fazzone

The Linux kernel can be configured at build time to expose its own headers as /proc/kheaders.tar.xz. Not sure if distribution kernels generally have this option enabled or not.

This is pretty much how clojure atoms [0] work. It's basically a Clojure wrapper around a Java AtomicReference, but Clojure's immutable data structures make an atomic reference type really useful because it is very cheap to read a "snapshot". It doesn't do upfront allocation, because like you mentioned, that requires you to have some knowledge about how the accessing code works. Additionally, whatever you are doing in Clojure is pretty likely to allocate memory anyway, so it probably wouldn't be that beneficial.

[0] https://clojure.org/reference/atoms

The problem is that the condition is parsed as

  (user=671156 AND permission=16 AND org=101) OR 102

  org IN (101, 102)

Some years ago I worked at a technology-oriented summer camp where I personally taught several bright young children to program. Minecraft mods were an incredibly popular project and a big driver of interest for many of the kids. Many of my counselor colleagues got into programming themselves modding Warcraft III and other games of that era. The absolute conversion rate may be low because the playerbase is so huge, but I think something like Minecraft is definitely the first (or zero-th) step in the journey of a lot of kids these days who become interested in technology and will later develop into the next generation of programmers & technologists.

This sounds like sort of reasonably OK layman's explanation of Weak DH https://weakdh.org/

In Clojure unresolved names are a compile failure, and Spec would not be involved. Spec is just a nice library for writing runtime assertions about data. An example would be calling a third-party API and then validating the response with Spec to make sure it matches your expectations, or even validating the responses you are about to return from your own API. Similar to a type system the Spec can also serve as a sort of documentation about what's in the data structure.

Wow, I really can't imagine an acquisition that I would less rather learn about than FB hypothetically buying reddit. Considering how FB has all the money in the world it is probably wise to start editing/deleting all my reddit comments now because once FB slurp the whole DB it's already over - the fact that I would never visit reddit again wouldn't save me.

To expand on the debugging bit, Datomic knows and can tell you the exact transaction responsible for every piece of data in the database. You can also see the complete state of the DB as of any transaction, but the complete history is more useful. It's like git blame for the database.

I think the strongest point in the article is this: "After the next breakthrough [in AI], we still won’t know how many more breakthroughs are needed, leaving us in pretty much the same epistemic state as before." That means that if we aren't prepared to start work on AI alignment now, there's not likely to be any sort of future event that will convince us of that.

Paredit is great but it is not the same as AST editing, it even describes itself as a 'cheesy imitation,' see https://github.com/takeoutweight/paredit/blob/master/CREDITS

Isn't this what IncrediBuild does on Windows?

I am playing around with an idea like this for editing Clojure in Emacs with cider-mode. You have to manually build the 'thread' that such a code unravel-er would return, but then you can read and edit functions from many files in one buffer. You start with the highest-level function you are interested in, and then pull the definitions of functions it calls into view as needed.

https://github.com/fazzone/multifiles.el/tree/cider-hacks

A guideline like "three arguments to a function" says nothing about the type of those three arguments. You are right that there are many jobs that require 8 or more pieces of information, but often those parameters can be grouped into structures which each hold multiple pieces of information. The point is not to create types for every function just to hold their arguments, but that well-thought-out data abstraction can simplify things by factoring data into named structures that are used in multiple functions. For example you could have a drawRectangle function which takes four parameters (x1, y1, x2, y2) of some primitive numeric type, or you could create a Point type and use just two of those. This improves clarity because it is obvious just from the signature of drawRectangle(Point a, Point b) that you are specifying the endpoints of the rectangle, whereas four numeric arguments could mean (x1, y1, x2, y2) or (x, y, width, height).

I think that Google's TPUs (Tensor Processing Unit) use Altera FPGAs (speculation based on [1]). I'm sure there's some degree of collaboration if Google is filling cabinets with high-end FPGAs.

[1] https://twitter.com/danluu/status/705802676945072129

My point is that they far over-prioritized fuel economy. If you actually drive like that you will be spending a lot more time changing gears than necessary, and that almost any acceleration will require a downshift. I wouldn't recommend trying to stretch the gas mileage in the snow, for example, since you might need power to recover if you start to slide.

gear-shift indicator in the dash.

My VW has this. It's useless, I never look at it. It is very simple and clearly intended only to give the best results on the controlled gas mileage testing - I believe that if a manual car has a gear-change indicator, the EPA test drivers are supposed to follow it. The advantage of a human selecting gears is that they can use more information than just the speed and engine rpm; for instance there is no point shifting into 6th gear if you see a big hill right in front of you. It also never indicates for the driver to skip a gear. Maybe you will save a few bucks on gas if you follow the indicator religiously but you'll probably break even after you have to replace your clutch early.

There's probably no reason to restrict the computer the below human pro players (200-300 APM). My guess is the engine just wasn't designed with superhuman APM in mind and could be exploited. If that's the case then the AI would in some sense not really be playing the same game as the humans.

I have seen this reported as a rumour with circumstantial evidence to back it up (why would they release a brand-new flagship product with a 2-year-old chip). Has it been confirmed anywhere yet?

Well, a lot of the papers published in the field present results like "We designed a neural net to perform <task> and achieved X% accuracy." The design of the net is novel and interesting enough to merit its own publication. If there was some sort of theoretical framework, results like that would not be interesting, because presumably the theory would explain which NN architectures are good at different tasks and why. I think that we will get there eventually, but right now I don't we have enough data for patterns to emerge and hint at some sort of Theory.

Quick, name three uses for this technology that aren't annoying advertisements!

Have any of Altman's critics taken a step back and realized that they are telling him that he is morally obligated to terminate a productive business relationship over political differences? Are they suggesting that half the country should blacklist the other half and not refuse to do business with them? I think views like this are a result of us segregating ourselves into very narrow filter bubbles - I think it is possible that Marco Arment for example does not personally know any Trump supporters. Scott Alexander has an excellent post about this which I highly recommend (see section III specifically): http://slatestarcodex.com/2014/09/30/i-can-tolerate-anything...

I thought your thesis was that there was no actual collusion between politicians and the media, just aligned incentives. I agree with your point about the incentives, but don't you agree that this email shows that there was in fact actual collusion going on as well? The evidence seems pretty compelling to me.

It's not in an obvious place, and for some reason it's not listed on their YouTube channel, but they do have a timelapse video showing the logo from the main video being cut from start to finish. (https://www.youtube.com/watch?v=x7XafvOKoIU) Looks pretty good to me.

I agree that 5x10 is small and hard to read, but what use would a bitmapped font have for subpixel rendering?

Re: #3, I think when Reddit first implemented comment tree collapsing they also put the button on the right, but I saw a thread where someone pointed out that since the length of usernames is variable, putting the button on the left makes it impossible to mass collapse/uncollapse several comments by only moving the mouse vertically. After that they moved the collapse icon to the left, which I think is a much better UX.

The way I see it, it is certainly possible for one entity to gain enough computing power to control 51% of the Bitcoin network. However, once you do that, if you actually do any of the bad things that you could do, people will notice, and consequently trust Bitcoin less, so the value goes down. So why make the huge investment to control Bitcoin for nefarious purposes if you will make it worthless in the process?

Pascal-style strings have the length stored in the first byte, so a string can only be 255 characters long. Netstrings have the length in text at the front of the string instead, so you need only preallocate a buffer of 20 characters to store the text representation of string lengths up to 2^64 bytes.

I agree. In fact, it is almost like the market has figured out 'a way for both countries to work on what they're really good at' (paraphrasing TFA). "Governments that at least partially cooperate," on the other hand, I agree would be a welcome change.

If you are interested in that sort of thing, I highly recommend the novel Cryptonomicon by Neal Stephenson - it is partly about the activities of a joint British/American counterintelligence unit whose mission is to make sure the Axis does not figure out that the Allies have broken their codes. It's also a great read in general.

Parametric type means to take the type as a parameter; parent was probably asked to use Generics in C# or Java or C++ templates.