Yup.

So what's next? Well, this research dovetails beautifully with what we've been doing in DOE's ESRA (which funded some of us working on this) and sets up a whole series of questions we look forward to digging into. Stay tuned.

It's scary when bedrock is busted. Many nearest neighbors of the same charge in bulk and bulk-ish conditions call into question a fundamental tenet of my understanding. But rest assured that thermodynamics predicts these states; these are not guesses based on motivated reasoning.

The team brought receipts. Quantum chemistry indicates the possibility. MLIP shows it applies to a range of commercially relevant systems. And a smoking-gun NMR shows these clusters exist in your cell phone battery. And me, scratching my head and saying, 'whoa'.

When Rich suggested Li+ forms effective clusters of (e.g., 3-ish Li+ as nearest neighbors), my brain melted a bit. Electroneutrality is violated, and we're not even at a double layer? Reader, I was skeptical. Sure, quantum fluctuations are wild, but are they relevant?

I've been thinking about batteries for ~25 years, and I've idealized electrolyte behavior for too long.

End result.

Coding agent coordination of datasheets, circuit layouts, and OpenSCAD/whatever is fucking magic.

I now see claude speak everywhere. A running thread. Note that many of these pages are interesting and filled great ideas and cool results. But I wish it didn't sound like a ChAId.

Is this the watermarking?

I find that Claude’s writing style, in code at least, has gotten very clickbaity. Increasingly verbose and jargony. I/ we wrote skills and code to fix the problems, but I find the style annoying even in problem-solving mode. anyone else?