Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
ausbah
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
45 ms
·
541.
▲
by
ausbah
5y ago
are commissions, paywalls, and just general donations avenues to supporting artists? is the "ownership" of NFT something people really care about? I know I don't
542.
▲
by
ausbah
5y ago
even worse, a field will take term from another field and misuse it enough such that "wrong" definition" becomes the default definition
543.
▲
by
ausbah
5y ago
is there? are the rates of queer people reddit in tech that much higher than other fields?
544.
▲
by
ausbah
5y ago
I've heard this a lot on HN, any data to back it up?
545.
▲
Georgia’s University System Takes On Tenure
(nytimes.com)
45 points
by
ausbah
5y ago
|
69 comments
546.
▲
by
ausbah
5y ago
money is fake but "printing all that money" cushioned many of the economic side effects of the pandemic
547.
▲
by
ausbah
5y ago
why would you reveal yourself if you don't have guanareteed protection?
548.
▲
by
ausbah
5y ago
I think your point on scarcity and attention are well put, but can't apply to NFTs in the way as physical commodities because I think there has to be some notion of originality intertwined. Like for art pieces, sure people can make cop
549.
▲
by
ausbah
5y ago
that's the whole point! if randomness can give returns higher than most "financial gurus", what do they really know about the market?
550.
▲
by
ausbah
5y ago
stuff like the lightning network has similar fees right? I can see the benefit for the smaller scale of tips though
551.
▲
by
ausbah
5y ago
what does bitcoin really enable here that normal payment systems don't
552.
▲
by
ausbah
5y ago
sounds like SOTA ml research
553.
▲
by
ausbah
5y ago
people need to trust science because the average person doesn't have the scientific background necessary, much less the time, to evaluate the accuracy of everything scientific they deal with - especially in the realm of medicine like i
554.
▲
by
ausbah
5y ago
deep learning is mostly a dead end rn for any sort of "general reasoning", like great image labeling is much better but beyond that who knows
555.
▲
by
ausbah
5y ago
jesus the lack of enforcement on stuff like this is outrageous. at the very least the company should be fined the garnished wages + some, the executives involved fired with no severance, and an external organization should do a compliance r
556.
▲
Judge Orders Apple to Ease Restrictions on App Developers
(nytimes.com)
20 points
by
ausbah
5y ago
|
5 comments
557.
▲
by
ausbah
5y ago
reinforcement learning deals with sequential decision making to maximize future rewards, time series data is any data over time you should be more clear in your title because the two domains are very different
558.
▲
by
ausbah
5y ago
I think the issue is more when people are still having roommates when they're in their 30s and older. it's hard to have families when you still can only a afford a bedroom, driven mostly by debt and low wages if I had to guess
559.
▲
by
ausbah
5y ago
another contender is the history federal funding of research involving fetal issue and stem cells
560.
▲
by
ausbah
5y ago
the question I have is why shouldn't we believe that works in these areas might not be used as justification for deterninal policies a la eugenics? what is different this time around to prevent someone with power from using some resear
561.
▲
by
ausbah
5y ago
There have also been strong reactions against ROGD within the scientific community [1]. Like much science, the study has its own share of methodological flaws and biases that should make anyone sensible person the results of the study with
562.
▲
by
ausbah
5y ago
this is a terrible analogy the US purposefully withdrew per negotiations with Taliban - something almost everyone wanted but no president had the guts to do until now sure the US totally failed in its mission and shouldn't have been t
563.
▲
by
ausbah
5y ago
this just seems like enhanced liability protection
564.
▲
by
ausbah
5y ago
as a total outside to this sort of stuff, doesn't this have the big issue of error propagation?
565.
▲
by
ausbah
5y ago
it seems people in this thread like this idea because it could give a platform to not-mainstream-but-not-totally-crackpot ideas in physics, math, sociology, etc. The sorts of ideas that might disregarded by experts in a field but not things
566.
▲
by
ausbah
5y ago
seems like a of the all to familiar trend of a lack of affordable housing in major cities + crime at the margins in the form of squatting empty lots? either way, seems like there needs to be some sort of tax if there properties that are jus
567.
▲
by
ausbah
5y ago
Mongo has strong relational capabilities I believe
568.
▲
by
ausbah
5y ago
real men use punchcards write bug free code from the get go
569.
▲
by
ausbah
5y ago
i am sympathetic to this viewpoint because the incentives for homeowners are mostly orthogonal to that of increasing home ownership for all. why would a homeowner support increasing the supply of housing when it drives down the price of the
570.
▲
by
ausbah
5y ago
I had a general understanding of the company before interning there, just not to nearly the same depth as the article explained it because it was never necessary for my job
More ›