frankling_

That's not a way around publishing at the established venues, it's done in addition to it. Everyone still submits to the established journals and conferences because this is what is being evaluated for career advancement and funding.

I'm not suggesting to publish every paper twice. The author submits only to one of the two journals. If it's the original journal, the article appears exactly as is common now, behind a paywall, and that's it. If it's the new sibling journal, the publisher of the original journal has nothing to do with it, no copyright transfer occurs, and the paper appears only on a free and open platform.

The main issue is that currently there's no way around publishing at the established journals and conferences, because of the reputation they've built up. Funding and career advancement hinge on publications in these venues. If we accept that publishers don't offer much at all in return for the publication fees/library subscription costs (barely any editing, reviewers work pro bono, hosting PDFs is very cheap nowadays), the main issue is one of "reputation transfer".

One way to cut out the middle man would be to convince journal editors to run sibling journals alongside the existing journals, so for each "Transactions on XYZ" there would be an "Open Transactions on XYZ" (as close in title as is legal). Importantly, each sibling journal would be run by the exact same academics (who are doing the real work on tax money anyway), and according to the same process as the original journal, just without involvement from a traditional publisher. PDFs would be hosted on a site like arXiv. The goal would be that submitting to the open "sibling" would be the obvious rational choice (same people, same decision-making, no fees, open access), which in time even the funding agencies and tenure committees would have to acknowledge.

In essence, you're deriving an "ought" from an "is" [1]. I'm not convinced that the presence of an inborn instinct X justifies a "right to X".

[1] https://en.wikipedia.org/wiki/Is–ought_problem

That's a great point. If they didn't, it's clear what their actual risk assessment is.

ft off last time and just continue from that point onwards. The essential part is to leave something over for next time, instead of finis

What would be an example use of the word free that you would find appropriate?

Nope, it's really annoying at this point. Try doing a search on Google Scholar for "considered harmful" -"go to statement considered harmful".

"About 53,000 results."

I'll skip the usual list of personal anecdotes about PA failing in surprising ways I've never seen in 15+ years of using Alsa, but I will say that since a few applications have started depending on it, "apt-get purge pulseaudio" seems to always be in my shell history.

Alsa has had software mixing via dmix since 2003.

https://www.alsa-project.org/wiki/Main_Page_News

Don't worry about it, it was close enough to get your point across.

I recently reread Vonnegut's "The Sirens of Titan" and was really surprised to notice to what degree Adams drew inspiration from that book. He did acknowledge it as an influence [1], but some ideas my naive self had attributed to Adams are closer to well-done rearrangements/retellings of elements from Sirens of Titan (which certainly wasn't written in a vacuum either) in a different comedic voice. Both works are fantastic, though.

[1] http://www.darkermatter.com/issue1/douglas_adams.php

This is one reason I prefer simple guitar pedals to VSTs and such. With their instant response and tight controls, pedals feel like solid tools you can build muscle memory on. When unzipping plugins and awkwardly dragging sliders with a touchpad, it feels like it's myself who's the tool.

Insightful comment. Similarly, competitive people prefer and create highly structured environments and situations, where it is clear what to min-max for and where they stand on the highscore list. Nuance and soft qualities go out the window that way, but others can balance that out at some cost.

That has never been true: on NVIDIA hardware the 32 threads of each Warp run in lockstep, certainly not all threads. And starting with Volta, each thread even got its own program counter and call stack.

I'd really love to see this plotted as a histogram of the age of founders of the top N companies according to some metric, and maybe even founder's age vs. some monetary evaluation after a certain number of years.

Good point, a lot of work really falls below the threshold where replication would still make sense.

However, I do think there is a lot of incremental, piecemeal type of work that over the years could amount to a decent step forward. It's just that currently, from looking at the code behind quite a few published papers, the statements often just cannot be trusted enough to actually build on these works. It is my view that some subsubfields in CS sustain themselves by avoiding the most pertinent questions, because their answers would reveal that the entire subsubfield has been superseded, or was never that promising to begin with. Unfortunately, that type of noise generation is actually more profitable than a single paper saying "nope", although the value of the latter in terms of knowledge generation is enormously high.

Some ACM conferences are now handing out "badges" that mark papers where the associated "artifacts" are publicly available or have been independently reproduced or replicated: https://www.acm.org/publications/policies/artifact-review-ba...

The main issue with that right now seems to be the lack of real incentives: those badges are nice to have, but in the end you're evaluated by publications in highly regarded journals and conferences, citations, and funding brought in. Those things of course tend to let people focus on novel, fancy, exciting work, instead of the time-consuming process of making things easy to replicate by others, or even replication of other people's experiments. Incentive-wise, the publication business follows the principle of fire-and-forget.

What's needed, but I believe very difficult to introduce, is the strict expectation of independent replication. That would mean replication as part of the core review process, which would A. create a lot of work for reviewers B. initially reduce the attractiveness of whichever conference introduces it first.

One thing that would be easy to introduce in computer science would be to make it mandatory to share all code used in experiments - if you cannot/won't do that, you should not able to take part in this game.

They list the well-known machine learning trends of the past decades, then say they don't know what's next. "Maybe it'll be something other than deep learning in the next decade."

The title should be appended with ", and failing".

Just as an exercise, it would be beneficial for you to make your argument explicit. You would notice that it falls apart/there is none.

I don't follow that line of reasoning at all. "Killing in the name of religion" would be to protect or spread some sort of metaphysical belief. The equivalent would be something like going to war with another country to build universities there.

I guess this goes into definitions of consciousness, which I don't have the background nor inclination to do. But if we define it, as the grandparent does, as "first person experience of reality", then there is something special about this particular illusion. Usually, one would reject the interpretation of an experience, not the experience itself. For instance, this could go as follows: "yes, from your vantage point, the experience was that a rabbit was pulled from an empty hat, BUT what actually occurred was that the rabbit was pulled from a hole in the table below."

With consciousness, "the experience of the trick" and "the trick" are one and the same: if I have a "first person experience of reality" (which I have), then I struggle to see how it can be viewed as a trick. Due to the property of consciousness being a subjective state, it seems that eliminative materialists claim that even the EXPERIENCE of the rabbit being pulled from the empty hat is an illusion, which to me is just nonsensical (but I assume that with suitable definitions you can work around that).

Looks interesting, but reading the text is nigh impossible on my phone due to excessive javascript acrobatics.

I fail to see how the package management in a Linux distribution could be considered a walled garden. Commonly it's cited as the exact opposite of that (see the table here: [1]). What type of freedom are you missing in something like Debian, where you can edit and recompile the core components of your system at a whim, the system even supporting you through source packages, and where anything outside the package management is just a "git clone" and a compilation away? I really don't know what more anyone could ask for. Certainly we don't want to go back to the 90s way of downloading unsigned .exe files from some random guy's web site and then going hunting for matching DLLs?

[1] https://en.wikipedia.org/wiki/Closed_platform

Not sure how much these terms mean nowadays, and you could even say that the left/right positioning game is identity politics in itself. Maybe it's more interesting to consider whether the ideas have merit.

Yeah, for some models, especially small-scale ones, it can be hard to get a performance gain. There are actually ways of avoiding the locking, e.g., by staging the desired actions and then carrying them out afterwards, resolving conflicts as needed. We actually have a paper on this as well, but I'm going to stop pushing my research at this point ;)

As for the performance aspect, these kinds of simulations (typically using fixed time steps and with per-agent decision-making) are a natural fit for many-core devices. We recently did a survey on that subject [1]. There is a lot of work being done along these lines right now.

Some groups are also trying to target sequential, parallel and, say, GPU execution from the same model specification language, which could actually help with the usability of running such simulations in practice.

[1] https://arxiv.org/abs/1807.01014

Honest question: what do you mean by "communists who run US campuses these days"? Is this about bureaucracy, or particular leftist views being upheld, or something else entirely?

Isn't that a bit strict? It seems that "x is a young man's game" can be read either in an empirical ("historically, only young men do x") or normative ("only young men should do x") sense, the former of which might be slightly inaccurate because there are exceptions, but otherwise seems fine.

A view held by some seems to be that since even the empirical observation may "perpetuate" the situation, stating it is "problematic" in itself, but I don't suppose this is what you mean here.

If all three of these are research, what is engineering?