancien président de la FAF est apparu en premier sur .

C’est une pièce d’histoire que l’on croyait perdue dans les méandres des « souvenirs » de guerre. L’arme de combat du colonel Amirouche, chef historique de la […]

L’article Après plus de dix ans de mystère : la carabine du colonel Amirouche refait surface, on vous dit tout est apparu en premier sur .

Les échanges commerciaux entre la France et l’Algérie marquent le pas. Après plusieurs années de croissance, l’exercice 2025 se clôture sur une baisse de 14,8 […]

L’article Échanges commerciaux : la France perd du terrain face aux nouveaux partenaires de l’Algérie est apparu en premier sur .

Written by Antonio Vale.

Introduction

The past few years have seen considerable interest in generative AI, particularly large language models (LLMs). This has translated into massive investment amounting to hundreds of billions of euros per year, especially in the US, in AI data centres designed around Graphics Processing Unit (GPU)-based platforms. Such breakneck expansion is increasingly running into constraints, particularly with regard to electricity availability.

Running AI models requires large amounts of power (as well as water, much of which is used to produce the electricity required), with data centres responsible for 1.5 % of global electricity consumption (2 % in the EU) and growing at 12 % annually. Moreover, they are often geographically concentrated, for example in Ireland, where they account for over 20 % of electricity consumption. Future scenarios suggest that this demand could continue to increase rapidly, although this should be taken with the caveat that investment in AI might be a bubble, LLMs may be supplanted by other models with different compute needs, and chip design innovations beyond GPUs may provide energy efficiency gains.

This situation has given rise to the idea of deploying compute in space to take advantage of the free, abundant solar energy. Originally focused on orbital processing of observational data and space mission support, the concept has rapidly evolved into the deployment of AI data centres in orbit to service ground-based needs. Recently, the strongest push has come from the US, with the merger between SpaceX and xAI linked to a request to put a million satellites in orbit, as well as interest from Google with project Suncatcher, and startups such as Starcloud and Axiom. Meanwhile, China has also launched pilot satellites intended to be the first in a future constellation, and in the EU the Horizon Europe-supported ASCEND project has concluded a feasibility study, aiming towards an operational system from 2030.

Potential impacts and developments

Launch costs represent a key constraint for any orbital infrastructure. The introduction of reusable rockets has led to a considerable decrease in recent times, to around several thousand euros per kilo of payload. This reduction is expected to continue thanks to improved heavy rockets and reusable second stages, with the European Space Agency (ESA) aiming for €280/kg with a new super-heavy lift launcher. Most ideas for future space data centres would involve either large constellations or modular construction, allowing build-up to occur over time. Even so, this would require a very high launch cadence, with a complete data centre likely needing upwards of one hundred launches, followed by a significant proportion yearly to replace satellites at end of life; this compares to around 300 space launches overall in 2025.

The main attraction of placing data centres in space is solar power: for objects located above the atmosphere, insolation (incoming solar radiation) can be several times greater than on the ground. The ideal choice would be a terminator sun-synchronous orbit, allowing satellites to keep pace with the dawn/dusk line and ensuring constant solar exposure on one side, while keeping the other dark to assist with cooling. Solar panels would need to be very large – up to a gargantuan 4 km per side, as envisaged by Starcloud for a 5 GW data centre; a small satellite with the equivalent of a server rack might make do with a more manageable 60 m2 and 28 kW, as deployed on the International Space Station (ISS). Newer thin-film solar panel technology may help keep the weight down.

If power is the main advantage, cooling is possibly the major challenge. Although space is cold, it is also a vacuum, meaning cooling can only take place via radiative emission. This can be achieved by coupling a coolant loop (the ISS uses ammonia) with large radiators pointing towards deep space, which would be of comparable size to the solar panels but considerably heavier. The spacecraft’s cooling system is particularly vulnerable: any rupture, for example from a meteoroid strike, can cause coolant loss and damage the electronic systems. Given radiative cooling scales as the fourth power of temperature, further advances may come from lighter radiators running at higher temperatures. The other main concern in orbit is radiation, which can cause random bit flips and whose impact over time can lead to a degradation of performance or malfunction. Recent work from Google and Starcloud, which has deployed a NVIDIA H100 chip in orbit, has given promising indications, but fault tolerance, error correction, redundancy (deliberate duplication of critical components or systems), and shielding are all required.

Any assembly or maintenance would pose a significant challenge. Heavy AI workloads can lead to relatively high chip failure rates, which, added to radiation effects, imply short lifespans of a few years. Depending on the concept, this would require redundancy or satellite replacement, with a weight or cost penalty, or else robotic maintenance in orbit, which still needs further development. Finally, there is the issue of communications. Large amounts of data from the ground, to be used for training, may simply be physically carried by ‘data shuttles‘, while server-side communications, needing high data rates, could use optical communication between satellites, in turn implying close proximity. Google’s plans, for example, envisage satellites hundreds of metres apart. With space debris and collisions being a critical issue, this would represent a major challenge in terms of the coordination of collision avoidance manoeuvres, which may be frequent given the sizes of the constellations being proposed.

Anticipatory policymaking

Deploying data centres in space poses important challenges, but does not appear to face insurmountable technical barriers and might be feasible even with current technology. The main hurdle is rather economic, with a mildly optimistic estimate placing near-future costs around three times those on the ground, although opinions are divided on whether such optimism is justified or not. Further innovation could help, with the evolution of launch costs a key determinant. This may lead to interesting synergies, with further technological and skills development benefiting other potential uses of space such as space-based solar power.

The current legal framework leaves space data centres in a grey zone: the United Nations’ Outer Space Treaty establishes no sovereignty in outer space, with launch states (a concept that presents its own issues) instead bearing responsibility and liability for space activities. Drafted in the 1960s, this treaty lacks explicit provisions regarding data. Article VIII of the treaty refers to jurisdiction over a space ‘object, and over any personnel thereof’, which has prompted some stakeholders to urge regulators to explicitly consider the concept of a ‘digital flag state’. Furthermore, relevant laws and treaties relying on the territorial location of data may require clarification. Examples include the GDPR‘s concept of transfers of personal data to third countries and the recently signed UN Convention against Cybercrime, which includes ships and aircraft but not satellites under its jurisdictional provisions. Likewise, legislation dealing with space activities may need to account for considerable processing of data originating from the ground rather than space. Extending the definition of space-based data and primary providers of space-based data in the Space Act, for example, could offer additional clarity. The overall situation is complex, involving potential multiple layers of overlapping jurisdiction. In the future, in-orbit assembly and AI agents risk further increasing this complexity. These issues highlight that extraterritorial application, as conceived in the GDPR or the Space Act, will be a crucial factor in the future regulation of space data centres.

The potential scale of orbital data centres is also important to consider. A 1 GW data centre, similar in scale to the largest under construction on the ground, could require a total payload upwards of 10 000 tons, or over three times the total payload mass launched in 2025. This risks potential infrastructure bottlenecks, such as the limited availability of launch facilities or liquid oxygen. It also raises sustainability questions, given that lifetime emissions may be larger than on the ground. Furthermore, the pollution of the upper atmosphere that would be caused by de-orbiting large numbers of end-of-life satellites is still poorly understood. Finally, it poses a critical, geopolitically relevant question regarding orbital congestion, as international regulation of slots in low Earth orbit is currently only done indirectly through radio spectrum assignment by the International Telecommunications Union, generally on a first-come, first-served basis.

What ifs are two-page-long publications about new or emerging technologies aiming to accurately summarise the scientific state-of-the art in an accessible and engaging manner. They further consider the impacts such technologies may have – on society, the environment and the economy, among others – and how the European Parliament may react to them. As such, they do not aim to be and cannot be prescriptive, but serve primarily as background material for the Members and staff of the European Parliament to assist them in their parliamentary work.

Read this ‘at a glance’ note on ‘What if AI data centres were put in space?‘ in the Think Tank pages of the European Parliament.

Il fut un temps où les Perses dirigeaient le plus grand empire que le monde ait connu jusqu'alors. Leur culture et leurs réalisations extraordinaires ont eu un impact sur le monde entier.

Donald Trump Junior était en visite mardi à Banja Luka, où il a rencontré de hauts responsables de la Republika Srpska. L'année dernière, le fils du président américain s'était rendu à Belgrade pour une réunion avec le président Vučić, mais qui paye ces voyages ?

- Le fil de l'Info / Bosnie-Herzégovine, RS sécession, USA Balkans, Courrier des Balkans

Donald Trump Junior était en visite mardi à Banja Luka, où il a rencontré de hauts responsables de la Republika Srpska. L'année dernière, le fils du président américain s'était rendu à Belgrade pour une réunion avec le président Vučić, mais qui paye ces voyages ?

- Le fil de l'Info / Bosnie-Herzégovine, RS sécession, USA Balkans, Courrier des Balkans

Le juge d’instruction de la deuxième chambre près le tribunal de Dar El Beïda a ordonné le placement en détention provisoire de trois mis en […]

L’article Alger : un employé communal placé en détention provisoire, voici les raisons est apparu en premier sur .

Very few people are ever prosecuted following terror attacks with suspects often held without trial.

Die Schadenfreude Europas darüber, dass es Trump nicht gelungen ist, einen Regimewechsel in Iran herbeizuführen, ist ebenso unklug wie verfrüht.