Hacker News new | ask | show | jobs
by carabiner 1050 days ago
No I mean why is this "huge" when each time that's happened it's been deflated within 12 hours:

- initial LK-99 paper upload on Arxiv: HUGE, then it's probably nothing until replication (waiting)

- DFT release: HUGE, then probably nothing (DFT has poor predictive power)

- 110K SC: HUGE, then ... ?

In every case it's been the laymen saying HUGE, then the experts saying it's probably insignificant. Then the laymen settle on what the experts said.

3 comments

Derek Lowe is a layman? I grant his specialism doesn't precisely match the field, but it's easily close enough I'd expect him to be able to smell bullshit on this if there was any, and his latest "In the Pipeline" suggests much more the scent of roses.
Derek Lowe is saying 'this is maybe possible', commentators on Hacker News are saying 'this is practically a done deal, and here are the implications'
Just temper your expectations a bit and don't get hyped up by people on the internet. All of these are exciting developments, but no one is going to know the truth of it until you have multiple independent confirmations in one direction or another. No single piece of evidence is going to be convincing unless it comes from one of the big labs, and they're unlikely to publish quickly because they're going to wait until they have to k solid proof.

As someone with experience working on superconductors, the DFT results and this paper are exciting because they show that at the very least this is likely a new class of superconducting materials at least as good as the current industrial ones. Knowing that the authors are on to something and that the initial claims aren't totally nuts is exciting and fun to post about, but it'll take time to be sure about any of this.

As someone who has worked on superconductivity I'd say that all of these are all potentially HUGE, but meaningless individually because they require experimental replication. They point to at minimum a new class of high temperature superconductors at least as good as current industrial ones. To know if it's truly transformative or not though we'll need multiple confirmations from big name labs. That's going to take time, so this trickle is exciting but won't mean much until the dust settles.