Comment by storywatch
5 hours ago
It's good for the world but terrible for the practitioners in academia. Imagine you are in your last year of your PhD and then something new comes out and completely invalidates/obsoletes your results. Cough natural language processing PhDs during COVID when GPT-3 came around..
Academic here. It's actually great for us. You're usually the first to see the implications for your niche area and have a chance to get a hot paper out rather than a moderately impactful one.
I think it depends on who has the hot paper and who has the outdated results. I was in academia for quite some time and when sometime cool comes out that renders other results less important, there are always winners and losers beccause the only thing that matters in academia is impact and the latest and greatest - not the blind alleys along the way.
There isn't really an "us" in academia like there is in a private company, who all have a common goal: in academia the goal is to get there before another does and be on top.
The perspective changes if you're in your last year of PhD as GP said. I'm sure the system is built in a way that you'll still graduate because your supervisor knows this is about learning how to conduct research and not about being right. But it's likely still soul crushing to come out with no published paper (unless you still publish knowing you're wrong, which in some cases might be valuable).
If your soul can bounce back, finding ways your previous research might help, or combine with the breakthrough (or not !) can help. I've used many previous formal/analytic SOTA algorithms and methods for dataset tagging automation, or as feedback for RL, or to speed-up post-training.
Three/four years of (honest) work is usually worth to publish. The result might not be high impact (and thus end up in a lower tier journal, sadly) but knowing what didn't work is also a valid output.
This happens in biology all the time.
PHDs will spend years mastering some intricate multi-day long procedure to do some thing and a new discovery will obviate the need for the procedure completely.
Preach. The way out is to find truly basic problems that you’re attacking at a fundamental level.
That rarely gets funded. And even more rarely handed off to grad students. In a sense the failure of a bunch of grad students is priced in to the way science is done.
Which makes it so perverse their lives and careers depend on success as newcomers.
Einstein wanted to measure how fast Earth moves through ether as his research thesis.
For those unaware - the most famous experiment to measure this(both high school and university physics classes have drilled this into my brain): https://en.wikipedia.org/wiki/Michelson%E2%80%93Morley_exper... a long history of experiments failing to find evidence of an ether https://en.wikipedia.org/wiki/Tests_of_special_relativity
It's different when you invalidate your own research.
That’s life. Sometimes you see it coming and ride the wave, sometimes you’re late to the party.