Calavar

42 ms·

421.
by Calavar 4y ago
Private practice is quickly going extinct in the US. It's generally not an option for US residency and fellowship graduates these days unless they are in one of the specialties that has cash payors (plastics, dermatology, orthopedics,
422.
by Calavar 4y ago
No, they key is to differentiate SEO'd pages with useful content from SEO'd pages with useless content. This is a game as old as search engines. In 2005, it meant filtering out sites that were just lists of keywords, not coherent
423.
by Calavar 4y ago
> I think you are just paraphrasing his point, so don't take this as a shot in your direction, but anybody who remembers computers 20 years ago being faster either has some serious rose tinted glasses, or was not using computers 20
424.
by Calavar 4y ago
You're applying a 21st century post-globalization lens to a 18th century world. Sure, today a German is welcome in France or the UK and seen as more or less equal. In the 1700s, there wasn't even a concept of equality for subcultu
425.
by Calavar 4y ago
There is a lot of circumstantial evidence that suggests that Kaedim's 3D models aren't AI generated, but manually crafted: https://mobile.twitter.com/danjohncox/status/156028399204932...
426.
by Calavar 4y ago
Oh man, I remember this attitude from a lot of my classmates back in undergrad. "Why are we using Java like dinosaurs? All jobs are in Ruby/Rails!" (Today it would be Node/Typescript instead of RoR) It really amazed me h
427.
by Calavar 4y ago
Same. I see a lot of links to conference aggregator websites and a couple of medical society websites. Not a single sponsored link.
428.
by Calavar 4y ago
I'm not sure I agree that there is a distinction between scaling laws and a model being "capable of scale." RNNs are Turing complete, so from that perspective they should in theory they should be sufficient for AGI. But of co
429.
by Calavar 4y ago
One big issue is how exactly we'll continue to scale. Exponential growth is hard to maintain. An example: In the Chinchilla paper [1], the authors suggest that most big transformer models are undertrained, and that we will probably see
430.
by Calavar 4y ago
The first source I quoted said a "a few hours or days," but I guess that wasn't convenient for you, so you clipped it. The numbers 5 and 10 are means. "A few hours or days" represents a range. If you look at the fir
431.
by Calavar 4y ago
The 2nd Edition was published in 1988. I would guess that this this 2012 version just adds an extra foreword and some errata?
432.
by Calavar 4y ago
I'm glad I could be of help There are some write ups out there: https://wexnermedical.osu.edu/blog/covid-19-vaccine-long-ter... > Once the body creates that spike protein using the mRNA instructions, the body q
433.
by Calavar 4y ago
I'll start with a breakdown of how RNA vaccines work. They deliver an mRNA sequence into your body. mRNA is a relatively short-lived form of RNA that is transcribed into protein a handful of times before being degraded by endonucleases
434.
by Calavar 4y ago
The COVID vaccine isn't a radical new class of treatment; it's a vaccine - something which physicians have been giving on the regular for about 250 years. The oldest treatment in modern medicine. We understand their side effect pr
435.
by Calavar 4y ago
In the US, the terms university and college are used more or less interchangeably. In the UK, a university is a collection of semi-independent colleges, each with their own students and faculty and their own departments for all the various
436.
by Calavar 4y ago
I am not a lawyer, but my understanding is that the idea that juries are for matters of fact comes from English common law. The constitution is a rather barebones framework. There is a lot that it doesn't cover.
437.
by Calavar 4y ago
Scroll bars are Web 1.0 garbage.
438.
by Calavar 4y ago
That's not how health insurance works - at least the US. Insurance companies reimburse either on a fee-for-service model, in which the hospital is paid for every procedure they do (with no cap); or bundle style model (capitation or dia
439.
by Calavar 4y ago
> Yes, it can save some typing, but programing is never hard because of typing, programing is hard because its hard to understand what is going on. I agree that optimizing number of keystrokes is a bad goal, but in my opinion that isn&#x
440.
by Calavar 4y ago
Insurance related quota?
441.
by Calavar 4y ago
First, the majority of the biomass of the earth is water, not organic molecules, and big fraction of what's left is lipids and carbohydrates, not nucleic or amino acids. So a simple multiplication of biomass by Avogadro's number i
442.
by Calavar 4y ago
No, it's not as simple as that. The probability of self-replicating RNA arising by random assembly of bases is very low. There was a good Nature article [1] published a couple years back that digs through some of the numerical issues.
443.
by Calavar 4y ago
Adding to my previous comment: If a student asked "how did life originate?" a response about about nucleic acid synthesis would be a bit disingenuous, because that covers only very, very tiny fraction of the surface area of the pr
444.
by Calavar 4y ago
It's progress, but the degree of progress is vastly oversold by the media, and sometimes even by academics. Kind of similar to those "amazing new semiconductor" articles that come out every 6 months.
445.
by Calavar 4y ago
I have always felt that applying the label "origins of life" research on amino acid/nucleic acid synthesis is extremely misleading. An analogy would be writing an article about the natural abundance of silicon on beaches and
446.
by Calavar 4y ago
I feel like we must have read two entirely different articles. The author has a very precise definition of stupidity. He very explicitly states that level of knowledge/education is completely orthogonal to being stupid. (As prevalent a
447.
by Calavar 4y ago
Based on a skim of your Twitter, it looks like you were accused of committing a lewd act with a minor. You believe that this was an unfounded accusation. The police seem to agree because they have decided not to charge you with a crime, but
448.
by Calavar 4y ago
> We'd probably need a whole new degree that's like a PhD, but whose dissertation isn't a novel extension of the field, but reproduction of other studies. I don't think that would work - those positions would go empty
449.
by Calavar 4y ago
There is an excellent three part video series about the Jan Hendrik Schön fraud on YouTube: https://youtu.be/nfDoml-Db64
450.
by Calavar 4y ago
I agree with everything on your list. Most of these problems are quite difficult to solve, but I think (2) is doable if we can convince politicians to unfreeze science spending that has been stagnant for decades. We need a new set of early

More ›