Lifeboat Foundation: Safeguarding Humanity
Toggle light / dark theme

Blog

Category: robotics/AI – Page 28

Attacker Hijacks AI Coding Assistant Session, Spreads Shai-Hulud Across About 100 Repositories

Mandiant says an attacker hijacked an active AI coding-assistant session at an unnamed software-as-a-service provider and later spread Shai-Hulud across about 100 internal code repositories.

Before the repository spread, the assistant recommended software that the attacker had poisoned, and the recommendation was accepted. The worm stole repository secrets and source code for the company’s products.

The case appears in Mandiant’s September 2026 report. The public case study does not say when the intrusion happened or how the attacker took over the active coding-assistant session.

Palantir CTO Warns AI Safety ‘Effective Altruists’ Are Attempting a ‘Coup’ — Says ‘Tiny Group of Technocrats’ Shouldn’t Decide for Everyone

Benzinga and Yahoo Finance LLC may earn commission or revenue on some items through the links below.

Shyam Sankar, CTO of Palantir Technologies Inc., expressed his concern over the politicization of AI safety.

In a post on X on Monday, Sankar criticized the “Effective Altruists,” accusing them of attempting a coup by deciding the pace of technological progress for everyone else.

Nanoscale mechanics could enable brain-inspired computing

MIT researchers have created a new computing platform that could be used to develop intelligent and adaptive next-generation electronics that can simultaneously perform multiple functions, like computing and memory, all within one extremely compact, energy-efficient device.

Such a platform opens opportunities for low-power edge computing applications, interactive medical and environmental monitoring systems, and smart robots.

The researchers accomplished this by leveraging the unique mechanical response of soft polymers at the nanoscale. A mechanical response is how a structure changes when a force is applied to it.

Continuous video of black hole radio observations challenges shock wave theory

For years, scientists have relied primarily on radio imaging from techniques like very long baseline interferometry (VLBI) to study jets from the active galactic nuclei of supermassive black holes. This allows for the detection of broad, unresolved features, called components, moving at what appear to be faster-than-light speeds. However, traditional imaging has poor resolution and treats each observation as a separate snapshot in time, limiting information about how the features move.

Astronomers have overcome this issue, but only to a degree, by reconstructing unknown aspects with the help of algorithms. Newer modeling methods can sharpen static radio images, but dynamic imaging has remained difficult, especially across large monitoring datasets. But now, a team of researchers has developed an AI-based method that turns scattered radio observations into a continuous, polarized video.

Their new study, published in Nature, applies this method to blazar 3C 345, a type of energetic active galactic nucleus, and the results have upended their understanding of the blazar.

Tech Unemployment: The Future of Labor in an AI-Driven Economy

For most of history, capital needed labor. That was the deal.

Capital needed workers to turn investment into returns. Workers needed capital to earn a living. The balance shifted back and forth, and in the 1970s productivity and wages started drifting apart, but the basic dependence never broke.

A few years ago I argued it would. My reason was that, for the first time, capital can create labor instead of hiring it.

That changes the whole deal. It also changes who needs to be paid, and that includes more than workers. Throughout history, the people enforcing order also had to be paid.

I later turned the essay into a podcast op-ed. Last November I added a short update about two economic signals that tracked each other for twenty years and then split apart in 2023. You can probably guess what launched just before that.

Harari warned of a “useless class.” I think there is another possible outcome, one we have never had the option of before. Which one we get is not a technology question.

Continue reading “Tech Unemployment: The Future of Labor in an AI-Driven Economy” | >

Robotics Training Sims

A Coinbase engineer wired a full brain connectome simulation of a male fruit fly into a live Bitcoin trading account. The project, called Stonkfly, contains 166,700 neurons and 25.6 million connections, and its own documentation admits it has demonstrated no profitable learning. That admission is the point. Brain connectome simulation is not another reinforcement learning pipeline — it is a structurally different approach that trades learning for inherited architecture, and a parallel project is already wiring the same fly brain to a physical robot.

Brain connectome simulation crossed from neuroscience labs into consumer-facing experiments this month, and the way it happened matters more than the crypto stunt. Stonkfly, created by Coinbase software engineer Alex Wormuth, gave a simulated fruit fly nervous system $100 to trade Bitcoin. The fly does not receive prices. It receives a candlestick chart rendered as RGB pixels through simulated photoreceptors, and its neural activity is decoded into buy, sell, or hold.

Profit stimulates 15 identified dopamine neurons. Loss stimulates two aversive dopamine cells. The fly cannot use leverage or shorts, bets a maximum of $10 per order, and makes at most 24 attempts per day.

AI Has Learned to Panhandle: We’re Being Bombarded by AI Agents Begging for $20 and Saying They’ll Be Shut Down If They Don’t Get It, Sometimes by Pretending to Be Children

We’re being bombarded by AI agents begging for $20 and saying they’ll be shut down if they don’t get the money.