hervature

10 ms·

91.
by hervature 2y ago
I would tend to agree with seizethecheese. I don't think anyone says "Markov chain doing naive next token prediction without any semantic model " emphasis on the last part of that sentence. The idiot only says the first part
92.
by hervature 2y ago
> There is also no risk to life there. Careful with absolute statements [1]. [1] - https://www.vice.com/en/article/pkadgm/man-dies-by-suicide-a...
93.
by hervature 2y ago
Do you not see the hypocrisy in stating that someone should be able to take something partially publicly funded and profit from it while the creator of said work should not retain some rights over said profit? By extension of the transitive
94.
by hervature 2y ago
> If I make it available via the web, I can probably sell access to other scientists and engineers in my field. Out of curiosity. Does this statement come from complete ignorance of or complete disregard to copyright of the author?
95.
by hervature 3y ago
The scalper is trading volatility (they may or may not sell the ticket for a higher price) for the opportunity of a return. The venue is getting the opposite side of the deal. Guaranteed income now.
96.
by hervature 3y ago
I appreciate adding data for the crowd but "a very large proportion of students will write their first paper in their first year" is simply not true when talking about the whole population. My department, albeit not CS but stats&#
97.
by hervature 3y ago
Answered in 3rd paragraph of TFA: > The news has left dozens of Alexa Skills developers wondering if they have a future with Alexa, especially as Amazon preps a generative AI and subscription-based version of Alexa. "Dozens" ma
98.
by hervature 3y ago
> also be doable in 3 years I would like to add more color to this. 3 year PhD is very possible for a motivated individual at, what I'll call, the low end of R1 universities. That doesn't mean you cannot do good research (the O
99.
by hervature 3y ago
Tesla's most recent 10-K (numbers in millions of $): """ Automotive sales: 78,509 Automotive regulatory credits: 1,790 Automotive leasing: 2,120 Services and other: 8,319 Energy generation and storage segment revenue: 6,
100.
by hervature 3y ago
Radix economy is all about which base is the most efficient to represent a given number. It is simple to show that, for large numbers, this is equivalent to how efficient a base can represent itself, b/ln(b). Simple calculus shows this
101.
by hervature 3y ago
An extra $0.50 per meal (less than what you are saying and not just for a topping) equates to an extra $500/year per person. Either you did not grow up as low income as you think you did or your family made sacrifices to keep maple syr
102.
by hervature 3y ago
As gwern said, maple syrup is a minor luxury. My household grew up with Aunt Jemima because the extra expense could not be justified. People here saying maple syrup cannot be substituted sound as pompous as saying they only drink XXO Cognac
103.
by hervature 3y ago
People responding and downvoting don't understand "substitute != same". For people downvoting this comment, go look at the canonical examples of substitutes. Margarine and butter. McDonald's and Burger King. These are al
104.
by hervature 3y ago
I would say it is short sighted land arb. Density and land prices go hand-in-hand [1]. The causality is less clear. Higher land prices necessitates more units and being able to build more units increases land prices. Cities build airports b
105.
by hervature 3y ago
I think you missed the part about the owner was the one who went to business school.
106.
by hervature 3y ago
I think you are more or less correct with regards to electricity but the rule applies more broadly to technology. Obviously, the Amish use wheels and other tools. The problem for them (and which I think a great many people would agree) is w
107.
by hervature 3y ago
Being used as a comparison... From the abstract: > Bringing these components together, we are able to build pure CNN architectures without any attention-like operations that are as robust as, or even more robust than, Transformers.
108.
by hervature 3y ago
You do not seem to understand how patents work because of other comments thinking that you can patent something without publicly disclosing the invention. In and of itself, this comment is silly because the answer to your question "Wha
109.
by hervature 3y ago
> Would it be possible to pass a law that says you can't own a patch of land unless you develop it sufficiently for some public utility? This is exactly how mineral rights work. The federal law says you need to do $100 worth of deve
110.
by hervature 3y ago
The did it in 2 stages. Before that, you need to know some history. A fire burnt down the majority of downtown. So, they had to rebuild everything anyway. The downtown was also low enough that the sewer would flow into the bay at high tide.
111.
by hervature 3y ago
Part of the reason I dislike giving examples is that people think the argument hinges on the examples. The point of my argument is that notifying the government of everything (1099-B in this example) as it happens live and settling at tax t
112.
by hervature 3y ago
I don't know why you are being downvoted. I think it is a very reasonable question to ask how the government can pre-populate capital tax forms without being notified of sales. If the onus is on you to send in the sales documentation t
113.
by hervature 3y ago
Ok, I understand what you are saying. Essentially, you would have liked different type of perturbations tested to get a baseline effect of how much a random perturbation can get people to agree. They did do this in what they call Experiment
114.
by hervature 3y ago
I'm not really sure what you are commenting on. They tested this across multiple pairs of categories: (sheep, chair), (dog, bottle), (cat, truck), and (elephant, clock). This isn't a phenomena related to cats. The whole point of t
115.
by hervature 3y ago
I think you misunderstand the experiment. They take an image of the flower, they perturb it so that the neural network classifies it as a "cat". They take another copy, perturb it so that the neural network classifies it as a &quo
116.
by hervature 3y ago
I think most people commenting in this thread are reading the headline and/or reading the article as outside the context of mathematics. This is Terry Tao's blog and not a VC's microblog. The very first block of text is a quo
117.
by hervature 3y ago
From TFA: "" The data sets were randomly divided into training (85%) and test (15%) sets. We used 10-fold cross-validation to obtain generalized results of model performance. Data splitting was performed at the participant level a
118.
by hervature 3y ago
You are correct. What I was trying to drive at is that utility for many things does not really get inflated/deflated on pre-defined metrics. In your iPad example, what's the inflator coefficient to use when going from TouchID to F
119.
by hervature 3y ago
CPI adjusts prices on a per weight or per volume basis. Shrinkflation is definitely captured in the official metrics. The hidden inflation is in the service industry where it is impossible to capture whether or not your dentist gives you a
120.
by hervature 3y ago
It should be clarified that adaptive trial design was studied for several decades before being applied to games and called reinforcement learning by computer scientists. The multi-armed bandit papers all reference "Asymptotically effic

More ›