Monday, June 08, 2026

Latest Gallup Survey Shows Support for LGBTQ Issues Slipping

Good news! For far too long this  human fad, infatuation and obsession with LGBTQ has been going on!

Of course, one survey does not make a trend!

"Support for same-sex marriage, the morality of gay and lesbian relationships, and endorsement of gender changes are all down from their early 2020s highs. ....

The Values and Beliefs survey, conducted May 1–17, is based on telephone interviews with a random sample of 1,001 adults. ..."

"Story Highlights
  • Support for legal same-sex marriage down six points from 2022-2023 high
  • 62% viewing gay/lesbian relations as moral is lowest since 2016
  • Perceived morality of changing one’s gender down eight points since 2021
..."

Survey Shows Support for LGBTQ Issues Slipping





Armenia PM Pashinyan’s Ruling Party Wins Polls: Early Results

Good news, I suppose! Elections have consequences!

This is not a Radio Yerevan joke anymore! 😊 

Armenia PM Pashinyan’s Ruling Party Wins Polls: Early Results "Armenian Prime Minister Nikol Pashinyan’s party led early parliamentary election results on Monday, putting him on course to strengthen Armenia’s shift toward the West despite Russian pressure and allegations of interference. Pashinyan hailed a “historic victory” and vowed to pursue peace with Azerbaijan while maintaining ties with Moscow."

P.S. A supposedly Radie Yerevan joke: " Is it possible to drive a Moskvitch on a curve at 100 km/h? — In general yes, but only once"

Armenian Prime Minister and leader of the Civil Contract party Nikol Pashinyan holds a press conference following the parliamentary election at the party's headquarters in Yerevan early on June 8, 2026.


US adds Alibaba, BYD and other Chinese tech champions to military company blacklist

Serious stuff! How much are these Chinese companies controlled by the communist party of China?

"The Pentagon signalled on Monday that it was adding Alibaba, BYD, Baidu and dozens of other Chinese companies to its list of entities it says are linked to China’s military, widening a blacklist that increasingly targets sectors at the heart of US-China technological competition.
In a Federal Register notice scheduled for publication on Wednesday, the US Department of Defence designated a broad range of Chinese firms as “Chinese military companies” under Section 1260H of the National Defence Authorisation Act.
These include electric vehicle makers, artificial intelligence companies, battery manufacturers, biotech firms and solar suppliers. The designation can complicate companies’ access to US capital markets and government business, although it does not automatically trigger sanctions. ..."

US adds Alibaba, BYD and other Chinese tech champions to military company blacklist | South China Morning Post

Over 8.5 Million Ukrainians refugees Abroad Since 2014

This is huge! The current population of the Ukraine is estimated at about 39.5 million.

In 2014, Putin the Terrible annexed Crimea and President Obama and the rest of the West did not do much about it! Very regrettable and a huge mistake!

Over 8.5 Million Ukrainians Displaced Abroad Since 2014, Ombudsman Warns "Ukrainian Parliament Commissioner for Human Rights Dmytro Lubinets announced that 8.5 million Ukrainians have fled abroad since the onset of Russian aggression in 2014. Speaking at the Global Ukrainian Summit, Lubinets cited UN data showing that more than 5.7 million of these departures occurred after the 2022 full-scale invasion began. The Ombudsman called for the continuation of international temporary protection statuses until active hostilities end, while condemning illegal foreign seizures and adoptions of Ukrainian children."

Polish-Ukrainian startup develops radar to track elusive, low-flying drones

Good news! No escaping!

"Polish-Ukrainian anti-drone radar company Molfar Defence said it is developing a new generation of tactical radar systems for drone types that have doggedly managed to penetrate Ukrainian defenses.

The technology aims to find small, low-altitude drones, including those connected to operators via long command wires, a setup that makes them immune to defensive electronic warfare. ..."

Polish-Ukrainian startup develops radar to track elusive, low-flying drones




France to test its own AI-powered battlefield command in June NATO exercise

Vive la France! Which country is next!

"France will test its artificial intelligence-powered battlefield command system with allies during a NATO interoperability exercise this month, as an alternative to the Maven Smart System developed by Palantir Technologies, said Gen. Patrick Justel, deputy chief of the French Army staff.

The French have been developing the system with local companies including Mistral AI, Safran.AI, Thales and Airbus, Justel said in a media briefing on Thursday. The French Army has already tested the system, dubbed Arcadia, in exercises including Dacian Fall in Romania and Orion 26 in France. ..."

France to test its own AI-powered battlefield command in June NATO exercise

Microsoft AI Introduces MAI-Transcribe-1.5: 2.4% WER and Up to 5x Faster Long-Audio Transcription

Impressive! Really polyglott!

WER = word error rate

"... It transcribes a full hour of audio in under 15 seconds. Best-in-class on FLEURS accuracy. Leads the accuracy-speed Pareto frontier. ..."

"... The model handles 43 languages with a single system. It is optimized for diverse accents, dialects, and real-world acoustic conditions. ..."

Microsoft AI Introduces MAI-Transcribe-1.5: 2.4% WER on Artificial Analysis, Best-in-Class FLEURS Accuracy, and Up to 5x Faster Long-Audio Transcription - MarkTechPost "Microsoft's Superintelligence team has shipped MAI-Transcribe-1.5, a production-focused speech-to-text model with expanded language coverage, domain-aware keyword biasing, and faster long-form inference."

MAI-Transcribe-1.5 (official website)

Report on 41 PepsiCo’s driverless trucks on public roads in the US

Good news! Perhaps, there will be a driverless truck near you soon! 😊

"The on-time arrival performance of PepsiCo’s driverless trucks, after factoring out uncontrollable variables like weather and traffic, the company said. The 41 vehicles in Arizona, Texas and Arkansas transport products between bottling plants, storage facilities and stores, making PepsiCo the first major U.S. consumer-goods company to disclose the real-life, large-scale use of autonomous trucks on public roads."

Wall Street Journal What's news

Driverless Trucks Are Here—and They’re Delivering Bags of Doritos (behind paywall) "PepsiCo has 41 trucks on the road in Arizona, Texas and Arkansas, bringing the technology into the mainstream"

On permanence of private art collections

Nice advertisement! 

Source



President Trump struggled to rein in Israeli Prime Minister Benjamin Netanyahu’s strikes on Iran. Really!

I bet, the news media like here the Wall Street Journal are trying too hard and too much to create the impression of a dispute/disagreement between Trump and Netanyahu over Iran!

Makes for good headlines, but may not be a good representation of the real situation!

"Trump had hoped to contain the flare-up in fighting to keep it from disrupting work on a peace deal he’s trying to hammer out with Tehran. The U.S. and Israel’s priorities are diverging sharply, particularly over Lebanon. On another call this morning, after several rounds of back-and-forth attacks between Israel and Iran, Trump asked Netanyahu to end the attacks, people familiar with the matter said. The Israeli leader later publicly agreed to refrain from further strikes—unless Tehran launched new assaults on his country."

Street Journal What's news

Trump Struggled to Rein In Netanyahu’s Strikes on Iran (behind paywall)" Climbdown in phone call with Israeli leader shows president’s diminishing control over fighting"



Trump DOJ files to strip Jamaican fraudster, Indian H-1B scammer, Haitian pervert, and 14 others of US citizenship

Bravo President Trump!

For decades, Americans have been known worldwide to be very generous, lax and lenient when it comes to enforcing the law (especially immigration laws)! Too many criminals have benefitted from or taken advantage of it!

Then there is the strong American belief in the benefit of doubt!

"Federal officials filed a total of 305 denaturalization cases between 1990 and 2017 — an average of 11 per year. Like the first, the second Trump administration appears keen to make those previous numbers look like child's play.

The Trump Justice Department announced on Monday that it filed denaturalization actions in a handful of federal courts against 17 individuals accused of various crimes including child sex abuse and fraud. ..."

Trump DOJ files to strip Jamaican fraudster, Indian H-1B scammer, Haitian pervert, and 14 others of US citizenship | Blaze Media "The Trump administration is pushing to denaturalize scammers, degenerates, and other foreign-born duds who apparently illegally procured US citizenship."

On When AI builds itself by Anthropic

When superstition trumps common sense! Or is this naive, wishful thinking?

This recent blog post by Anthropic appears to have stirred up quite some public controversy and fears about the future of machine learning and AI!

I would say these public fears are quite overblown! So far humanity has managed other potentially dangerous, catastrophic technologies like nuclear weapons and biological weapons quite well! Why would or should it be different this time?

To halt or slow down AI development as suggested appears to be an extreme and rather unnecessary measure. Plus, such a halt or slowdown could e.g. not be verified worldwide with confidence!

"For most of AI’s history, humans drove every step in its development cycle. But at Anthropic, we are delegating a growing share of AI development to AI systems themselves, which is speeding up our work.

Taken far enough, and given enough compute, that trend points to an AI system capable of fully autonomously designing and developing its own successor. This is called recursive self-improvement. We are not there yet, and recursive self-improvement is not inevitable. But it could come sooner than most institutions are prepared for. ...

We believe it would be good for the world to have the option to slow or temporarily pause frontier AI development to enable societal structures and alignment research to keep up with the advance of the technology. ..."

When AI builds itself \ Anthropic

Action Alert: Opportunity for AI Pause by MIRI Institute (MIRI is a leading, ML & AI research institute in Berkeley, CA)




Hookworms have spent millions of years surviving inside a human host. Really!

Last time I checked, the earliest humans came about only roughly several hundred thousand years ago!

"“The hookworm has spent millions of years perfecting how to assure long-term survival inside a human host and how to get molecules out of its body and into ours,” said co-author of a new study Makedonka Mitreva in a statement. “We asked: What if we could add one more molecule to the roughly 1000 things the worm already secretes, something therapeutically useful to people?” ..."

ScienceAdviser



Makedonka Mitreva (Source) knows something about hookworms! 😊


How similar are the protein names of p16INK4a and p16-ARC? Hundreds of scientists were confused in more than 300 papers

An odd story! 

For lay people, these two names are not similar at all, I would presume! Maybe you have to be an egghead to confuse these two names! 😊

"Hundreds of scientists who study cancer and aging have deployed the wrong antibody to test for a key protein, according to a researcher who exposes errors in the biomedical literature. Instead of antibodies that recognize p16INK4a, a tumor suppressing protein that may also promote aging, these researchers used versions that tag the similarly named protein p16-ARC, which helps shape the cell’s molecular skeleton.

The gaffe appears in more than 300 papers, including some published by top journals such as Nature, Nature Medicine, Cancer Cell, eLife, and Science Advances, reports Sholto David, a molecular biologist at the U.K. biotech OXB and a part-time error hunter. David’s latest revelations, posted on 2 June on the blog For Better Science, have sent researchers digging through old lab notebooks and, in at least one case, back to the lab to rerun experiments. ..."

ScienceAdviser

Chemists design impact-resistant plastics for shoe soles, tires and other products

Amazing stuff! In my next life I'll become a chemist! 😊 

What about more bounce and springiness for soles?

"With help from a novel cross-linking molecule, MIT chemists have shown they can substantially improve the ballistic impact resistance of common polymers, including polystyrene and a type of rubber used to make shoe soles. ...

To make the polymer more resistant to sudden impact, the MIT team added weak bonds scattered throughout the material as cross-links, which allows the material to dissipate energy much more effectively under deformations. When struck by a projectile, these weak bonds selectively break at the site of impact to open up pathways for enhanced energy absorption.

The researchers found that this approach can also fortify styrene-butadiene-styrene rubber, and they are now investigating whether it will also work for other types of polymers such as latex or the rubber that is used to make tires.  ..."

From the abstract:
"Mechanical failure is a marked limitation for plastics used in structural, protective and coating applications. In particular, perforation under high-rate deformation is difficult to mitigate through conventional molecular design.
Cross-linking is widely used to improve the thermal and chemical stability of polymers, yet under mechanical deformation, it typically renders materials more brittle, limiting impact resistance and functional lifetime. Overcoming this fundamental trade-off between stability and toughness remains a central challenge. Here we demonstrate that embedding a small fraction of force-sensitive mechanophores as cross-links into common polymers fundamentally reverses this trade-off, producing materials with substantially enhanced ballistic energy dissipation.
At strain rates exceeding 107 s−1, we show that mechanophore-cross-linked networks absorb up to about 115% more energy than conventional thermosets and surpass even their uncross-linked thermoplastic counterparts.
We attribute this behaviour to a force- and adiabatic-heating-driven local thermoset-to-thermoplastic transition, in which selective mechanophore scission facilitates viscoplastic deformation at the impact site while preserving network integrity in the surrounding regions.
We demonstrate the generality of this strategy in both glassy polystyrene and rubbery styrene–butadiene–styrene triblock copolymers.
These results establish mechanophore cross-linking as a design principle for converting commodity polymers into impact-resilient materials and open directions at the intersection of polymer mechanochemistry and extreme-strain-rate material behaviour."

MIT chemists design impact-resistant plastics | MIT News | Massachusetts Institute of Technology "Introducing weaker bonds into polystyrene and rubber helps these materials dissipate energy, making them more resistant to destructive forces."






English for trippers: A gene, a game

A genie!

Chinese President Xi Jinping pledges ‘unwavering’ support for North Korea and Kim Jong-un. Really!

Too bad! The 72 years old Xi Jinping missed a great opportunity to go down in history as a great leader by not suggesting a reunification between South and North Korea.

Developing | Chinese President Xi Jinping pledges ‘unwavering’ support for North Korea and Kim Jong-un | South China Morning Post "The Chinese leader began his first foreign trip of the year by pledging that the ‘traditional friendship’ between the countries will not change"




Disclaimer

Since end of February, I  am blogging from behind the Great Firewall of China.

My Internet service in China is very spotty. Thus, I am not able to blog as usual.

Sunday, June 07, 2026

Nach Auto, Stahl und Energie streichen auch Pharma-Konzerne Investitionen in Deutschland

Wieder nur ein paar Einzelfälle oder wieder mal nur Medien Hysterie?

Nach Auto, Stahl und Energie streichen auch Pharma-Konzerne Investitionen in Deutschland | NIUS "Für die Regierung in Berlin gilt es als Landespolitik. Doch es ist ein weiteres Zeichen dafür, was in Deutschland gerade passiert. Zwei große Pharmakonzerne streichen Milliardeninvestitionen in Rheinland-Pfalz. Pharma gehört zu den Top-5-Branchen der deutschen Exportindustrie. Jetzt hat es also nach Auto, Stahl, Chemie und Energie auch diese Branche getroffen: Abbau und Abzug."

On MetaWorld: Scaling Multi-Agent Video World Model from Single-view Video Data

This could be an interesting new research paper by Dacheng Tao and his team!

From the abstract:
"Video world models are a foundational generative technology for embodied AI and the Metaverse, yet existing approaches are inherently limited to a single agent observing from a single perspective.
Extending these models to multi-agent settings introduces two critical challenges:
data scarcity (coordinated multi-view recordings are prohibitively expensive to collect for general open-domain scenarios) and
world state alignment (independently generated video streams cannot ensure that shared physical environments and events evolve consistently across views).
To address these challenges, we propose MetaWorld, a novel framework that scales multi-agent video world models to open-domain environments directly from single-view videos.
First, we introduce Monocular World-State Unrolling (MWSU) to explicitly decompose monocular footage into the camera operator's ego-motion and the visible subject's spatial trajectory. This camera-trajectory decomposition naturally extracts synchronized multi-agent motion data within a shared 3D space, completely bypassing the need for multi-camera setups.
Second, for precise visual control, we develop the Subject-Aware World Generator to enable appearance-driven simulation conditioned on per-agent identity images.
Finally, to ensure both views are grounded in the identical physical reality, we propose World-State Alignment, a per-frame inter-branch cross-attention mechanism inserted at every transformer layer of the video DiT.
By jointly synchronizing the denoising process, WSA enforces both static geometric consistency and dynamic motion consistency, encouraging that the shared 3D environment and physical events remain well-aligned across both egocentric views.
Extensive experiments demonstrate that MetaWorld achieves superior cross-view consistency and identity fidelity, establishing a highly scalable, physics-driven paradigm for multi-agent video world modeling."

[2606.02753] MetaWorld: Scaling Multi-Agent Video World Model from Single-view Video Data




English for trippers: A stale state

Is it stable or a stalemate?

On Edge of Stability Selectively Shapes Learning Across the Data Distribution

This could be an interesting research paper by Tomaso Poggio and his team!

From the abstract:
"Existing analyses of the edge of stability (EoS) treat it as a global property of optimization.
We show that it is also selective: the stability constraint redistributes learning across subsets of the training distribution, amplifying progress on some groups while suppressing progress on others.
Using a branching intervention that enters or exits the EoS regime from the same training state, we causally demonstrate this trade-off and identify two necessary conditions for a group to benefit.
First, its aggregate gradient must align with the top Hessian eigenvector. We isolate this mechanism with a controlled perturbation that preserves distance but randomizes direction, destroying alignment and eliminating the advantage.
Second, the group must sustain non-vanishing gradient magnitude over time. Under cross-entropy loss, gradient saturation decouples confidently classified groups, shifting the advantage to output-outliers, whose gradients persist.
Together, these results show that EoS functions not only as a stability boundary, but as a mechanism governing the allocation of learning across the data distribution."

[2606.04212] Edge of Stability Selectively Shapes Learning Across the Data Distribution




Sorry Google Scholar

This is what I got this morning when I tried to look up a preeminent ML & AI researcher Quoc V. Le) on Google Scholar from behind the Great Firewall of China using SetupVPN.




On LEAP: Supercharging LLMs for Formal Mathematics with Agentic Frameworks

Google with Quoc V Le and hist team is bringing mathematics into the age of AI! Expect a lot of progress regarding mathematics in coming years thanks to ML & AI!

Exciting times!

From the abstract:
"Large Language Models (LLMs) exhibit strong informal mathematical reasoning but struggle to generate mechanically verifiable proofs in formal languages like Lean.
We present LEAP, an agentic framework that enables general-purpose foundation models to achieve state-of-the-art performance on automated formal theorem proving.
LEAP leverages foundation model capabilities, such as informal reasoning, instruction following, and iterative self-refinement. By decomposing complex problems into smaller units, the system bridges formal proof construction with informal blueprints through continuous interaction with the Lean compiler.
To provide a rigorous evaluation beyond increasingly saturated benchmarks, we introduce Lean-IMO-Bench, a benchmark of IMO-style problems formalized in Lean, with short statements yet highly non-routine and multi-step proofs across a wide range of difficulty levels.
Empirically, on the latest 2025 Putnam Competition, an annual mathematics competition for undergraduate students in North America, LEAP solves all 12 problems, matching recent breakthroughs by frontier formal mathematical models. On Lean-IMO-Bench, LEAP boosts the one-shot formal solve rate of general-purpose LLMs from below 10% to 70%, notably surpassing the 48% benchmark set by a specialized, gold-medal-caliber IMO system.
Furthermore, we demonstrate LEAP's research-level utility by autonomously formalizing complex proofs for open combinatorial challenges, including a verified proof for a key subproblem in Knuth's Hamiltonian decomposition of even-order Cayley graphs."

[2606.03303] LEAP: Supercharging LLMs for Formal Mathematics with Agentic Frameworks




A Supreme Court Judge Branded India’s Youth as ‘Cockroaches’—and Unleashed a Political Movement

Sometimes words have consequences!

"The chief justice of India’s Supreme Court at a May hearing referred to many young people as “cockroaches” who don’t work. He inadvertently spawned a political movement of young Indians frustrated by high unemployment and low wages that has embraced the moniker and adopted it as the name of a new party."

A Judge Branded India’s Gen Z ‘Cockroaches’—and Unleashed a Political Movement - WSJ "Frustrated by low wages and a lack of jobs, the country’s youth are pushing for change"




Saturday, June 06, 2026

David Barnea shares Mossad role in Nasrallah assassination, Iran war details

Don't mess with Mossad at your own peril!

"On Tuesday, David Barnea completed his astounding five-year term as Mossad chief. During that time, he transformed the spy agency from a unit that carried out one or two operations in the shadows at a time, to a juggernaut that could influence the course of war and peace on multiple Middle Eastern fronts – from Iran to Hezbollah in Lebanon, to protecting Jews from terrorism overseas.

He has spent decades in the agency, with five years at its helm, including some of the most crucial in Israel’s history. He helped hobble Iran with hundreds of agents, sabotaging the country in June 2025. There were the exploding beepers in September 2024, which killed Hezbollah operatives. That same month, he oversaw the assassination of Hezbollah chief Hassan Nasrallah in Lebanon. ..."

David Barnea shares Mossad role in Nasrallah assassination, Iran war details | The Jerusalem Post "After decades in the agency, and five years at its top, including some of the most crucial in Israel's history, Barnea also has some critical viewpoints which Israel and the world can learn from."


Head of Mossad David Barnea attends a ceremony held at the Yad Vashem Holocaust Memorial Museum in Jerusalem, as Israel marks the annual Holocaust Remembrance Day. April 23, 2025.


Chinese Sodium-ion batteries could become a low-cost rival to Tesla's batteries

Good news! We need more cheap and better batteries! Hina from China!

I also just blogged here in German about this story!

"A popular sodium-ion battery designed by the company Hina and used in cars and large-scale energy storage systems in China matches performance parameters and production quality of Tesla's lithium-ion batteries, finds new research published in Cell Reports Physical Science. ..."

From the highlights and abstract:
"Highlights
Sodium-ion cell performance rivals established lithium-ion batteries
High capacity retention observed at extreme charge rates and low temperatures
• Unique NaCu1/9Ni2/9Fe1/3Mn1/3O2 cathode composition identified in sodium-ion cells
• Archimedean spiral method accurately estimates jelly roll electrode length

Summary
Transparent insight into the design and materials of commercial sodium-ion batteries is essential for assessing their maturity and benchmarking against lithium-ion technologies.
We report postmortem analysis and electrochemical testing of the most commercialized sodium-ion battery technology from Hina Battery, which employs a tabless double-aluminum architecture and a novel cathode composition, NaCu1/9Ni2/9Fe1/3Mn1/3O2. This cathode features a distinctive spatial separation of copper from the other transition metals (Ni, Fe, and Mn) within individual particles.
Using an Archimedean-spiral analysis of computed tomography data, we estimate total cathode length with 2% deviation, without additional image optimization.
The cell delivers performance and production quality comparable to state-of-the-art Li-ion batteries: impedance variation across 120 cells is 5.3%, capacity remains over 100% at 4C (25°C), and usable discharge energy above 80% at −20°C, though this value drops to 56% in the charging direction."

Sodium-ion batteries could become a low-cost rival to Tesla's batteries

Cell teardown and characterization of a Hina commercial sodium-ion battery (open access)

Credits: Akku für E-Auto & Co.: Natrium statt Lithium? RWTH Aachen testet | FAZ (behind paywall) "Natrium statt Lithium: Eine neue Batterietechnologie könnte die Rohstoffabhängigkeit beenden. Forscher von der RWTH Aachen haben ein aktuelles Modell aus China zerlegt – und waren überrascht von der technischen Reife."


Figure 7 SEM and EDX of anode and cathode
Morphological and elemental analysis of sodium-ion battery electrodes.


Houston fiercest, most fabulous drag queens

Showtime in Houston! Texas, more than longhorns, cowboys, and crude oil! 😊

Houston, we have a problem! Drag queens everywhere! Just kidding!

"Houston's drag scene is one of the most dynamic and diverse in the country. We have it all, from pageant to comedy to dancing queens. Seven days a week, you can see your favorite queen at karaoke, bingo, brunch or your favorite club. ..."

Houston fiercest, most fabulous drag queens




New propulsion system could make tiny satellites both fast and fuel-efficient

Amazing stuff!

"MIT engineers are testing a new propulsion system that combines the power and speed of conventional chemical thrusters with the precision and fuel-efficiency of electrical thrusters. 

The system could enable the design of nimbler, more flexible small satellites, which could perform both fast, powerful maneuvers and slower, precise adjustments, depending on the mission and moment at hand.

The key to the new system is a special propellant that can power both chemical and electrical thrusters, which traditionally have required separate, bulky fuel sources.  ..."

New propulsion system could make tiny satellites both fast and fuel-efficient | MIT News | Massachusetts Institute of Technology "For satellites as small as a briefcase, getting around in space just got a whole lot easier."

On Graph Neural Networks Are Not Continuous Across Graph Resolutions

This could be an interesting, new paper by Michael Bronstein and his team!

Caveat: I did not read the paper.

From the abstract:
"We show that contrary to conventional wisdom in the community, graph neural networks (GNNs) are not continuous with respect to all natural modes of graph convergence.
As a result, GNNs may generate substantially different latent representations for graphs that are very similar.
In particular they assign vastly different latent embeddings to graphs that represent the same underlying object at different resolution scales.
We trace this failure of continuity back to a structural obstruction arising from commonly used information-propagation schemes.
Building on this insight we then derive a principled modification to standard GNN architectures which equips models with continuity across scales. The proposed modification enables consistent integration of distinct resolutions and reliable generalization between them. We systematically validate our theoretical findings in a wide range of numerical experiments."

[2605.31315] Graph Neural Networks Are Not Continuous Across Graph Resolutions




'World-first' vaccine designed by artificial intelligence to provide broad protection from thousands of variants of viruses

Amazing stuff!

"Artificial intelligence has been used to develop a "fundamentally new" type of vaccine that could protect against large swathes of viruses and prevent pandemics, say researchers.

The team at the University of Cambridge say it is the first time a vaccine's key component has been designed entirely by AI and then trialled in people.

The vaccine was engineered to work on all coronaviruses which would include all Covid variants and viruses that infect animals, but could start the next pandemic. ..."

"The first human clinical trial of a universal Sarbeco coronavirus vaccine, developed by the University of Cambridge and spin-out DIOSynVax (DVX) Ltd, has shown that the vaccine is safe and has no significant side-effects. ..."

From the highlights and abstract:
"Highlights
• pEVAC-PS is a novel, needle-free DNA vaccine targeting all sarbecoviruses
• pEVAC-PS was safe and well tolerated in this phase I study
• Followup Phase 2 to provide data on the breadth and durability of immunogenicity to pEVAC-PS

Summary
Background
Coronaviruses such as SARS, SARS-CoV-2 and related Sarbeco-Coronaviruses continue to pose global health threats, underscoring the need for vaccines capable of inducing broad cross-sarbecovirus protection.
The pEVAC-PS vaccine was developed using Digitally Immune Optimised Synthetic Vaccine (DIOSynVax) technology and pre-clinically selected for the ability to induce broadly protective immune responses across the Sarbecoviruses including SARS, SARS-CoV-2, and related viruses representing potential zoonotic spillovers. For this first-in-human study, the antigen was delivered as a DNA vaccine to enable thermostability and needle-free intradermal administration to support future deployment in resource-limited settings.

Methods
This open label phase I dose escalation study investigated the safety, tolerability and immunogenicity of the pEVAC-PS vaccine candidate against SARS, SARS-CoV-2 and related Sarbeco Coronaviruses via needle-free intra-dermal delivery using the PharmaJet Tropis Device. ...

Findings
Between December 2021 and September 2023, a total of 39 volunteers were vaccinated. The vaccine was well tolerated at all four doses with no significant safety concerns elicited. Interpretation of immunogenicity outcomes was influenced by high baseline antibody levels and heterogeneous exposure histories due to ongoing waves of Omicron variant infections during recruitment, which differed across dose-escalation cohorts and introduced unavoidable immune bias.

Interpretation
Needle-free intradermal delivery of this novel computationally designed PanSarbeco vaccine was safe and well tolerated. Although immunogenicity was modest in the context of substantial pre-existing immunity, participants developed measurable responses to conserved, vaccine-encoded sarbecovirus epitopes, supporting the feasibility of this antigen design strategy."

'World-first' vaccine designed by artificial intelligence

New ‘universal vaccine’ technology could protect us from future virus outbreaks (original news release) "A Cambridge-led team has developed a way to engineer better vaccines that could provide broad protection from thousands of variants of viruses - such as coronaviruses or Ebola - in a single vaccine. This represents a fundamental new vaccine technology that could prevent future pandemics before they begin."



Fig. 2 Vaccination and SARS-CoV-2 variant timeline. The timeline plot shows the vaccinations periods for each group in relation to the dominant SARS-CoV-2 variant circulating in the UK at the time.


On Using Reward Uncertainty to Induce Diverse Behaviour in Reinforcement Learning

This could be an interesting paper by Doina Precup and his team!

From the abstract:
"Classical reinforcement learning (RL) typically seeks a deterministic policy that maximizes the expected sum of a scalar reward. Yet, modern applications such as language model fine-tuning or scientific discovery demand diversity.
Existing remedies such as entropy regularization or diversity bonuses often require fragile trade-offs that sacrifice performance for stochasticity or rely on heuristic metrics that can misalign policy rankings. We argue that diversity is more naturally understood as the rational response to uncertainty in the reward. When the reward function is not perfectly known--as is the case with ambiguous preferences or imperfect reward models--committing to a single action can be sub-optimal.
Building on this, we propose a fundamental reformulation of the RL objective by replacing the scalar reward with a distribution over reward functions, and applying a non-linear objective over sets of actions.
The result is a framework in which calibrated behavioural diversity emerges naturally, remains controllable through the reward function distribution, and is obtained without sacrificing expected reward.
Focusing on the contextual bandit setting, we derive a principled gradient estimator for this objective and prove that our formulation naturally generalizes both vanilla policy gradient and more recently developed action-set approaches. Our empirical results demonstrate that this framework offers a robust and theoretically grounded alternative for complex RL tasks where the traditional formulation of the problem fails to induce the desired breadth of agent behaviour."

[2606.03962] Using Reward Uncertainty to Induce Diverse Behaviour in Reinforcement Learning




Google annoyances again

Just tried to look up Doina Precup (a well known machine learning and AI researcher) on Google Scholar. Here is what I got:


 Working from behind the Great Firewall of China is not easy despite VPN!

Satellite maps of sinking coastlines are contradictory and come under scrutiny

When scientists believe in the global warming/climate change hoax, their impartiality is seriously affected! Also called confirmation bias!

"But satellite radar, a popular tool for identifying fast-sinking, vulnerable shorelines, may be leading scientists astray, according to a comparison of two influential studies of subsidence along the U.S. Gulf Coast. The studies reached strikingly different conclusions about which areas are sinking, according to the analysis, published today in Earth Observation. ..."

From the abstract:
"Interferometric Synthetic Aperture Radar (InSAR) is widely used to monitor surface-elevation change in subsiding coastal regions, but inconsistencies between studies hinder understanding of the processes driving vertical land motion (VLM). Here we compare two recent InSAR datasets from the central U.S. Gulf Coast which yield similar mean rates (−2.8 ± 2.8 and −3.3 ± 1.8 mm yr−1) but show negligible spatial correlation (R2=0.05), except in medium to highly developed urban areas (R2>0.5). Using 41 Global Navigation Satellite System records from adjacent Pleistocene uplands with minimal shallow subsidence and sediment accretion, we find a median VLM of −1.2 mm yr−1, largely driven by glacial isostatic adjustment that is higher than previously believed. InSAR data exhibit larger uncertainties and are presently unable to capture this rate.
Given the struggles of InSAR in vegetated landscapes, we recommend that vertical velocities below 5 mm yr−1 are interpreted with utmost caution."

Satellite maps of sinking coastlines come under scrutiny | Science | AAAS



Figure 1Maps of InSAR-derived surface-elevation change (SEC) rates and corresponding frequency histograms.


On No More K-means: Single-Stage Sparse Coding for Efficient Multi-Vector Retrieval

This could be an interesting paper by Stefanie Jegelka and her team!

Caveat: I did not read the paper.

From the abstract:
"Multi-vector retrieval (MVR) models, exemplified by ColBERT, have established new benchmarks in retrieval accuracy by preserving fine-grained token-level interactions. However, this granularity imposes prohibitive storage and retrieval efficiency bottlenecks: to manage the immense memory footprint and computational overhead of billion-scale token vectors, state-of-the-art systems are forced to rely on aggressive dimension reduction and complex clustering (e.g., K-means).
This compromise introduces two critical limitations: excessive indexing latency of clustering large-scale corpora and semantic information loss inherent to compression. In this paper, we propose Single-stage Sparse Retrieval (SSR}, a paradigm shift that replaces expensive clustering with efficient sparse coding. Instead of compressing features into low-dimensional dense vectors, we utilize Sparse Autoencoder (SAE) to project token embeddings into a high-dimensional but highly sparse representation. This transformation enables us to bypass vector clustering entirely and leverage inverted indexing for precise, high-throughput retrieval. Extensive experiments on the BEIR benchmark demonstrate that SSR achieves a "trifecta" of improvements: it reduces indexing time by 15x compared to ColBERTv2, halves retrieval latency, and simultaneously improves retrieval performance over leading baselines."

[2605.30120] No More K-means: Single-Stage Sparse Coding for Efficient Multi-Vector Retrieval




Surprisingly intelligent bumble bees

Amazing stuff!

"In a new study bumble bees solved a completely novel object-manipulation task. What makes this behaviour especially remarkable is that the bees had never been trained. The findings challenge the long-standing assumption that spontaneous problem-solving is restricted to humans and other large-brained vertebrates.

More than 100 years ago, psychologist Wolfgang Köhler famously showed that chimpanzees could solve novel problems by suddenly combining objects in new ways, such as stacking boxes to reach an out-of-reach banana. These experiments became some of the earliest and most influential demonstrations of insight and spontaneous problem-solving in animals.

Now, researchers from the University of Oulu, the University of Helsinki, and the University of Turku report strikingly similar problem-solving abilities in bumble bees. ..."

From the editor's summary and abstract:
"Editor’s summary
Recent research has revealed that bumble bees are much more cognitively advanced than previously thought: They play with balls, count, recognize faces, and even feel rhythm. However, it has not been shown that they could achieve one of the highest peaks of cognitive performance: the ability to spontaneously solve a problem. Bhambore et al. tested this ability by providing bees with a ball that could be used as a tool to reach an otherwise unreachable flower reward. Bees that had been allowed to play with a ball and experience the flower spontaneously learned to move the ball to access the flower when they were present together.

Abstract
Problem-solving using novel solutions without explicit training is often considered a hallmark of cognitive flexibility. We investigated whether bumble bees (Bombus terrestris) could solve a novel object manipulation task spontaneously.
Bees trained to associate a blue ring (“flower”) on the floor with a reward successfully moved a ball underneath a flower relocated to the ceiling to reach the flower. In control experiments in which the flower was out of sight when ball movement began and remained hidden during transport, bees still succeeded in the task. These results suggest that these were goal-directed actions rather than reinforcement-based associations driven by perceptual feedback. Our findings provide evidence that bumble bees can exhibit spontaneous problem-solving, challenging the notion that such advanced cognitive abilities are exclusive to large-brained vertebrates."

In Science Journals | Science



A bumble bee standing on a ball beneath the artificial flower containing reward, illustrating the experimental solution.


Friday, June 05, 2026

UK Prime Minister kneels for a lifelong, petty criminal and drug addict who tried to fake an emergency before arrest. Really!

Incredible! What a fool and idiot! Appalling! God save the UK!

UK quo vadis! Thatcher and Churchill are spinning in their graves!

Credits: Keir Starmer kniete für George Floyd, schwieg aber monatelang zu Henry Nowak

PM Keir Starmer and deputy Labor leader Angela Rayner (Source)



Zelensky Issues Open Letter to Putin the Terrible: ‘Enough War – Let’s Meet to End It’

Good news! Worth a try after so many years and so many war casualties!

Don't expect Putin the Terrible to become an angel of peace!

"President Volodymyr Zelensky proposed a face-to-face meeting with Russian President Vladimir Putin in a rare open letter to the Russian leader Thursday, saying he was also ready for a “full ceasefire.”

The letter was released by the Office of the President of Ukraine, amid recent Ukrainian long-range drone strikes near the opening of a major Russian economic forum in St. Petersburg.

“Ukraine proposes ending this war through direct engagement between us – and you. I am proposing a meeting,” Zelensky said in the letter, adding that “Ukraine is ready for a full ceasefire for the duration of the negotiations.” ..."

‘Enough War’: Zelensky Calls Putin to Direct Talks and Monitored Ceasefire "President Volodymyr Zelensky has publicly addressed Russia’s leader, arguing that 26 years of Putin’s rule have turned bilateral relations into a ledger of strikes and losses, highlighting Russia’s mounting casualties and economic strain. He offered a path to end the war through a bilateral summit hosted by a neutral state, along with fully swapping prisoners and broader security guarantees from Europe and the US."

Akku für E-Auto & Co.: Natrium statt Lithium? RWTH Aachen testet Chinesische Batterie

Hat man nicht vor früher China der Industriespionage bezichtigt?

P.S. Ich konnte leider keine anderen Artikel dazu finden (inkl. RWTH Aachen website).

Akku für E-Auto & Co.: Natrium statt Lithium? RWTH Aachen testet | FAZ (behind paywall) "Natrium statt Lithium: Eine neue Batterietechnologie könnte die Rohstoffabhängigkeit beenden. Forscher von der RWTH Aachen haben ein aktuelles Modell aus China zerlegt – und waren überrascht von der technischen Reife."

Typische Männer- und Frauenberufe. Wirklich!

Gleichberechtigung hat nichts mit Gleichverteilung zu tun! Was ist daran auszusetzen?

Im Laufe der Zeit (innerhalb der letzten Jahrzehnten) haben sich bereits die Anteile verschoben. 

An der zum Teil extremen Ungleichverteilung in verschiedenen Berufen ist grundsätzlich nichts auszusetzen, wenn sie auf freiwilliger Berufswahl beruht.

Typische Männer- und Frauenberufe - iwd.de "Ob in der Ausbildung oder im Studium – Frauen und Männer entscheiden sich jeweils nach wie vor häufig für bestimmte Berufe. Maßnahmen, um die Geschlechterklischees bereits in der Schule aufzubrechen, sind auch aus wirtschaftlicher Perspektive begrüßenswert."





With over 1600 signatories, mathematicians issue a declaration warning of AI. Really!

I bet, some mathematicians were also complaining when the Abacus was introduced more than 4,000 years ago! 😊 Same old story!

Of course, machine learning and AI is coming to mathematics, the queen of sciences (according to a famous quote from Carl Friedrich Gauss). I am already curious what ML & AI will contribute to mathematics. Apparently, it has already provided some proofs to some very old mathematical problems.

It is probably safe to largely ignore these warnings!

"Few communities have felt that pressure more acutely than mathematicians, who have haplessly watched AI get frighteningly smart, frighteningly fast.

Today, 16 math specialists have turned that unease into a public cry for help—and call to action. Part warning, part manifesto, the 11-page Leiden Declaration on Artificial Intelligence and Mathematics cautions that unchecked automation threatens not only how math is practiced, but what the discipline stands for. It also lays out principles for using AI in ways that support, rather than erode, the field. ..."

Mathematicians issue warning as AI rapidly gains ground | Science | AAAS

Leiden Declaration on Artificial Intelligence and Mathematics "This declaration calls for action to address the challenges posed by the use of artificial intelligence within mathematics research. It is the result of a community initiative and is endorsed by the International Mathematical Union (IMU)."

Brazil plans to buy 20 more Swedish Gripen fighter aircraft

Good news for Sweden and the European defense industry!

"Brazilian Defense Minister José Múcio Monteiro Filho announced today that he wants to buy an additional 20 Gripen E jets, which would bring Brazil’s fleet to 56 fighters.

The announcement came during a press conference today, wrapping up the minister’s four-day visit to Sweden. ...

After Saab beat out the Boeing F/A-18 and Dassault Rafale in 2014, Brazil ordered a total of 36 Gripen fighters: 28 single-seat Gripen E models and eight two-seat Gripen F variants. Eleven aircraft have been delivered to date. ..."

Brazil plans to buy 20 more Gripen fighters - Breaking Defense "Gen. Marcelo Kanitz Damasceno, the Brazilian air chief, told Breaking Defense that the need for extra jets was born from a serious look at Brazil’s air power."


Saab Gripen


Habits form far faster than science previously thought, research shows

Amazing stuff! As they say we are creatures of habit! 😊

I remember, there were rumors and remarks decades ago that it takes only one cigarette to become addicted. It was meant as deterrence!

"... Scientists have long believed that habits emerge gradually after long periods of repetitive behavior. But the new research shows that the transition into habitual action occurs faster than previously understood. And the research suggests that a particular brain region may play a key role in the transition—a discovery that could point to ways to alter entrenched habits.

Key Takeaways
  • The formation of habits may not transpire gradually as science has long believed.
  • New research show that the transition to habitual behavior occurs far faster than previously understood.
  • There may be a brain region that plays a key role in that transition, a discovery that could point to ways to alter entrenched habits.
"For over 100 years the theory of how habits form has been one of gradual strengthening and repetition ..."

From the abstract:
"The speed of goal-directed to habit transitions has been debated since Clark Hull asked in 1943: is habit formation slow or sudden?
To address this, male mice were given home-cage access to citric-acid water that reduced—without eliminating—reward-seeking for plain water in an auditory go/no-go task. Animals learned to discriminate quickly but exhibited ongoing state-like fluctuations in engagement.
Strikingly, these fluctuations abruptly ceased (transition) long after discrimination stabilized, with HMM-GLM modeling pinpointing a ~3-trial transition.
We confirmed this as a goal-directed to habit transition using sensory-specific outcome devaluation, DLS lesions, motor stereotypy, and pupillary responses. Dual-site fiber photometry showed equally abrupt DLS dynamics at the transition: outcome-related activity dropped and stimulus-response activity sharpened, suggesting a switch-like mechanism that recruits a readily available habit circuit rather than gradual changes across a threshold.
Thus, habits can emerge suddenly, mediated by an abrupt dorsostriatal shift from outcome- to stimulus-driven processing."

Habits form far faster than science previously thought, research shows | Hub "A Johns Hopkins is team is studying whether a region of the brain plays a key role in developing habitual behavior, a discovery that could point to ways to alter entrenched habits"



Fig. 2: Abrupt state-like transitions from goal-directed to habitual behavior appear spontaneously within individual animals.