RC_ITR

21 ms·

211.
by RC_ITR 3y ago
It’s so interesting that people understand that it’s a machine mimicking speech, until it starts mimicking something they want to believe is real.
212.
by RC_ITR 3y ago
Old tool made one wavelength of light for lithography using a specific laser you are an expert in. New tool makes a different wavelength of light by shooting high energy beams at molten tin and all your PhD research isn't relevant to t
213.
by RC_ITR 3y ago
The injunction applies specifically to SF as it relates to a specific court case. Is that not clear to you? >SF has all the resources they need. They just need to use them. I guarantee to you the Federal government has more/better.
214.
by RC_ITR 3y ago
> Medicare, section 8 housing and social security are (relatively) new inventions I guess we just have different time horizons of what counts are historical, since Medicare is ~3x my age.
215.
by RC_ITR 3y ago
I am not implying the strawman hypothetical you proposed. I think the confusion might be that the strawman doesn't even follow logic, TBH. Were did the Sacramento Democrats comment come in? I simply stated that the article highlighted
216.
by RC_ITR 3y ago
The biggest propaganda win of the century has been the following: In a country with complete freedom of movement (and relatively cheap long-distance transportation), the mainstream has been convinced that homelessness is only caused by th
217.
by RC_ITR 3y ago
>As San Francisco progressives are learning, government can build more shelter, but that doesn’t mean the homeless will use them. This is literally the opener to the article. The whole argument it is making is a complete mess.
218.
by RC_ITR 3y ago
As the article states: >San Francisco is under a federal injunction that bars officials from enforcing laws against camping or sleeping in public spaces as long as its homeless population exceeds available shelter beds. As we recently ex
219.
by RC_ITR 3y ago
Interesting this article primarily cites federal policies that have led to the current situation and then vilifies Democrats in Sacramento and Progressives in SF. No comment on the politics, just seems like an inconsistent and thin argume
220.
by RC_ITR 3y ago
If you’re a government that sets a fixed exchange rate, one way you do that is by saying - for every dollar you give me, I will give you 6.5 RMB. Since they’re the entity that makes the RMB, and a lot of people want to buy stuff from them i
221.
by RC_ITR 3y ago
I just use 'degradation.' Not everything has to be jargon.
222.
by RC_ITR 3y ago
> that 85% of all property crime is committed by less than 600 individuals. I’d like to know how an organization that claims they cannot find/arrest the suspects in nearly any property crime can also claim with such certainty that t
223.
by RC_ITR 3y ago
I agree with you but it does feel so disappointing that there’s all this work going into something that’s just regularity arbitrage. The USD does all the same things that you describe crypto as doing for Argentinians, so now they just have
224.
by RC_ITR 3y ago
It’s amazing that people accept we live under capitalism and accept that MIT physics graduates will work 80+ hours a week at a quant trading firm just to make money, but we somehow believe that people who have no other options won’t resort
225.
by RC_ITR 3y ago
Guaranteed housing for all people? EDIT: Just for context, relatively free-market oriented Singapore and Hong Kong both invested heavily in government funded housing to solve similar homelessness crises in the 20th Century. So much so tha
226.
by RC_ITR 3y ago
To answer the headlines question: no [0] San Francisco’s “Crime Woes” will continue to exist until US National policy creates a stronger safety net for the least fortunate among us or San Francisco’s politics take a hard turn right toward
227.
by RC_ITR 3y ago
The argument from the article is that the fairy circles live longer thank the termite colonies persist, so you’ll always only see termites at some of them (both sides agree termites are sometimes but not always found).
228.
by RC_ITR 3y ago
A huge misconception is that MoE is an ensemble of discrete models, when it is in fact multiple FFNN modules that share an attention and embedding module. Basically the idea is that there's some pars of the model (attention/embedd
229.
by RC_ITR 3y ago
Interesting that the video of levitation isn't mentioned at all in this rebuttal. Assuming the scientists acted in good faith and this isn't a complete scam, maybe we found something else that doesn't conform to our curre
230.
by RC_ITR 3y ago
He’s just some guy at an aerospace startup who is trying to make some lol. Problem with his approach is that the synthesis requires some materials that are restricted to academic/scientific labs (like red phosphorus), so he’s probably
231.
by RC_ITR 3y ago
Fascism has the effect of instilling a lot of order on public spaces, but the concern is that historically, cages are never truly built for just one group. A lot of people calling for the death of drug dealers assume that they themselves wo
232.
by RC_ITR 3y ago
A lot of people are talking about very cool practical use cases, but I'm here just thinking 'all of our trains could float, wouldn't that be neat?' [0] [0] https://en.wikipedia.org/wiki/SCMaglev
233.
by RC_ITR 3y ago
To put a big fat disclaimer on this - N is <50% of the entering Freshman class, data is self-reported, the statistics (like self-reported race) do not line up with official Princeton statistics. My guess is that if you're a legacy t
234.
by RC_ITR 3y ago
Yes, but I think your comment has the foundational misconception that it's the first or even main place where bias is put into models. LLMs are just pattern identifiers and repeaters. They are trained on inherently biased training
235.
by RC_ITR 3y ago
Both do backpropagation, the difference is what you are backpropagating towards. Think of it this way - there are an equal number of rude and polite comments online (actually probably way more rude ones). If a model is trained on that dat
236.
by RC_ITR 3y ago
>A quant trader analyzes real world events, both inside and outside the market, to predict stock prices. This is a proven and successful strategy, and in some ways it is essentially automating what people already do with fundamental anal
237.
by RC_ITR 3y ago
RLHF does change the parameters. The way to think about it is that backpropagation changes the parameters of a model so they get closer to some sort of desired output. In pre-training and SFT, the parameters are changed so the model does a
238.
by RC_ITR 3y ago
This is interesting, but said differently 'when we build models to do a really good job of representing 32 words/tokens as vectors, you can very easily backsolve for 28/32 words just using the vectors. These results are no
239.
by RC_ITR 3y ago
GP likely means training transformers on raw data (similar to protein folding transformers) to find patterns that humans cannot (due to lack of context, bias, or whatever). Problem with the assumption though is that transformers are good at
240.
by RC_ITR 3y ago
This was very noticeable in sharing. Early days - share anywhere with a single link, autoplays in most messengers, no account needed to scroll forever, etc. It was almost comical how quickly they went from that to more or less total lock-do

More ›