sdenton4

33 ms·

631.
by sdenton4 3y ago
The difference is that, by design, the best (cheapest!) Bitcoin you can buy is the Bitcoin you buy today. It only gets more expensive as more people use it and the supply of new Bitcoin decreases. By contrast, the ai we have today is the wo
632.
by sdenton4 3y ago
When I was living in Kenya around 2013, open pit burning was the main way of dealing with trash... And included a lot of plastics, batteries, you name it.
633.
by sdenton4 3y ago
In grad school I knew a guy who was reeeeally into the idea of building benches out of reused plastic and cob. The basic design was plastic bottles stuffed with plastic bags acting as bricks, which are then cemented together with cob. Clean
634.
by sdenton4 3y ago
To be hosted at the Global Technical Forum for Ornithology.
635.
by sdenton4 3y ago
Use StockFish to predict the next move, and only store the diff between the actual move and the prediction. Or, better, get a ranked list of all possible moves from stockfish, and use a variable-length integer to encode the position in the
636.
by sdenton4 3y ago
I'm not sure how you get from 'the Canadian government must respect the treaties that it signed' to 'identity-based solutions.'
637.
by sdenton4 3y ago
I'm not clear how you think the ML of 2017 would have been useful for generating these videos? This points to specific actors and creators, who I expect were using whatever was cheap and easy to produce a pile of content with in 2017.
638.
by sdenton4 3y ago
"On November 23, French-Canadian outlet Tabloïd released a video investigation about Toy Monster, a channel linked to Webs and Tiaras. They confronted the videos' creators – based out of the south shore of Quebec City – who refuse
639.
by sdenton4 3y ago
Humans responding to the rewards of the algorithm, as it turns out... https://en.m.wikipedia.org/wiki/Elsagate
640.
by sdenton4 3y ago
Easy if all the weights are zero...
641.
by sdenton4 3y ago
And humans in cars are meat-boxes of stochastic behavior, prone to all kinds of dangerous regressions. I feel safer, as a cyclist, around Waymos than I do around many human-driven cars. That's even with the cyclist crash last week: it
642.
by sdenton4 3y ago
Some people are just gluttons for pun-ishment.
643.
by sdenton4 3y ago
That was Elon Musk's promise... He sucks, I agree. But he is not the whole story, here.
644.
by sdenton4 3y ago
Waymo is doing just fine...
645.
by sdenton4 3y ago
It's worth it to ask, when reading Gary Marcus pat himself on the back, how many of his predictions were actually falsifiable.... Few-to-none. Furthermore, there has been insane progress over the last few years, which presumably would
646.
by sdenton4 3y ago
These tend to be galactic algorithms. Furthermore, I don't know that they do any work involving parallelism, or considering the narrowing gap between time needed for addition vs multiplication - typically only multiplies are counted.
647.
by sdenton4 3y ago
I would imagine that breaking them up instead of dropping as a single unit could be beneficial - higher surface to volume ratio, leading to more thorough burn up in the atmosphere.
648.
by sdenton4 3y ago
I feel that people who go on and on about 'clickbait' simply don't like reading... Feed the article to GPT if all you want is a bulleted list of points. Or just skim!
649.
by sdenton4 3y ago
See also: Every T&C page ever published.
650.
by sdenton4 3y ago
(and after one more thought about it, one would zero out the row /and column/ to preserve the symmetry of the matrix, if applicable, and thus keep the eigenvalues real. and one would probably want to think a bit about whether this
651.
by sdenton4 3y ago
Just guessing, but.... A common noise cancellation technique is to throw away small eigenvalues, as in PCA. This result relates eigenvalues to the structure of the matrix, so might be helpful for reducing ev's without bothering with di
652.
by sdenton4 3y ago
Equivariant machine learning is a thing that people have tried... Tends to be expensive and slow, though, and imposes invariances that our model (a universal function approximator, recall) should just learn anyway: If you don't have en
653.
by sdenton4 3y ago
A nice way to formulate (most) data augmentations is: a family of functions A = {a} such that our optimized neural network f obeys f(x) ~= f(a(x)). So in this case, we're explicitly defining the set of desired invariances.
654.
by sdenton4 3y ago
A similarly important benchmark is DOOM fps: https://youtu.be/J2qU7t6Jmfw?si=YNuNBDS7ti8gmpqc
655.
by sdenton4 3y ago
Same; imagining a comment calling the model stupid, leading to worse LLM code suggestions...
656.
by sdenton4 3y ago
And sometimes a side project doesn't want to be a side hustle; dealing with payments and tax implications is time not being spent on the core project. It's an individual choice as to whether the time cost of accepting payment is w
657.
by sdenton4 3y ago
Paid loot boxes, obviously. /s
658.
by sdenton4 3y ago
Absolutely love this. I've been getting into mixing for miniature painting myself, but with a bit of added perspective which (I think) isn't captured here. It turns out there's more than just raw color going on here. Acrylic
659.
by sdenton4 3y ago
It sounds like you don't understand how business loans work. This is not a VC play.
660.
by sdenton4 3y ago
Presumably there's a business loan with a fixed contract and terms. That's a very different situation than an open ended commitment to maximally enrich shareholders... Pay your bills and the lender has nothing to say about how you

More ›