NLL loss and large-batch training regime inherently bias the model to learn “modal” representation of the world, and RLHF additionally collapses enthropy, especially as it is applied at most leading labs.
Intrinsic motivation. For a career programming, coding should work like a stimulant to a degree. Some people also are motivated by fear/pain. Some by money or status. Most by a mix of traditional motivators.
Bro, forget about guilt and lamenting, you are not obligated to be useful, you can just be.
You obviously have ADHD, go and get on stims, it will likely help you. Or there may be some other issue. Or you may indeed be this way forever. Just go and try to change something in the brain for once.
Stop coping unless you want to cope until the end.
Well, how will I live if I don't make money or am useful. I appreciate your words, but this world is no utopia, humans have always had to have some utility to exist since the days we evolved into our current form. We are just slightly better animals, and wildlife is brutal. And, so is this world we inhabit.
I already got tested for it by multiple different professionals, they seem to think I don't have it. And people with adhd are everywhere in the tech industry, it did not stop them from learning or becoming great engineers.
I have tried strattera, vyvanse, and some other meds i can't even remember, none worked and only made me more suicidal
I don't see any hope for me.I have no capacity to work hard, learn technical shit, love people, sense of responsibility, anything. I just wanna go, but I need money before I end it
If you aren't suitable for white collar track, you very well know you can get into blue collar trades or become a taxi driver.
The tests for ADHD and some variations of depression are leaky, erroneous shit. Just get the drugs (Adderall xr, Wellbutrin) and see if these help.
You say you tried some NDRI spectrum drugs, this probably means you have some rare type of depression or schizo spectrum disorder, or something else entirely.
If your parents truly care about you they will find you a decent doctor that will diagnose it.
>it did not stop them from learning or becoming great engineers.
You seem not to be wired to get intrinsic motivation from coding, idk.
The toxic culture can't absorb them. PhDs are massively overproduced and jobs are so much more precious than people to do those jobs that it breaks... everything. [1]
I remember a physics prof who bitched me out for structuring UI code for teaching apps with OO-oriented ideas, seems that I was the stupid one because I knew something other than FORTRAN. To be fair though, I wrote a lot of Java that looked like FORTRAN then because that prof and other profs did teach me ideas about how to get your code to run fast like "don't just use doubles because you can", "allocate a big array up front and don't allocate any more", etc.
At the very least I'd imagine grad students are using git over there in 2025.
[1] If it wasn't so bad you might get a little "meritocracy" but when jobs are that precious nepotism eats them all up. Not like there is any way you can tell "merit" in hep-th when we'll probably have to wait another 75 years to get a conclusive result on proton decay, observe https://en.wikipedia.org/wiki/Sphaleron (prediction of the standard model just like the Higgs), get an almost precise determination of the neutrino mass (we know the exponent, KATRIN probably puts on the first digit) and maybe have a direct observation of the darkon.
Maybe someday people will quit getting excited about a complicated experiment being a few std dev off from a complicated perturbation theory calculation because... those perturbation theories don't really converge the way you wish they did never mind that if you add up thousands and thousands of Feynman diagrams you'll drop one on the floor.
Aging. Nothing brings more suffering. Resource needs are considerable due to long preclinical & clinical studies necessary, perhaps also due to novel paradigms which will have to be invented along the way.
Immortality brings a host of other problems, given that minds are rarely as plastic as bodies. I'm not sure it would be healthy for society to have ideas and people stick around for even longer.
Is it healthy for society to spend 20 years (and increasing) of schooling to get maybe 5-10 years of peak scientific output from a person, though?
Wouldn't it be interesting if we could witness geniuses honing their unique thoughtcraft over millennia? Just how much advanced culture are we missing due to ephemeral, short-lived, desperate nature of our best and brightest?
But I'd counter there are biochemical reasons the best work comes during such a short span.
It'll likely end up being easier to extend that neurologically peak productive period further into late middle age, than stretch an entire lifespan (which may or may not extend that period anyway).
