695
science hell
(fedia.io)
A place for majestic STEMLORD peacocking, as well as memes about the realities of working in a lab.
Rules
This is a science community. We use the Dawkins definition of meme.
Oh man. I seem to get this a lot more on lemmy than I did on reddit.
My theory is that coders & tech enthusiasts think they're smarter than everyone else, so the dunning Kruger effect is just run rampant on lemmy with an over representation of that demographic.
I do talk about a lot of shit that I really don't know much about and I like to engage in unsupported supposition, but I'm very quick to admit my limitations. Occasionally though, I do talk about fields related to my work and do know more than a lay person.
Lemmy comments do come off as incredibly pedantic sometimes, especially when you point out an observation about something and people feel the need to reply in paragraphs about the topic, as if you never actually understood it to begin with.
Um ackshually 🤓
The reason people on Lemmy do that is because everybody on Lemmy is just an idiot, except me of course - I am not an idiot. So that's why I reply with multiple paragraphs over minor things.
See, replying to people with long winded ranty, over pedantic messages is likely to make them realise they're actually an idiot and don't know what they're talking about. Regardless of whether all the posted is just a 1 sentence meme referencing the OP, and, say, the office, it's just not good enough. They need to learn to be smarter.
So in closing, the reason this always happens to you and all the other lemmings here is because I'm actually smarter than you lot are actually. If you just started prefacing your jokes with 47 paragraphs of explanatory text, clearly explaining that you actually aren't an actual idiot this wouldn't actually happen to you anymore.
Git gud
We don't know what we don't know. Maybe 5 minutes is all it takes to understand the essence of a problem. Maybe several lifetimes. There are examples of people who have studied something for a long time yet have come to more incorrect conclusions than someone who reads a single paper on the subject might. (There are physicists who believe consciousness is "real" but "unphysical", biologists who think life must has been created and nurtured by a god, and healthcare specialists who think vaccines are bad.)
That doesn't justify being arrogant and naive or dismissive of people more knowledgeable in a subject matter, but it enables someone to decide that a person they're arguing with is one such example because "the truth is bloody obvious".
It's painful to read people's takes on things you know something about. At the same time, most of us do the exact same thing whenever we share our take on something we don't know as much about because we think we don't need to.
Honestly I don't think it's that hard to approach topics with the assumption that there's a lot you don't know, and I like to think that's the way most people behave in most cases.
It's fun to make assertions about things we don't know much about, provided we acknowledge that we're probably wrong.
I think reddit and now lemmy kind of supported a "never concede that you're wrong" learned behaviour, both because they're anonymous and because there's no cost to being an idiot.
I find it hard. I really try to check my assumptions and state my reasoning where I think it's relevant but if I am to draw a conclusion about anything then I have to make a lot of necessary assumptions first. Some I am more confident about, others less so, but they are almost always given the same status in a statement in the interest of brevity.
I make an unthinkable number of implicit assumptions every time I communicate any information to anyone, and considering the problem of infinite regress I don't even know where or if ever the assumptions end.
Some people are better at not coming across as assertive or arrogant (my partner, for one), and I admire that. I'm more the kind to throw a statement out there after thinking about it for a while and error correct if my assumptions are being challenged. The downsides to that method are many; unintentionally spreading misinformation being a major one, but also that people are frustratingly bad at criticising premises and instead often attack the conclusion itself and assume ill intent, or at best just disagree without further explanation.
Oh and a big problem is also that, in the process of making weakly founded assumptions and working with those, the more you work with them, the more they blend in with all the other heuristics you've accumulated in your life until you no longer remember they're even there - much less questionable.
Lemmings also seem to love moving the goalposts and then arguing that whatever I said doesnt fit their goals. I posted about how we could possibly use machine learning to optimise traffic (adjusting speed limits and light timings) with the goal of getting everyone to their destinations just a little bit faster and reducing the time each individual car is running.
Some Lemming just hooked in that it wouldnt work because we need better public transport, more trains, cycleways, people need to drive less... blah blah fucking blah. Treating me like I'm a fucking idiot because it didnt solve a problem he decided it should.
You probably already know but this is text book straw man argument.
Search up logical fallacies. It's a bit of a rabbit warren.