adw

11 ms·

121.
by adw 2y ago
You’re using a tool and choosing to give the tool different names based on how you feel about it. If I’m the landlord of a pub in Glasgow, and Scotland get to the finals of Euro 2024 (we won’t, but work with me here), betting on Scotland lo
122.
by adw 2y ago
I would describe every single one of these as active risk management - and the difference between one sort of active risk management and another is vibes, not reality. There’s an implied moral judgment here that gambling is inherently bad.
123.
by adw 2y ago
In gambling you pay someone to take risk for you. In insurance you pay someone to take risk for you. They are the same tool; we just give it a different name when we’re hedging from when we’re speculating.
124.
by adw 2y ago
No, the other poster is actually right. In betting on outcomes, there are two roles - bet (“long”) and lay (“short”). An insurance company is exactly a bookmaker; they are buying a bunch of risk and being paid a vig (or in the case of long-
125.
by adw 2y ago
Dust generation is a function of vehicle weight, so people and bikes generate massively less.
126.
by adw 2y ago
Not really - often what I’m doing is running something with a conditional breakpoint which will only be hit a long way into process lifetime (potentially nondeterministically), so I really do need “a session which will persist indefinitely
127.
by adw 2y ago
The first is Fatah/PLO, who are in many ways much closer to, eg, the IRA (also nominally religiously inspired) than what we understand as modern Islamist terrorist groups.
128.
by adw 2y ago
This dynamic – a political leader trying to run away from justice – has, historically been a very common way in which states fail; which is why a lot of people who pay close attention to such things are very concerned about the state of US
129.
by adw 2y ago
I believe that 'tptacek's point could be summarized as "the facts are, in a real sense, not material to this conversation, as we are operating entirely with the domain of realpolitik rather than morality". Biden, in _the
130.
by adw 2y ago
This is very cool and it's being treated unfairly, though it's also obviously not ready for prime time; it's an existence proof. To illustrate that, many people on here have been losing their mind over Kolomogorov-Arnold Netw
131.
by adw 2y ago
> First Things (FT) is a journal aimed at "advanc[ing] a religiously informed public philosophy for the ordering of society" It's (explicitly) religious propaganda.
132.
by adw 2y ago
"Obsessive" is pretty loaded language! Also: normal has shifted, as it always does, so you should expect permanent changes in behavior. It used to be normal to smoke indoors; one of the few good outcomes of Covid is that people ar
133.
by adw 2y ago
If you're at risk, you're probably better off taking the antivirals than not, and it lets you know whether you should ask your doctor for some?
134.
by adw 2y ago
You don’t know their personal circumstances. Maybe they’re in the same situation as you (I’m assuming “generally in good health”), but they are just more anxious than you and you believe that’s unnecessary. On the other hand, some people -
135.
by adw 2y ago
I think you can make a case for “Things We Lost In The Fire” by Low extremely easily; or for “After Murder Park” by the Auteurs. As a recording engineer he had a lot more range than people gave him credit for, largely because his own music
136.
by adw 2y ago
Those (hip-hop) producers are musician-songwriters. They’re more in the Brill Building/Phil Spector mode. Steve Albini recorded the songs you brought in with you. Just a different thing.
137.
by adw 2y ago
> What happens when the best methods for computational fluid dynamics, molecular dynamics, nuclear physics are all uninterpretable ML models? A better analogy is "weather forecasting".
138.
by adw 2y ago
Ian Jackson (former Debian project leader) has been on this train for about 25 years now: https://www.gnu.org/software/userv/
139.
by adw 2y ago
> They haven't (eaten the world etc). To clarify what I mean on this specific bit: the SOTA results in 2D and 3D vision, audio, translation, NLP, etc are all transformers. Past results do not necessarily predict future performance,
140.
by adw 2y ago
> And MCTS still the SOTA search algo in Go etc It's often forgotten that Rich Sutton said the two things which work are learning (the AlphaGo/Leela Zero policy network) and search (MCTS). (I think the most interesting researc
141.
by adw 2y ago
> much more inefficient The tricky part here is that "efficiency" is not a single dimension! Transformers are much more "efficient" in one sense, in that they appear to be able to absorb much more data before they sat
142.
by adw 2y ago
> The language of continuous functions is alien to most humans and it's very difficult to encode one's intuitive, common sense knowledge into a system using that language In other words; machine learned models are octopus brai
143.
by adw 2y ago
Yeah, all of those architectures are _themselves_ hacks to get around having insufficient compute! They absolutely were encoding inductive biases into the network to get around not being able to train enough, and transformers (handwaving ha
144.
by adw 2y ago
> Why is Yudkowsky taken seriously? People like religion, particularly if it doesn't affect how they live their life _today_ too much. You get all of the emotional benefits of feeling like you're doing something virtuous withou
145.
by adw 2y ago
People have advanced that argument a lot, and it's often worked for a short while; then the statistical models get better. Chess was a game for humans. It was very briefly a game for humans and machines (Kasparov had a go at getting &q
146.
by adw 2y ago
Given Python has a long history as a teaching language, that strikes me as the kind of grandiose statement without evidence which Perl fans used to be fond of. Even more so given there are approximately no “non-programmers” writing Ruby and
147.
by adw 2y ago
a) most libraries are not UI frameworks undergoing significant perpetual rewrites (and a lot of the distaste towards those frameworks comes from the perpetual rewrites) b) even given that, they share a mindset, and being able to talk about
148.
by adw 2y ago
I think they're suggesting "change the name", which neglects the value of identifiers as _things people talk about_. It's a fine approach if module names are equivalent to ISBNs, but they're not really.
149.
by adw 2y ago
Pageant is `ssh-agent`.
150.
by adw 3y ago
[Three is always funnier]( https://en.wikipedia.org/wiki/Rule_of_three_(writing) ).

More ›