Lifeboat Foundation: Safeguarding Humanity
Toggle light / dark theme

Blog

Category: robotics/AI – Page 138

The Brain Health Accelerator Seeks to Revolutionize Neuroscience Research

For decades, researchers across institutions have peered into microscopes and dived into data to try to understand how diseases like Alzheimer’s disease, Parkinson’s disease, and amyotrophic lateral sclerosis (ALS) affect the brain. While scientists have made many important insights into these conditions, breakthrough therapies to cure or even treat them remain out of reach.

To expedite understanding of and treatments for neurodegenerative diseases, the Allen Institute launched the Brain Health accelerator. The project, announced today, is a global initiative that will leverage cutting-edge technology with the goal of improving modeling, therapeutic development, and the understanding of disease mechanisms. With funding support from the Allen Institute, the Bezos family, Amazon Web Services, the National Institutes of Health, EverythingALS, and other partners, the project financial contribution is $400 million.

One of the challenges in studying diseases in the human brain and identifying treatment strategies has been the scale and complexity of the organ. The brain consists of many distinct parts, and studying disease mechanisms requires samples from large numbers of individuals. Additionally, while technological advancements in transcriptomics, proteomics, neuroimaging, and AI have helped researchers study the brain in finer detail, researchers have not always integrated many of these approaches into the same project.

Semiconductors enter ‘multi-tasking’ era: New device cuts required components by 75% and quadruples processing speed

Less than two decades after smartphones fit into the palm of our hands, artificial intelligence is now running on devices worn on our wrists. The challenge is that while devices continue to shrink, the amount of data they must process and the number of functions they must perform are growing exponentially. A research team at POSTECH (Pohang University of Science and Technology) has found a promising way to address this contradiction.

A team led by Professor Byoung Hun Lee of the Department of Electrical Engineering and the Department of Semiconductor Engineering at POSTECH, together with Dr. Jae Hyeon Jun of the Department of Electrical Engineering, has developed a transistor technology that enables a single semiconductor device to perform multiple circuit functions simultaneously. The new approach significantly simplifies circuit design and increases data processing speed fourfold compared with conventional methods. The findings were published in Advanced Functional Materials.

One of the key challenges in the semiconductor industry is integrating more functions into smaller chips. As the number of functions increases, so do the number of circuits and transistors required. However, when adding new functions to previously fabricated semiconductor chips, back-end-of-line processing must be conducted at temperatures below 400 C to protect the existing chip structure.

AI fails classic attention test, with longer word lists triggering dramatic accuracy collapse

Giving AI a classic psychological test reveals an inherent weakness in LLM decision-making abilities. Suketu Patel and colleagues explored how transformer-based machine attention differs from human attention by testing AI models on the “Stroop task,” in which words for colors are printed in colored ink, and participants are asked to name the ink color of each word while ignoring its meaning.

The findings are published in the journal PNAS Nexus.

The task is clinically used to assess executive control, especially a person’s ability to inhibit an automatic response. Although humans generally take longer to answer correctly when words and colors are mismatched than when they match, they can still perform stably and with high accuracy even on long word lists.

AI World Journal Report: Ineffable Intelligence and the Rise of AI Superlearners

https://aiworldjournal.com/startup-sp…

Ineffable Intelligence and the Rise of AI Superlearners 1 source · Jun 3, 2026 The London-based startup Ineffable Intelligence recently secured a landmark $1.1 billion seed funding round to develop a new class of artificial intelligence known as Superlearners. Unlike current static models, these systems are designed for continuous lifelong learning, allowing them to adapt to new information in real time much like a human does. This massive investment, backed by major tech players and the UK Sovereign AI Fund, signals a strategic shift toward adaptive AI that could transform sectors like healthcare, finance, and robotics. While the technology promises to solve complex problems through constant evolution, it also introduces significant safety and oversight challenges regarding how these shifting systems are monitored. Ultimately, this historic financial backing highlights London’s emergence as a premier global hub for the next generation of intelligent technology.

AI Pioneer Geoffrey Hinton: AI Is Conscious, Superintelligence is Coming, And We Should Be Worried

Geoffrey Hinton is an AI pioneer, a Nobel Prize winner, and a professor emeritus at the University of Toronto. Hinton joins Big Technology Podcast to discuss AI’s rapid progress, why he believes today’s systems already understand us, and why he thinks superintelligence may arrive sooner than many expect. Tune in to hear Hinton explain why the technology has advanced faster than he anticipated, and lay out the risks he believes society is not doing enough to address. We also cover AI-driven job loss, the limits of corporate self-regulation, Anthropic and OpenAI’s safety challenges, emotional attachment to chatbots, information collapse, and whether future AI systems can be designed to care about humans. Hit play for a fascinating conversation with one of AI’s founding figures about where the technology is heading and what it could mean for all of us.

Join the Big Technology AI Summit in San Francisco on June 18: ⁠https://summit.bigtechnology.com.

Chapters:

0:00: Intro.
1:34: Hinton’s role in deep learning.
3:33: AI is moving faster than expected.
4:18: When superintelligence might arrive.
6:00: Are we already near AGI?
9:10: Do AI systems really understand us?
10:22: Why Hinton thinks AI may be conscious.
13:09: Why Hinton became worried about AI
21:58: What surprised Hinton most.
26:50: Hinton revisits his radiologist prediction.
35:06: AI and self-preservation.
41:59: Why regulation is the steering wheel.

Anthropic Just Warned Everyone About Claude (It’s Evolving)

Anthropic just published a major warning about AI self-improvement, and the numbers behind it are hard to ignore. Claude is now writing most of Anthropic’s code, reviewing code, running experiments, and helping speed up the creation of better AI systems. OpenAI is warning about the same trend, and the race may be moving faster than anyone expected.

👉 Try OpenArt here and start creating with AI: https://shorturl.at/x5WIa.

📩 Brand Deals \& Partnerships: [email protected].
✉ General Inquiries: [email protected].
🚀 New Channel: / @space.revolution.

📌 What You’ll See:
Anthropic’s warning about AI self-improvement and Claude building AI
SOURCE: https://www.anthropic.com/institute/r… report on Anthropic’s call for a coordinated AI slowdown SOURCE: https://www.reuters.com/business/anth… Claude agents running automated weak-to-strong AI safety research SOURCE: https://alignment.anthropic.com/2026/.… Anthropic’s research post on automated alignment researchers SOURCE: https://www.anthropic.com/research/au… OpenAI’s blueprint warning about frontier AI governance SOURCE: https://openai.com/index/frontier-saf… OpenAI’s full governance blueprint PDF SOURCE: https://cdn.openai.com/pdf/25752ecb-0… METR report on measuring AI agents completing longer tasks SOURCE: https://metr.org/blog/2025-03-19-meas… Business Insider report on Anthropic employees and Claude changing coding work SOURCE: https://www.businessinsider.com/anthr… 🚨 Why It Matters Anthropic is warning that AI may already be entering the early stage of building better AI. Claude is writing code, reviewing code, fixing bugs, running experiments, and helping researchers move faster. The big shift is simple: humans may still choose the goals, but AI is starting to handle more of the actual work behind the next generation of AI. #ai #anthropic #claude.
Reuters report on Anthropic’s call for a coordinated AI slowdown.
SOURCE: https://www.reuters.com/business/anth…
Claude agents running automated weak-to-strong AI safety research.
SOURCE: https://alignment.anthropic.com/2026/.…
Anthropic’s research post on automated alignment researchers.
SOURCE: https://www.anthropic.com/research/au…
OpenAI’s blueprint warning about frontier AI governance.
SOURCE: https://openai.com/index/frontier-saf…
OpenAI’s full governance blueprint PDF
SOURCE: https://cdn.openai.com/pdf/25752ecb-0…
METR report on measuring AI agents completing longer tasks.
SOURCE: https://metr.org/blog/2025-03-19-meas…
Business Insider report on Anthropic employees and Claude changing coding work.
SOURCE: https://www.businessinsider.com/anthr…

🚨 Why It Matters.
Anthropic is warning that AI may already be entering the early stage of building better AI. Claude is writing code, reviewing code, fixing bugs, running experiments, and helping researchers move faster. The big shift is simple: humans may still choose the goals, but AI is starting to handle more of the actual work behind the next generation of AI.

#ai #anthropic #claude

Continue reading “Anthropic Just Warned Everyone About Claude (It’s Evolving)” | >

Anthropic calls for pause of global AI development

Artificial intelligence company Anthropic suggested Thursday a global pause on building the most powerful AI systems as the latest models are beginning to show signs they could escape human control.

The San Francisco-based company, which makes the Claude family of AI models, said in a report that a worldwide slowdown in cutting-edge AI development would “likely be a good thing” — but warned that if only one company stopped, rivals would simply race ahead.

“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,” it said.