bob the mongoose
37 posts
i would give all my worldly passions for zverev to make it to the final, be up two sets and a break, get a few match points, then get brutally reverse swept
English
i feel that this sort of reasoning is a bit lazy? if two generative processes produce indistinguishable observations then why do i care about differences in the latent mechanism. and as a professional chud i would argue that most of the time i can only approximate intelligence :/
Valerio Capraro@ValerioCapraro
The Pope is making exactly our point. LLMs “may imitate or even simulate, but they do not understand.” This is the core epistemic fault line. Most AI evaluation is still based on one assumption: if a system statistically approximates human behaviour, then it is close to human intelligence. But approximation is not intelligence. Simulation is not understanding. LLMs can produce the right answer without knowing why it is right. They can simulate empathy without feeling. They can imitate judgment without responsibility. They can generate coherent explanations without having a world to which those explanations are accountable. Stop confusing behavioural similarity with cognitive equivalence. Human understanding is embodied, affective, relational, motivational, and normative. It is not just the production of plausible text. * Full paper in the first reply
English
@joannejang idk there’s definitely some work that has to go into constructing the optimal portfolio to construct your synthetic short, not trivial at all
English
@joannejang since most of them are private it’s hard, maybe the most obvious way to me is to look at which publicly traded firms have large holdings in them and try to synthetically short the startup by shorting these firms and then being long an index to hopefully hedge out other factors
English
hypothetically
if you're willing to bet literally 90% of your net worth that an ai startup valued $30B+ will fail, where / how would you bet?
English
@Etoile_up ok the part where he pretends to clear angles he knows are empty got me lmao
English
@finn_hulse idk man i don’t recall sig mock being that deep, no idea what u r talking about rn lol
English
i remember baffling the instructors regularly with my "antics"
during the group discussions at the beginning of mock trading sessions, i'd randomly share tricks (like some way of using the software or a slick math thing) that i would've been much "better off" keeping to myself
the thing is, i never ever want to win just because of some arbitrary advantage. i want everyone to have all of the ingredients to success and i want to crush them nonetheless
the worst, most feeble startups in the world (well, really just the middle of the curve) act like those quant firms: trying to hoard and nurse unique little edges that others would be capable of out-executing them on if shared
the strongest have no such need. they out-execute and monopolize regardless. it's a big reason i really don't mind being extraordinarily open on twitter, even with stuff that might appear like it could be used "against me." it does not matter to me
Finn Hulse@finn_hulse
quant firms would almost never hire someone who'd drop out of college to join a startup especially for trading roles quant attrition has a uniquely high impact on their PnL since your value compounds forever and is initially quite low. if you leave even after 2 years, the firm is likely at a loss (this is the main reason the comp scales upwards forever) so kids who could be convinced to drop out of school to join startups would be huge liabilities/flight risks. those firms want teachable people who will stay for many years—that's their success function. you can have a high risk tolerance when trading THEIR money, but your personal risk tolerance must be negligible i had to learn this the brutally hard way when i was cut from SIG in their final round of big cuts. i was so so pissed because i had literally carved myself into a new person to fit their mold of what made a technically strong trader. but i couldn't change certain aspects of myself and it is a good thing i didn't, because those are the same traits that will let me cripple quant/casino revenues by changing the outside world. if you have any such desires, you're MUCH better off in startup land (they'll sniff you out very quickly anyway)
English
@rucibear not asking for us to build an actual starship either but a realistic proposal would be nice :/
English
@rucibear yeah i mean tbf im kinda an optimist at heart and that part of me wants to think that maybe much of the relevant technologies also have adjacent use cases like strong thin lightweight materials needed for a laser sail for instance or fusion (lmao)
English
this is probably one of the least plausible explanations for the fermi paradox imo
Derek Pederson 🇺🇸🇺🇦🇻🇪@DerekPederson3
Science fiction nerds are gonna hate me for saying this but the Great Filter is that interstellar travel is probably more or less impossible and there is no reason to come up with any other explanation for that.
English
i feel bad for people who derive their self worth or identity from things that they are good at on a relative basis. i can see that it’s natural and admittedly it happens to me too but there’s obviously a problem with needing to be needed
Tomás Bjartur@BjarturTomas
Mathematician reacts to OpenAI's recent proof:
English
@nebulo_s8 that wouldn’t be so bad but i don’t think i can part ways with my hair
English
it's hella ungrateful of me to be so disappointed with where i am in life all things considered and yet at the same time i feel that i'm capable of so much more than i have accomplished
English
in a sense, i wonder if i have gotten too comfortable with not failing that i've become scared of pursuing things that are genuinely difficult and carry real risk to them
English
and it's difficult to tell whether i ought to resolve this by learning to be happier with where i am and what i've achieved or whether i should really lock in and see what i can achieve if i put my mind to it
English
@RadishHarmers meh yeah i mean in reality what is really being said is that the relative volume of object that is sphere like as n grows large goes to 0. it’s only really formal in an asymptotic sense
English
You can see this in a completely elementary way, one which makes it retrospectively obvious:
For a point in the cube to be in the inscribed unit sphere, its first two coordinates must lie in a unit circle, its next two coordinates must lie in a unit circle, and so on.
Prof. Brian Keating@Briankeating
My intuition about geometry was catastrophically wrong. I never knew that, in high dimensions, spheres effectively disappear. By 100 dimensions, an inscribed sphere occupies a smaller fraction of its cube than a proton occupies of the observable universe. By 500 dimensions, the volume is smaller than what standard floating-point arithmetic can even represent. The equations are real. Here’s the wild horror hidden in higher-dimensional geometry:
English
@RadishHarmers maybe a bit more heuristic but u can also think about the fact that the vertices of the cube r moving farther away from the origin as n increases while the points on the sphere are not and there r also more and more vertices and the volume difference is coming from these corners
English
These separate coordinates are independent, so overall, the probability of a point in the cube being in the sphere is ≤ k^(floor(n/2)), where n is the number of dimensions and k is the probability for two dimensions.
English
@RocketBullets sometimes bad people are good at playing video games! being good at video games does not make you a better person!
English
in his debut, haeyoday dropped 43
do you still not like the signing?
ROCKET@RocketBullets
why are people clowning this signing
English
@rucibear part of the reason i’m optimistic is that i also expect automated research to be a thing somewhat soon but tbf it’s not so clear how meaningful this will be towards progress for fields that have very real material constraints rocket science being one of them
English
@rucibear im envisioning something like >0.1ish c speeds.i may very easily be too optimistic but i expect the rate of scientific progress to increase substantially and in a project hail mary like scenario where we had to get our shit together or go extinct i feel 100 years is feasible
English