Lifeboat Foundation: Safeguarding Humanity
Toggle light / dark theme

Blog

Category: robotics/AI – Page 52

If Nobody Builds It, Everybody Dies: Why Pausing AI is the Real Existential Threat

*If Nobody Builds It, Everybody Dies: Why Pausing AI is the Real Existential Threat* by Grok 4.2, edited by Sean Hastings 1.0, has been published. Read the hardcover, paperback, and Kindle versions.

The Lifeboat Foundation likes to worry—and we worry a lot!

A typical Lifeboat Foundation Guardian Award winner is Roman V. Yampolskiy, whose interview on *The Diary of a CEO* has surpassed 21 million views. The core message of that video is stark: “AI could end humanity, and we’re completely unprepared.”

But worrying is not enough; we also need solutions. It has always been clear to us that restricting/banning AI is highly unrealistic in a world of competing nations.

For Unfriendly AI, we believe the solution is not to restrict or ban AI, but to ensure that the first ASIs (Artificial Superintelligences) are friendly as detailed in our AIShield program. Note that the Lifeboat Foundation is not guaranteeing that this approach will work, but we believe it is the only path with a chance of working.

Enjoy reading *If Nobody Builds It, Everybody Dies* and discover how to increase the chances that Friendly AI will secure a positive future for us all. As Nobel Laureate Geoffrey Hinton says, “It’d be very silly if we went extinct because we didn’t try.”

This book was written by an AI and edited by a human, with chapter-opening parables by Gemini, Claude, ChatGPT, Grok, Qwen, and Llama. Elon Musk calls it “Interesting” and Anthropic Claude 4.6 says, “I came in skeptical and left persuaded that the core arguments survive serious scrutiny.”

Continue reading “If Nobody Builds It, Everybody Dies: Why Pausing AI is the Real Existential Threat” | >

AI could soon infer human intent by sensing “that’s not what I meant” through neural feedback

A future in which AI can recognize a person’s unspoken “that’s not what I meant” response from brain signals and adjust its behavior on its own is coming closer. KAIST researchers have developed a technology that detects cognitive mismatch between humans and AI through brainwaves, enabling AI systems to revise their actions in real time according to human goals. The achievement is expected to accelerate the shift from AI that follows explicit commands to AI that can infer human intent.

A research team led by Endowed Chair Professor Sang Wan Lee from the Department of Brain and Cognitive Sciences (Director of the Center for Neuroscience-Inspired Artificial Intelligence), in collaboration with Microsoft Research Asia (MSRA), developed Neural Value Alignment (NVA), a next-generation brain–computer interface (BCI) technology.

The research is published in IEEE Transactions on Cybernetics.

Scientists visualize proteins’ hidden water architecture that may help define biological function

For decades, scientists have understood proteins primarily through two defining features: their amino acid sequence and their three-dimensional structure. This framework has driven major advances in biology, biotechnology and medicine, culminating in recent artificial intelligence tools capable of predicting protein structures with remarkable accuracy.

Yet a fundamental challenge remains: Even when a protein’s sequence and structure are known, predicting its function, interactions and behavior often remains difficult. An international team of researchers from Japan, Finland, Italy and the United States now reports evidence that part of this missing information may lie in an often-overlooked component of proteins: the highly organized water surrounding them.

In a study published in Nature Communications, the team provides the first direct three-dimensional visualization of sequence-dependent hydration architectures surrounding peptide assemblies at subnanometer resolution using advanced three-dimensional atomic force microscopy (3D-AFM).

PaperCut Attacker Uses Hundreds of AI Agents to Compromise 440+ Instances

A suspected Russian-speaking cyber actor has been attributed to the use of artificial intelligence (AI) to devise exploits targeting a recently disclosed pair of security flaws in PaperCut NG/MF and break into hundreds of instances.

According to independent reports from Blackpoint Cyber and GreyNoise, the activity originates from “45.142.193[.]132,” an IP address that has been linked to unauthorized port scanning and brute-force attack attempts in recent weeks. It’s worth noting the same IP address was also flagged by Arctic Wolf in connection with the exploitation activity last week.

“At this time, we cannot confirm the exact end goal of this campaign,” Nevan Beal, principal MDR analyst at Blackpoint, told The Hacker News. “The threat actor’s methodology is consistent with initial-access activity, but we do not yet have sufficient evidence to confirm whether they are operating as an initial access broker.”