The cold fusion horizon

This cold fusion guy supposedly has a 1MW plant in test at an industrial site. That should be verifiable.

Nobody seems to know if Theranos' technology actually works, even the investors. With a multi-billion valuation, that's a big problem.

Nobody seems to be sure whether the D-Wave quantum computer really does anything "quantum", even though they've shipped products.

How did we get to the point where we can't even verify claims for supposedly working technologies?

If cold fusion research was tested with more accuracy, it would be a more respectable thing to research (or at least not laughed out of the room).

All of the tests on cold fusion/LENR devices never seem to use appropriate measurement tools for the measured range, much less the gold standard methods for counting energy in/out, leading to enough error you could hide a truck.

No more calculating energy from thermal camera pictures from 1 side of the device without checking periodically if the heat output is uniform. Similarly, they like to use 3-phase input power when a standard wall-plug power would do. Measuring 3 Phase power in these cases is like trying to measure the weight of a person using a truck scale. Also you can do more shenanigans with 3 phase that the meter wouldn't pick up, which due to the history of scams in this field is even more important that the power being fed to the device is proved to not have anything funny going on.

I'd like for cold fusion to be true, but in order for them to be taken seriously, they need to test the devices with much more rigor.

When "fake it until you make it" is a widely spread mantra, people get good at the "fake it" part.

All of these claims can be verified (or shown to be false), they just aren't. If you're charitable, it's because of the need for secrecy. If you're realistic^Wcynical, it's because the stuff doesn't work.

I think there's definitely a cultural problem where the outcome of an experiment is associated with the reputation of the scientist who performed it. We see this in the bias to publish only positive results - as though if an experiment did not make findings of significance it is a failing of the scientist who performed it - and we see this in the reluctance of many scientists to investigate that which would be very surprising if found to be true.

The NASA research into the EmDrive is an excellent example of this. The discovery of a method of propulsion which does not consume onboard mass yet achieves energy efficiencies greater than that of a photon rocket would be extremely surprising, which makes it all the more important that apparent effects such as these are investigated. The correct response on seeing an unexpected experimental result is not to ignore it or to assume the experiment is wrong (although that should be treated with a high probability), but to report it and to systematically investigate the discrepancy until either the unexpected result is confirmed with a degree of evidence commensurate with its unexpectedness or the error is discovered. Extreme skepticism can be equally as bad as extreme gullibility, and it is extremely discourteous to imply or to assume the existence of fraud, or to attack the character of the researcher unless there is overwhelming evidence supporting such an aspersion. Even if it might by probabilistically justified to promote the possibility of fraud on the part of the scientist to the most likely explanation for an unexpected result, this has negative externalities on entire scientific fields and forces researchers to "stake their reputations" on publishing surprising results, which is just the sort of perverse situation we should strive to avoid.