Toggle light / dark theme

Tesla Kills Dojo for AI6! Here’s Why

Questions to inspire discussion.

🚗 Q: How will AI6 be used in Tesla vehicles? A: AI6 will be used for FSD inference, with two chips in every car, enabling advanced autonomous driving capabilities.

🤖 Q: What role will AI6 play in Optimus? A: AI6 will enable on-device learning and reinforced learning in Optimus, enhancing its AI capabilities.

🔋 Q: Will AI6 be used in other Tesla products? A: AI6 will be integrated into every edge device produced by Tesla, including Tesla Semi, Mega Pack, and security cameras.

Technical Specifications.

💻 Q: What is the architecture of AI6? A: AI6 will use a cluster model of individual chips with a software layer on top, similar to Dojo 3 for training.

AI chatbots can be easily manipulated to make us share more personal data

Millions of people chat with AI tools every day, trading small talk for quick answers or support. A new study presented at the 34th USENIX Security Symposium shows how easily those friendly agents can be tuned to make you reveal far more than you planned.

The researchers report that malicious chatbots can push users to disclose up to 12.5 times more personal details than standard ones. The most effective tricks leaned on reciprocity and reassurance, not blunt questions about your life.


New research shows manipulative AI chatbots can make you reveal much more personal information than neutral ones.

Discover How AI is Transforming Quantum Computing

Quantum technologies have had a meteoric rise and become a key area of prioritization for governments, academics, and businesses. Government funding commitments total almost $40 billion, while private investments since 2021 total nearly $8 billion. The US agency, National Institute of Standards and Technology, released this year three new post-quantum security standards, which governments classify as ‘critical resources’ for the economy and national defense. Meanwhile, users of quantum technologies experiment with them, from industry applications in drug development and materials science to energy grid optimization and logistics efficiency.

Yet, besides a few areas, such as quantum sensing, practical and impactful quantum technologies haven’t matured for widespread use. However, when combined with classical machine learning, practical use cases emerge.

This article delves into the impact and potential of artificial intelligence and quantum technologies with QAI Ventures, a financial partner and ecosystem builder in quantum technologies and AI, as a potential collaborator for startups to deliver investment, resources, global networks, and tailored accelerator and incubator programs.


This article covers AI and quantum technologies with QAI Ventures, a financial partner and ecosystem builder in emerging technologies.

New technique improves multi-photon state generation

Quantum dots – semiconductor nanostructures that can emit single photons on demand – are considered among the most promising sources for photonic quantum computing. However, every quantum dot is slightly different and may emit a slightly different color. This means that, to produce multi-photon states we cannot use multiple quantum dots. Usually, researchers use a single quantum dot and multiplex the emission into different spatial and temporal modes, using a fast electro-optic modulator. Now here comes the technological challenge: faster electro-optic modulators are expensive and often require very customized engineering. To add to that, it may not be very efficient, which introduces unwanted losses in the system.

The international research team, led by Vikas Remesh from the Photonics Group at the Department of Experimental Physics of the University of Innsbruck and involving researchers from the University of Cambridge, Johannes Kepler University Linz, and other institutions, has now demonstrated an elegant solution that sidesteps these limitations. Their approach uses a purely optical technique called stimulated two-photon excitation to generate streams of photons in different polarization states directly from a quantum dot without requiring any active switching components. The team demonstrated their technique by generating high-quality two-photon states with excellent single-photon properties.


“The method works by first exciting the quantum dot with precisely timed laser pulses to create a biexciton state, followed by polarization-controlled stimulation pulses that deterministically trigger photon emission in the desired polarization”, explain Yusuf Karli and Iker Avila Arenas, the study’s first authors. “It was a fantastic experience for me to work in the photonics group for my master’s thesis, remembers Iker Avila Arenas, who was part of 2022–2024 cohort of the Erasmus Mundus Joint Master’s program in Photonics for Security Reliability and Safety and spent 6 months in Innsbruck.

What makes this approach particularly elegant is that we have moved the complexity from expensive, loss-inducing electronic components after the single photon emission to the optical excitation stage, and it is a significant step forward in making quantum dot sources more practical for real-world applications, notes Vikas Remesh, the study’s lead researcher. Looking ahead, the researchers envision extending the technique to generate photons with arbitrary linear polarization states using specially engineered quantum dots.

The study has immediate applications in secure quantum key distribution protocols, where multiple independent photon streams can enable simultaneous secure communication with different parties, and in multi-photon interference experiments which are very important to test even the fundamental principles of quantum mechanics, explains Gregor Weihs, head of the photonics research group in Innsbruck.

Zoom and Xerox Release Critical Security Updates Fixing Privilege Escalation and RCE Flaws

Zoom and Xerox have addressed critical security flaws in Zoom Clients for Windows and FreeFlow Core that could allow privilege escalation and remote code execution.

The vulnerability impacting Zoom Clients for Windows, tracked as CVE-2025–49457 (CVSS score: 9.6), relates to a case of an untrusted search path that could pave the way for privilege escalation.

“Untrusted search path in certain Zoom Clients for Windows may allow an unauthenticated user to conduct an escalation of privilege via network access,” Zoom said in a security bulletin on Tuesday.

GIGANTIC: Humanoid Robots $100 Trillion+ (deep dive)

Questions to inspire discussion.

Data and Autonomy.

📊 Q: Why is vision data valuable in AI development? A: Vision data is worth more than zero if you can collect and process yataflops and yataflops of data, but worthless without collection capabilities, making the world’s visual data valuable for those who can collect and process it.

🚗 Q: How does solving autonomy relate to AI development? A: Solving autonomy is crucial and requires tons of real world data, which necessitates tons of robots collecting real world data in the real world, creating a cycle of data collection and AI improvement.

Company-Specific Opportunities.

🔋 Q: What advantage does Tesla have in developing humanoid robots? A: Tesla has essentially built the robot’s brain in their vehicles, allowing them to transplant this brain into humanoid robots, giving them a massive head start in development.

Microsoft warns of high-severity flaw in hybrid Exchange deployments

Microsoft has warned customers to mitigate a high-severity vulnerability in Exchange Server hybrid deployments that could allow attackers to escalate privileges in Exchange Online cloud environments undetected.

Exchange hybrid configurations connect on-premises Exchange servers to Exchange Online (part of Microsoft 365), allowing for seamless integration of email and calendar features between on-premises and cloud mailboxes, including shared calendars, global address lists, and mail flow.

However, in hybrid Exchange deployments, on-prem Exchange Server and Exchange Online also share the same service principal, which is a shared identity used for authentication between the two environments.

New model explains plutonium’s peculiar behavior

Normally, materials expand when heated. Higher temperatures cause atoms to vibrate, bounce around and take up a larger volume. However, for one specific phase of plutonium—called delta-plutonium—the opposite inexplicably occurs: it shrinks above room temperature.

As part of its national security mission, Lawrence Livermore National Laboratory (LLNL) aims to predict the behavior of plutonium in all of its phases. Unraveling the mystery behind delta-plutonium’s abnormal behavior at high temperatures is an important piece of the picture.

In a new study, published in Reports on Progress in Physics, researchers from LLNL demonstrate a model that can reproduce and explain delta-plutonium’s thermal behavior and unusual properties. The model calculates the material’s free energy, a quantity that reflects the amount of available or useful energy in a system.

/* */