I once met a scientist who spent a week traveling to where there was a powerful x-ray laser. He used it to blast a thin film of something or other that was floating on the surface of some water. He left with a flash drive full of data and some FORTRAN titled LSQREFL, which allegedly could decode the laser results. He then spent the next 6 months trying to make it actually do that. Turns out you had to have a folder with today's date on it on your desktop, otherwise the program would crash. This was documented nowhere, he just eventually puzzled it out from the code.
I offered to put it on github for him, so that at least he didn't have to be the sole caretaker for this endangered bit of software, but he was afraid of running afoul of the original author's rights, so endangered it will stay.
This was maybe an unlikely occurrence, falling neatly in the not part of your:
> More often than not it's not anything software related
But it makes me think that there is still some juice left to squeeze out there. I mean, I'm having a good time with my one-class-per-semester, I'd just prefer to not have to do it for another decade before I'm enough of a biologist to get my hands dirty.
Sounds like he was doing an xray diffraction experiment? The last time (in my opinion) XRay diffraction based structure results meaningfully changed scientific discourse that affects human life was probably in the 80s or 90s. While it's important work it's no more important for Healthcare than some physics guy doing things with a random metal alloy. The point is there are interesting things but one shouldn't delude that this is the thing that's keeping us from unleashing human health prosperity.
There's two kinds of ignorance which come into contact when people work across disciplines.
In my own work helping ecologists, I see plenty of CS/ML folks who think they'll change the world by throwing a transformer at the problem. (which problem? you think we haven't tried that?) It takes some time and exposure to figure out what kinds of problems you can meaningfully contribute on.
On the other hand, I've met lots (most?) of ecologists who underestimate the impact of looking at their work through CS/ML lenses. Effective automation can greatly improve iteration speed, which ultimately leads to better outcomes than a slow but 'perfect' process. (and, indeed, the 'slow-but-perfect' process may not be sufficiently benchmarked, and not be perfect at all...)
You can do a lot of good by working closely with a practitioner, and identifying the places where they are spending a lot of time doing 'boring' stuff, and finding ways to automate or approximate the outcomes of that boring work. As you work with more people, you'll be able to identify boring stuff that everyone in the field is stuck doing.
So, in short, an excellent goal is to find ways to save people time through bottleneck analysis. Improve iteration speed and you improve the speed at which we can accumulate knowledge / make discoveries / etc. When you're done, it's "just a tool", but beforehand it's a problem holding back discovery.
There were https://en.wikipedia.org/wiki/Galectin proteins embedded in a thin lipid layer on the surface of the water. The goal was to understand what conditions triggered various conformational changes. I'm under the impression that such details end up in databases and get selected for further inquiry by drug design processes, particularly those targeting autoimmune diseases. Or at least, that's what I got from his talk on it. I'm still working my way up to the biochemistry classes.
But yeah I get your point about avoiding that delusion. Honestly I'd be happy enough to be doing something that I suspect is not actively harmful (this should be easy but SaaS tends to converge on products that control their users and not the other way around). I don't need to be humanity's savior. More helpful than harmful will be enough.
I admire your stance but consider this hypothesis: nothing is better than nonsense. Most research leads to nothing. Knowledge is better than ignorance but research funding isn't unlimited. I estimated the average biology paper to cost several hundred thousand dollars if it didn't involve animals and a million if it did. Does the parameter you find or establish even justify such a cost? Most of my and my colleagues work really didn't change anything in my opinion. Even if some of us didn't want to be useless like this, we had no choice. The system just forces you to do these projects. The only way to beat this system is to get out of it. That's what I did. I hope to start a garage lab sometime and do whatever I want as you say. At least I won't be taking public funding and I hope I might find something better without those limits after all.
I offered to put it on github for him, so that at least he didn't have to be the sole caretaker for this endangered bit of software, but he was afraid of running afoul of the original author's rights, so endangered it will stay.
This was maybe an unlikely occurrence, falling neatly in the not part of your:
> More often than not it's not anything software related
But it makes me think that there is still some juice left to squeeze out there. I mean, I'm having a good time with my one-class-per-semester, I'd just prefer to not have to do it for another decade before I'm enough of a biologist to get my hands dirty.