Sponsored by
Hello readers,
Welcome to the AI For All newsletter! Today, we’re talking about NASA’s AI project, how to master red-teaming for generative AI, and more!
AI in Action: Going to Space and Not Taking Orders
For decades, NASA has flown missions on chips designed in previous decades — reliable, radiation-hardened, but slow. Space-grade processors have to survive solar wind, cosmic rays, and temperature swings that would destroy commercial hardware in minutes. That durability came at a cost: the computing power of a smartphone from fifteen years ago, running missions that generate terabytes of scientific data. The agency has now partnered with Arizona-based Microchip Technology to build something fundamentally different.
The result is the High Performance Spaceflight Computing processor, or HPSC — a system-on-a-chip small enough to fit in the palm of a hand. Testing began in February at JPL, where engineers have been putting the chips through radiation exposure, thermal extremes, and shock tests while simultaneously running high-fidelity simulations drawn from real NASA landing scenarios. Early results are striking: the processor is performing at roughly 500 times the capability of the radiation-hardened chips currently flying on NASA missions. It also includes scalable vector computing for AI workloads and adaptive power modes that let functions dial down when not in use — a meaningful feature when electrical power is one of the most constrained resources in deep space.
The practical implications go beyond speed. Communication delays to the Moon run seconds; to Mars, they stretch to twenty minutes each way. A spacecraft that can only react after hearing back from Earth isn't really autonomous — it's slow. The HPSC is designed to close that gap, enabling onboard AI to respond to unexpected situations, compress and transmit large data volumes without waiting for ground commands, and eventually support life-critical systems on crewed lunar and Mars habitats. Once certified, NASA plans to incorporate it into future orbiters, rovers, and deep space missions. The agency's project manager at JPL put it plainly: this is “hardware that will enable NASA's next giant leaps.”
🔥 Rapid Fire
Commentary: The AI Industry Is Losing
OpenAI’s operating margin is negative 122% and ChatGPT growth has stalled
Uber COO says AI costs are getting “harder to justify”
Companies are questioning if soaring AI spending is delivering returns
Big Tech’s AI spending is depleting its free cash flow
NVIDIA is very close to hitting the circular financing brick wall
“NVIDIA's circular financing scheme had to gear up significantly over the past 12 months in order to keep the wheels spinning and avoid an abrupt halt that would have nuked the whole AI narrative in the blink of an eye”
Culper Research shorts NVIDIA over suspected illegal GPU diversion to China
“NVIDIA claims that its business in China went to zero after April 2025 U.S. trade restrictions. We believe that in reality, over 20% of NVIDIA's FY 2026 compute revenues remained driven by China — supported both by illegal GPU diversion and Southeast Asian intermediaries”
SoftBank insiders grow uneasy over OpenAI investment
The cost of using AI is more expensive than paying humans
Jensen Huang says AI excuse for layoffs is “lazy”
Altman, Amodei, and others walk back AI job loss claims
How AI startups use ARR to inflate revenue figures
AI is making scams hard to spot. Here’s how to protect yourself.
Why Tesla’s AI trainers don’t trust its self-driving tech or safety stats
Still setting up entities in every country you hire?
What’s changing in how companies expand globally?
Hiring internationally used to mean opening entities, navigating months of legal setup, and building local infrastructure before making a single hire.
That model is starting to shift.
More companies are using EOR not just as a temporary solution, but as a strategic way to access talent faster, test new markets with less risk, and scale globally without adding operational complexity too early.
But the biggest change may not be the hiring model itself. It’s how companies think about expansion.
Instead of building infrastructure first and hiring second, many teams are now hiring where the best talent already exists — and building strategy around that reality.
Oyster’s Strategic EOR Whitepaper explores how modern companies are using EOR to scale internationally, where the model works best, and why the global expansion playbook is evolving faster than most leaders realize.
📖 What We’re Reading
Red-teaming has become a key part of generative AI product development. It is the first step in identifying potential harms to measure, manage, and govern to mitigate AI risk. Commonly used in the IT industry, red teaming is now prominent for stress-testing generative AI and identifying a broad range of potential harms, including safety, security, and social bias.
Since AI models are deployed worldwide, it is crucial to design red-teaming solutions that not only account for linguistic issues but also address threats arising from political and cultural contexts. It is vital to test generative AI systems as they are being rapidly integrated into enterprise applications, as they might introduce new security challenges, ranging from prompt injection to hallucinated instructions and training data leakage.
