The company cited in the article is [Bluehart](https://blueheartenergy.com/our-technology)

One disadvantage that Bluehart lists of classic heat pumps is the need for a refrigerant fluid. But Bluehar uses helium internally; I wonder how often the helium would need to be replaced.

Does the remaining 13% of the ferrofluid end up in the filtered water or in the discarded microplastics? Or a mix? Ferrofluid doesn't sound like something I want to drink

Many (most?) people who would phrase that question are themselves already among the rich. There's always someone richer to blame; and if you're the richest person in the world there's always the poor who aren't playing by the rules. You can always find an excuse to cheat.

The reason to play by the rules is because society and all individuals end up better if we do so. See the iterated prisoner's dilemma

I guess the decision to put "zero electronics" in the title was made by some headline editor who thought it made for more clicks.

I abandoned Matlab for Julia 12+ years ago and find it such a better tool. It provides the hackability of Matlab with the power of a more powerful language

On a more positive note, one random person in the area unexpectedly confirmed that they thought global warming was indeed real.

I think this is one argument for a more progressive income tax (and/or a wealth tax). Sure, a sharply progressive income tax incentivizes work by low income people, but it also reduces the wealth and income of high-income people, indirectly making them and everyone happier.

https://archive.ph/HfXMr

> Our implementation is available at: https://github.com/jumerckx/Metapractice/tree/taco.

Unfortunately that URL doesn't work for me; maybe it should point to https://github.com/jumerckx/Tamagoyaki

> ... We present a new approach for structuring the optimization phase of a compiler. In our approach, optimizations take the form of equality analyses that add equality information to a common intermediate representation. The optimizer works by repeatedly applying these analyses to infer equivalences between program fragments, thus saturating the intermediate representation with equalities. Once saturated, the intermediate representation encodes multiple optimized versions of the input program. At this point, a profitability heuristic picks the final optimized program from the various programs represented in the saturated representation. ...

Fig 18 shows up to a 4x gain, which seems pretty impressive