I got things to share.
And share them I will!
But before that, just a note: I was invited to deliver another morning lecture on using AI for Science in the Oulu ICT Breakfast series, so I’ll be updating my thoughts on that too. Making it all available also through the PhD Power Trio (just in case you don’t have it yet).
One of the best things lately was attending HCOMP 2024 in Pittsburgh. And I mean that quite literally: attending was great, Pittsburgh was…well, let’s put it this way: that society is not well. There’s work to do.
But don’t we all! And fixing USA is not my job.
This is my job. I wrote a ton of notes during the conference, and as always even more than the research I’m interested in understanding the ways how people deal with academic life. The meta-academic-advice of sorts. And what I took away from that trip is too juicy not to share, so here are my top five takeaways I learnt from fellow people in the conference:
1: Don’t chase what everyone else is busy chasing
Ohhhh, ChatGPT API was such a game changer in my field. And now approximately 190% of all papers published are LLM-assisted “something”… competing for the attention of all other LLM-assisted somethings out there. Here’s blog post well worth reading on the topic.
But in a more general way, every now and then something hot appears, and everyone jumps on the bandwagon. Maybe, if you’re among the very first ones, it’s a good idea to jump in and grind hard to get your ideas out there. But when it’s 1-2 years later…and people still cannot think anything else than that one new shiny thing, yeah…just do something else.
Take a step back, and really think are you just intending to publish on the same thing out of FOMO or is there something you really can contribute there? When you work on the same thing as everyone else, you’ll:
- Have way more competition
- Are unlikely to stand out
- Have lower impact even if your work is good, as the noise is just too high
I thought this was a pretty good reminder. Then, second:
2: Publishing and investing are not that different
So, we all know it’s true: Publish and perish is as real as the Santa Claus (look it up, lives in Rovaniemi, Northern Finland). But it still makes sure to divide your attention a little bit.
So, just like in investing you’d want to have some safe investments along with a few more risky bets, in publishing you should all the time put out some “safe” work that most likely won’t make big waves but also work on risky things that have a high chance of failing. But when the risky stuff plays out…it pays out big time!
Of course, with things like funding you always have to take risks, but even with papers you should start thinking about the ideas you are almost too scared to pursue. And then devote some time for them anyway.
Go for it! Oh and speaking about publishing:
3. Literature reviews are a great way to put out your thoughts
I’ll admit. I used to not be a fan of literature reviews. A lot of work. And, I mean, you’re not really doing any work, you’re just talking about work done by others.
But little by little I’ve changed my mind, and indeed reviewing literature or even datasets can be a great way to make a statement. Or, to provide useful guides for the entire field on what to work on.
And of course, you can have your own say in all this. What do you think, based on the work of others, is going wrong in the field or what is going great.
Literature reviews offer a brilliant opportunity to voice out your bold thoughts and visions of what should the future be.
And this brings us to the most important one….
4. Just say it
As the very last takeaway… Matt talked about courage. Sharing an example from some of his early work, his biggest “regret” (not sure he really meant it…anyway, something in that direction) in academia was not saying what he really wanted in some papers or occasaions. And then, some of his early ideas were “capitalised” by other people later who simply stated those ideas more clearly.
Bummer.
I think in this time and age, this is a huge one. We are so encouraged to please everyone, whether it’s in online communications or to make sure the reviewers who hold the power of your paper getting accepted or killed are not offended.
And no, I’m not saying to deliberately look for drama.
But I am saying to think carefully what you want to communicate with your work and don’t water down the message just to please the readers.
Say what you mean.
Alright, hope to see you next time, and let’s figure out how to make this game better for everyone!