LK-99, Sclerotic establishments, Fitch downgrades US debt, Trump indicted... again

The episode opens with teasing aimed at David Sacks over his appearance and whether he used soap. The hosts also joke that he is running for office and ask whether he will come out of retirement as a lawyer to defend Trump.

Friedberg opens the science segment by describing the past ten days of global attempts to replicate LK-99 as a rare moment of shared optimism. Hundreds of labs followed the paper's recipe, and some live-streamed their attempts. He explains the prize physically: electrons in a normal wire collide with atoms, which creates resistance and heat, and the only way so far to cut that loss has been extreme cold.

He says roughly 70% of the energy produced is lost to heat and friction, adds the energy spent cooling data centers, and suggests available energy could double or triple. Another host adds a limit on timing:

assuming huge infrastructure investments, which would take probably decades to do

Near-term gains would come from components. Quantum computers already use small superconductors as qubits, and cooler AI chips could follow.

The strongest support is computational. Five independent labs modeled the described crystal and found electron pathways that could permit superconductivity in theory. One simulation suggests only a single lead atom must be replaced by copper, which would explain why baked samples give inconsistent results. Another team suggested gold might work better than copper, and some think conduction may occur only along a line of atoms, which would require many parallel lines to work at scale.

The experimental record is thinner. A Chinese lab reported zero resistance, but only near 170 Kelvin, about 100 degrees below freezing. Friedberg expects months to years before anyone knows whether the material can be industrialized.

Friedberg relays a story he says he cannot verify. Lee and Kim supposedly saw the material in 1999, could not reproduce it, and spent two decades in ordinary jobs before getting funding during COVID. A rumor says a cracked vacuum tube during baking changed the sample. Another figure, Kwan, was reportedly fired in March and is rumored to have taken the lab data and posted a paper on arXiv.

arXiv lets anyone publish outside peer review, so nobody knows whether a paper is real. One theory is that Kwan wanted his name attached because only three people can share a Nobel Prize. Lee, Kim and four collaborators then released a second paper within a day. Both are described as messy, with errors and redundancy, though one speaker argues the rush to claim credit and the funding 25 years later suggest something may be there.