Lifeboat Foundation: Safeguarding Humanity
Toggle light / dark theme

Blog

Category: robotics/AI – Page 83

The Android Show: I/O Edition | Gemini Intelligence

Introducing Gemini Intelligence, an intelligence system that knows what matters to you, helps you stay a step ahead and works proactively to get things done throughout your day, bringing the best of Gemini to our most advanced devices.

Join Mindy Brooks (VP, PM and UX, Android Platform), Dieter Bohn (Director, Product Operations), and Ruchi Bezoles (Director, Android Marketing) to see how we’re making Gemini Intelligence handle the busywork so you can get back to what brings you joy.

Watch the full show now to check out all of the innovations and breakthroughs coming soon to Android! → https://www.youtube.com/live/dXCCleAddEA

Learn more about Gemini Intelligence → https://android.com/gemini-intelligence.

Catch up on all things Android → https://android.com/io-2026

#TheAndroidShow.

Continue reading “The Android Show: I/O Edition | Gemini Intelligence” | >

The Commoditization of Intelligence: Why AI Aggregators Will Beat Foundation Models

Everyone is currently watching the major tech giants throw billions of dollars at the AI arms race, cheering for whichever foundation model happens to top the leaderboards this week.

It is an incredible spectacle to watch unfold, but focusing too closely on the tech itself might mean we are missing the actual business revolution happening right under our noses.

We have seen this exact economic shift before. The biggest winners of the internet era weren’t the ones who built the physical infrastructure or supplied the goods; they were the platforms that organized the supply and owned the user relationship. The same economic laws are now coming for artificial intelligence, actively turning “intelligence” into a basic, interchangeable utility.

The real value moving forward is no longer in the models themselves, but in the seamless interfaces that aggregate them. If you want to protect your business from vendor lock-in and position your team for ultimate flexibility, it is time to rethink your approach.

Read my full blog post to dive into why the future of AI belongs to the aggregators, and how your business can strategically capitalize on this shift.

We spend an enormous amount of time obsessing over the titans of the AI arms race. Every single week seems to bring a breathless new headline about OpenAI, Google, Anthropic, or Meta releasing a foundation model that edges out the competition on some obscure benchmark test. We find ourselves endlessly arguing over parameter counts, context windows, and raw reasoning capabilities, captivated by a multi-billion-dollar war unfolding in real-time.

Continue reading “The Commoditization of Intelligence: Why AI Aggregators Will Beat Foundation Models” | >

Garment humanoid robots, Zhejiang Humanoid lands order

Zhejiang Humanoid Robotics Innovation Center said on May 12 that it has signed a strategic partnership with Jack Technology and an order for 2,000 garment humanoid robots customized for garment manufacturing. According to Gasgoo, the company described the deal as the first mass deployment of humanoid robots in the global apparel industry. The announcement matters because garment handling combines flexible materials, tight tolerances, and repetitive production steps that have been difficult to automate with general purpose humanoids.

Garment humanoid robots face a hard manufacturing test

The source article frames apparel production as a demanding proving ground for embodied AI systems. Fabrics vary in material and shape, and they can wrinkle, shift, and deform during handling. Zhejiang Humanoid said alignment deviations for cut pieces such as collars and pockets must be kept within plus or minus 2 mm, while cutting and sewing tasks require motion precision of 0.3 to 0.5 mm.

Physicists create hybrid light-matter particles that interact strongly enough to compute

Eighty years ago, Penn researchers J. Presper Eckert and John Mauchly launched the age of electronic computing by harnessing electrons to solve complex numerical problems with ENIAC, the world’s first general-purpose electronic computer. Today, that same architecture still underlies general computing, but electrons are beginning to show their limits. Because they carry a charge, they lose energy as heat, encounter resistance as they move through materials, and become harder to manage as chips incorporate more transistors and handle larger volumes of data.

With artificial intelligence pushing today’s hardware to process, move, and cool more, Penn physicists led by Bo Zhen in the School of Arts & Sciences are looking to the electron’s massless counterpart, the photon, to shoulder more of the load.

“Because they are charge-neutral and have zero rest mass, photons can carry information quickly over long distances with minimal loss, dominating communications technology,” explains Li He, co-first author of a paper published in Physical Review Letters and a former postdoctoral researcher in the Zhen Lab. “But that neutrality means they barely interact with their environment, making them bad at the sort of signal-switching logic that computers depend on.”

Microsoft Exchange, Windows 11 hacked on second day of Pwn2Own

During the second day of Pwn2Own Berlin 2026, competitors collected $385,750 in cash awards after exploiting 15 unique zero-day vulnerabilities in multiple products, including Windows 11, Microsoft Exchange, and Red Hat Enterprise Linux for Workstations.

The Pwn2Own Berlin 2026 hacking competition takes place at the OffensiveCon conference from May 14 to May 16 and focuses on enterprise technologies and artificial intelligence.

Security researchers can earn over $1,000,000 in cash and prizes by hacking fully patched products in the web browser, enterprise applications, cloud-native/container environments, virtualization, local privilege escalation, servers, local inference, and LLM categories.

Von Neumann probes: Where are they all?

In the 1960s the Hungarian-born American mathematician John von Neumann wrote about machines that could make exact copies of themselves. He envisaged a kind of robot equipped with a computer brain that could be programmed to reproduce itself from raw materials taken from its surroundings. It wasn’t long before some people suggested that von Neumann machines, in the form of robot spacecraft, would be a great way for us to explore the Galaxy.

My other YouTube channels:
The Science Fiction Rock Experience (the music show I produce):
/ @sciencefictionrockexperience.
Discover Maths (with Juan Medina):
/ @discovermaths.
Science World (with Emrah Polat):
/ @scienceworld1

My website:
https://www.daviddarling.info

Yann LeCun on What Comes After LLMs

Yann LeCun, Turing Award winner and former Chief AI Scientist at Meta, joins Jacob Effron. The conversation centers on Yann’s contrarian thesis that LLMs are a dead-end on the path to human-level intelligence, despite being useful products — because they can’t predict the consequences of their actions, can’t plan, and fundamentally can’t model the messy, high-dimensional real world. He unpacks his alternative architecture, JEPA (Joint Embedding Predictive Architecture), which learns abstract representations rather than generating pixel-level predictions, and explains why this approach is essential for robotics, industrial applications, and any system that needs to operate beyond the substrate of language. Yann also reveals the real story behind his departure from Meta (he had zero technical influence on Llama, contrary to public narrative), the genesis of his Tapestry project for sovereign open-source AI, why he believes LLMs are intrinsically unsafe, where he diverges from his fellow Turing laureates Hinton and Bengio, and why he predicts the industry will recognize the paradigm shift by early 2027. Throughout, he offers candid reflections on the tension between research and product at major labs, and why he intentionally headquartered AMI Labs in Paris with zero Silicon Valley VC money.

0:00 Intro.
01:45 Why LLMs Aren’t the Path to Intelligence.
07:51 AMI and World Models.
12:07 The JEPA Architecture Explained.
15:55 Problems with Robotics Models Today.
20:37 Silicon Valley Herd Behavior.
28:18 Tapestry: Sovereign AI for the Rest of the World.
35:49 OpenAI Is the Next Sun Microsystems.
40:51 Why Yann’s Views Diverged from Hinton & Bengio.
44:32 LLMs Are Intrinsically Unsafe.
58:00 Why Yann Left Meta.
1:00:26 Reflections on FAIR
1:12:11 Advice for PhD Students.

LeWorldModel Paper: https://arxiv.org/abs/2603.

With your host:
@jacobeffron.
Partner at Redpoint.

A potassium-gated ion channel identified!

Potassium ions (K⁺) are essential for all cells and living organisms. Scientists have long believed that K⁺ merely passes through ion channels and transporters, rather than acting as an extracellular ligand or molecular “switch.” Indeed, there had been no clear evidence that K⁺ functions as a ligand for membrane proteins in animals or plants—until now.

“Unexpectedly, we made this discovery serendipitously while testing the effect of aspartic acid, with K⁺ added as a counter cation, on Alka, an ion channel located in the brain of Drosophila melanogaster,” said the author. “The compound was effective. At first, we thought the effect was due to aspartic acid, but we ultimately realized that it was caused by K⁺, meaning that Alka functions as a membrane receptor that detects extracellular K⁺ as a ligand.”

Ion channel currents in Alka-expressing cells changed significantly in response to K⁺ levels. The researchers combined electrophysiological analysis with AlphaFold3, an AI-based protein structure prediction tool. This allowed them to identify the K⁺-binding site in Alka. This site creates a chemical environment favorable for K⁺, similar to that found in aqueous solution or in the well-known selectivity filter of K⁺ channels.

/* */