A crystal was just discovered that acts like a metal and a glass at the same time.
Molybdenum oxychloride, announced June 1, 2026.
Metals conduct electricity freely. Electrons move through them like water through a pipe.
Glasses are insulators. Electrons are frozen in place.
These two states have been mutually exclusive for the entire history of materials science.
Until now.
MoOCl₂ has a split personality at the quantum level:
• Along one axis: metallic. Electrons flow freely.
• Along the perpendicular axis: insulating. Electrons don't move at all.
This is called extreme optical anisotropy.
Why it matters:
• Smart contact lenses that can display information without power-hungry screens
• Ultrathin AR glasses that process light differently in different directions
• Photonic circuits that route light the way copper routes electricity
We've had one fundamental state for each material for 200 years of engineering.
Now we have both in one crystal.
Materials science just got a new category.
Effective today, we are:
1) Doubling Claude Code’s 5-hour rate limits for Pro, Max, and Team plans;
2) Removing the peak hours limit reduction on Claude Code for Pro and Max plans; and
3) Substantially raising our API rate limits for Opus models.
Prime Minister of the Netherlands handed over the government and left on his bicycle, but every Nigerian man needs a car to boost his confidence😂😂
I can't promise all of you cars if I become President but you will have good roads and infrastructure!
I just published the most recent batch of optimizations we applied to Elixir's set theoretic types that is allowing us to push the envelope in Elixir v1.20: https://t.co/ce4XsiyZkT (in-depth technical article)
Mark Cuban just pronounced software dead, and the implications will destroy industries before most people understand what happened.
Cuban: “Software is dead because everything’s going to be customized to your unique utilization.”
Rigid SaaS dies. Businesses stop bending to static tools. AI molds around needs in real time.
Value chain shatters. Decade of value going to generalist builders ends. Transfers to customizers who translate capability into advantage.
Next trillion comes from customized intelligence sold to “33 million companies” too small for giants to service.
Software-as-a-Service over. Service-as-Software owns everything.
Cuban: “33 million companies aren’t going to have AI budgets, aren’t going to have AI experts.”
Problems exist. Data exists. The bridge doesn’t.
Cuban: “Learn all you can about AI but learn more on how to implement them in companies.”
Market bifurcates into builders and integrators. Integrators capture value builders can’t touch.
Cuban: “Every single job available for kids coming out of school because every single company needs that.”
Technical translator becomes the only role that matters. Contextualizing AI for specific use cases becomes more valuable than building AI itself.
Learn “the difference between Sora and Veo” and “how to customize a model” and you modernize any legacy business’s economics faster than they can process.
Alpha isn’t the model. Alpha is applying it to contexts nobody else understood well enough to solve. There are 33 million businesses waiting for that person who doesn’t exist yet.
To become an expert, immerse yourself in meaningful projects that challenge you, learning as you go. Share your knowledge in your own words to solidify your understanding. Remember, your only competition is who you were yesterday. Embrace growth, and keep pushing forward!
Oh sure, let’s just trust the "expert" SWEs who brought us the wonders of Microservices and Docker. Because nothing says "reliable technology" quite like a stack of JavaScript, MongoDB, and Kubernetes. Who needs skepticism when hype is just so comforting?
This summer I started working on Ruby's new JIT compiler—ZJIT, with zero JIT background
After two months of learning, I wrote about the questions that puzzled me most:
- Where does JIT code live?
- How does Ruby execute it?
- And more!
https://t.co/7AZ2oPZu9O