Lifeboat Foundation: Safeguarding Humanity
Toggle light / dark theme

Blog

Get the latest international news and world events from around the world.

Log in for authorized contributors

General-Sep29

The guidelines were published on Sept. 29, but OpenAI chose to announce proofs for the 722 problems anyway.

OpenAI proceeded with the release format the guidelines warned against (a headline drop generated by a closed model), but structured the drop itself to comply with many of the documentation and formalization standards laid out in those guidelines. https://agmai.org//

About the guidelines:

AI laboratories have begun using powerful artificial intelligence models to solve complex mathematical problems. However, this is creating a dilemma: AI systems can produce advanced proofs that humans do not yet understand, cannot easily verify, or cannot take responsibility for.

Historically, mathematics relies on human understanding, peer review, and open collaboration. To protect this foundation, a group from the mathematical community—backed by feedback from over 600 mathematicians—has issued guidelines on how AI labs should responsibly handle and share mathematical discoveries.

Core Principles.

* Prioritize Human Understanding: The ultimate goal of mathematics is human insight. AI discoveries shouldn’t remain black boxes.

* Lab Responsibility: If an AI lab releases a major math discovery that humans don’t yet understand, the lab is responsible for funding and supporting the effort to help human mathematicians understand it.

Continue reading “General-Sep29” | >

A quantum state of mitochondria in the living cell

Mitochondria are the power plants of the cell. New research suggests they might also be quantum machines.

Scientists have long wondered how living cells manage to produce energy so efficiently—something that’s hard to explain using only classical physics. Many researchers have suspected that quantum mechanics (the strange physics of tiny particles) might play a role, but proving this in living cells has been difficult because of a lack of direct experimental evidence.

In this study, the researchers combined laboratory experiments on living cells, tissues, and mitochondria (the tiny “power plants” inside cells) with a theoretical model. They discovered a special vibration occurring at a frequency of 71.0 terahertz (THz) that appears *only* in living cells and tissues—not in dead or disrupted ones. This vibration depends on mitochondria having an intact, healthy structure, and it doesn’t come from any single molecule.

To explain this, the team built a quantum model describing how light interacts with matter inside mitochondria. Their calculations suggest that a kind of quantum “superposition” state forms inside working mitochondria, created by the coupling of light with lipid (fat) molecules in the folded inner membranes of the mitochondria. This coupling splits a natural vibration of those lipids (at 87 THz) into two new levels—one at 71 THz and one at 103 THz. The 71 THz signal is the one seen only in living cells, while the 103 THz signal gets lost among other vibrations from biomolecules and water, making it impossible to detect separately.

Further experiments showed that this quantum state acts like an efficient control channel for regulating ATP production—ATP being the molecule cells use as fuel. In short, the findings offer a quantum-level explanation for how living cells work, and raise the intriguing possibility that this quantum state might serve not only as a channel for energy metabolism but perhaps even for transmitting information in living systems.

Bottom libe.

The core experimental findings—the 71 THz signal depending on intact mitochondrial structure, and the frequency-specific ATP response—are interesting and worth pursuing. The quantum interpretation is a model that fits the data, not a proven mechanism. Independent replication, direct measurement of energy transfer dynamics, and testing in more complex biological systems are the necessary next steps before any of the grander implications can be taken seriously.

#quantumbiology #mitochondria #ATP #quantummechanics

Continue reading “A quantum state of mitochondria in the living cell” | >

Decline of chaperone-mediated autophagy in aging impairs macrophage clearance of senescent cells

This study looks at how the decline in chaperone-mediated autophagy in aging alters senescent cell properties leading to an impairment in their immune clearance. Utilizing proteomic and metabolomic analyses, the authors show that restoring autophagy enhances macrophage function and reduces senescent cell accumulation.

NASA to Cover Northrop Grumman CRS-24 Spacecraft Departure

After delivering more than 11,000 pounds of supplies, science experiments, and other cargo to the International Space Station for NASA, Northrop Grumman’s Cygnus XL spacecraft is scheduled to depart Friday, Oct. 9, as part of the company’s Commercial Resupply Services-24 mission, or Northrop Grumman CRS-24.

Watch NASA’s live coverage of undocking and departure beginning at 12:30 p.m. EDT through a variety of platforms. Learn where to watch online:

https://www.nasa.gov/live

The Super Intelligence Era Will Be Defined by Trust

Most arguments about AI still begin with autonomy: how much should these systems be allowed to do on their own? It is a fair question, but it is not the one that decides whether an institution can use them.

A very capable system can be governed if people can read its rules, cap its authority, rerun a decision and see what an update changed. A modest chatbot can be impossible to govern if its behavior is spread across model weights, training data, prompts, retrieval, filters, tools, vendor policy and decisions no one recorded.

The industry has become good at controlling what a system says. It has not found a way to control the moment its behavior changes. When a vendor ships an update, most deployed systems give the customer no single place to see what changed, which rule produced a given result, or whether the limit the customer signed off on still holds. That gap matters more each month, as these systems move from answering questions to acting inside companies, hospitals and infrastructure.

Hierarchical Prototyoe Emergence in Hopfield Networks

How can an AI memory learn the patterns behind the patterns?

A new study, Hierarchical Prototype Emergence in Modern Hopfield Models, explores how associative-memory models can learn not just individual examples, but the deeper structure connecting them—and potentially use that structure to create new, sensible examples. https://arxiv.org/abs/2609.

The researchers study a modern version of a Hopfield network, an AI system designed to store and retrieve patterns. Instead of treating every memory as unrelated, they organize the memories into a hierarchy, much like a family tree: individual images belong to groups, groups share broader characteristics, and those groups may themselves belong to larger categories.

The key question is whether the network can go beyond simply remembering the images it was given. The researchers find that, under certain conditions, the network develops prototypes—stable representations that capture common features shared by multiple memories. In other words, rather than memorizing a particular example, the system can discover something like the “essence” of a group and reconstruct a new example from it.

Interestingly, the amount of information needed to achieve this kind of generalization grows only quasi-polynomially with the complexity of the hierarchy, suggesting that learning these higher-level patterns may be considerably more efficient than simply memorizing every possible combination.

The researchers also find similar behavior when testing Fashion-MNIST data: the transition between memorization and prototype formation depends on factors such as how many memories the network stores and the sharpness of its activation function.

The broader significance is that this provides a mathematical “toy model” for a much bigger question in AI: how can a system move from memorizing examples to learning hierarchical structure and generating something genuinely new from that structure? Understanding this process in Hopfield models could offer clues about more sophisticated generative architectures, including diffusion models.

This is especially interesting because it frames generalization as the emergence of stable higher-level representations, rather than simply as better recall of training examples.

Continue reading “Hierarchical Prototyoe Emergence in Hopfield Networks” | >

Industrial Robotics Training Sims Just Hit A Ceiling Nobody Budgeted For

A new benchmark from Dalian University of Technology (VA-Bench) tested 12 multimodal AI models on robot-arm manipulation tasks.

The results are consistent and uncomfortable: • Object location accuracy: ~100% • Task understanding: ~99% • Whole-task success (best model): only 53.93%

The more important gap is between detecting an error (73.6%) and correcting it in real time (46.7%). Dual-arm tasks collapsed further — single-arm success around 65%, dual-arm only 11%.

Simulation is teaching models what to see and what needs to be done. It is still failing to teach them how to reliably complete the action when conditions change.

For buyers evaluating robotics vendors: treat simulation demo success rates as an upper bound, not a production prediction. Ask for dual-arm success rates on held-out tasks, error-correction rates, and performance when object geometry varies before approving any pilot.

Full analysis:

#Robotics #Simulation #IndustrialAI #Procurement

Continue reading “Industrial Robotics Training Sims Just Hit A Ceiling Nobody Budgeted For” | >
/* */