laughingman2

Then I guess you're not looking hard enough. Recently swift integration with tensorflow also introduced easy way to import python. There into swift. Data science engineers need performant language to do lot of stuff, and rust fits the ball. But it's hard to abandon python ecosystem.

I and others hate the pattern in python to write anything performant in C, Cython or some other language and write bindings for python. The python as binding language paradigmn doesn't work when all the stuff you want don't exist in performant languages.

TLDR; this fits the perfect use case for machine learning/data science, to use python only when necessary and hopefully migrate projects (parts of them atleast) to rust.

Give it a chance and try spacemacs (and then after using it for sometime maybe start building something from scratch if you think its too big). The eVil (vim) mode works out of the box and mnemonic bindings are easy to memorize.

Oddly I left vscode for emacs for Tramp. Tramp makes it possible for me to have autocomplete, remote repl, remote debugging out of the box (with the python layer plugins in spacemacs). sshfs + terminal never felt integrated and autocomplete doesn't work from the python environment in the remote machine. That was the moment I thought that the flexibility of emacs is hard to match with text editors having a plugin system. I wonder if something like tramp will ever be written into vscode.

For me, I knew touch typing (not a master but enough to not see the keyboard while typing), that combined with little experience in vim bindings (basic navigation hjkl, v) etc was enough to become really comfortable with spacemacs. Spacemacs's mnemonic bindings will make it easy for you to start having a muscle memory of bindings to do stuff like opening file tree, moving between windows, switching layouts, killing buffers (all essential for a good workflow in emacs).

lol! Not at all surprised, if you can think of it, there is most probably an emacs plugin for it.

Nice feature for emacs home screen or scratch buffer to display a tip of the day ..

As someone who recently saw the light, and converted to Emacs (spacemacs), I am absolutely delighted that its going strong.

Emacs is the most delightful software I have ever used. I could evolve a flexible todo management productivity workflow, start writing a journal and start writing new posts in a distraction free environment all using Org Mode in just two days. A complete win for someone who went back to paper (rather unsuccessfully) to organize to-dos after finding digital tools inflexible.

And for python development on remote machine, I get editing and autocomplete on remote systems out of the box (spacemacs) due to Tramp. I was like wtf, why didn't I start using this beautiful piece of software years back.

The way Emacs is a programmable platform really changes how I view what a good software should be. For instance, I was documenting a machine learning experiment in org mode file. I wanted to mention a metrics result generated as json by my machine learning code. I could just embed sh into the org file, execute it (on the remote machine without friction) and print result directly into the org file. I can trivially compare results in two files and record it back into org mode easily, even on a remote system.

It has made me think that we need more software which erases distinction of programming as something distinct from using software.

Emacs has tramp, which will allow you to remote edit files through ssh with your own setup. You have emacs installed in your computer and you can open any file in any system with ssh authentication inside it.

I am running emacs with spacemacs.org

People who want to work on remote access with restricted setups, checkout Emacs. Tramp mode allows you to edit files over ssh, docker, adb etc without you worrying about anything.

And if you haven't tried about Org Mode, it is not exaggeration if I say it is life changing. It can help you organize notes, todos, agendas etc.

Any extensions for emacs?

>Free Will would have to be some other special category of events that is neither determined not random. We don't have a word or phrase to describe what these events would be like other than "Free Will." However, lacking the vocabulary and math to describe it does not make it non-existent.

That is not even non-existent, more like a misconception which one comes out of by thoroughly examining what "feels" correct. Like Wittgenstein said one of the jobs of philosophy is to free us out of semantic knots. I see free will such a knot arising due to our conception of a separate self from the causality, leading us to project a conception of freedom which makes no sense when examined upon.

Exactly, which makes "free will" a semantic nonsense term. Understanding that neither determinism nor in-determinism of the universe show anything about popular conception of free will (libertarian sense) allows one to exit a semantic misconception.

Lets assume determinism is epistemological unfeasible, ie not possible to evaluate all the states due to limitations. Unpredictability makes no difference to the ontological question of whether actions are due to casual chain or pure irreducible randomness (if it exists in fundamental level). There is no other alternative. Free will is ill defined/semantic nonsense in libertarian sense of its use as there are only two possibilities.

I applaud your attitude. The other child commentors have the naive view like your brother that "enabling ads" give some of a positive impact (supporting artists) than the negatives of current ad industry.

The only way to individuals can fight tracking ads now is to starve the industry out. I am tempted to say take the fight forward and install Ad Clickers https://adnauseam.io/ in every browser you get your hands on.

I hate to be under the thrall of Elsevier software. But the shitty tactics would force me to switch to Zotero anyways. I just wish Zotero packed in a consistent cross-platform pdf annotation tool and a mobile app. I wouldn't mind paying for these.

Information wants to be free.

> Reads a little like the writings of someone who so wants to say something bigger than everyone else that they end up not saying anything to anyone.

Exactly, similar to some of the (not all though) "insights" you have under the influence of psychedelics.

Could this be a manifestation of power law at work through historical accidents/events? i.e., the amount of information in the vogue about countries is distributed as a power law with ~20% of countries like the US, China taking 80% of the share of international headlines.

But I do remember Scientific American(mention unintentional, but could be the same power law) debunking power law's presence in many networks.

If anyone from the dev team there, can you look into integrating the "make the research into production" part into allennlp. Facebook currently has fairseq, this and other nlp repos. Allennlp makes it easier to model most classes of NLP problems with a clean dependency injectable interface with most common tasks abstracted out cleanly.

On one hand, you have rabid nationalists who don't acknowledge culture is cross-pollinated and on the other rabid postmodern leftists who invent terms like "cultural appropriation" without realizing that it is part of how culture gets created in the first place.

I think by calling someone's error out as "maternal instinct" instead of arguing against it rationally, you have become the very caricature that makes their argument more valid.

I think blank slate is the common prejudice and simplistic narrative. Calling out the paper as bad science without giving any valid reason is stupid. If the argument had been that a single study is not enough to conclude the author's conclusions (which are not too broad). But one can say that to any study in a new/nascent topic.

But what is in the thread is that such a study should not have been published because it violates ethics, without pointing out the scientific flaws in such a study.

Yeah, I was referring to that thread. Without explicitly stating what's wrong with the study (in which many factors could go wrong), Or stating how/why environmental factors could still derail the study despite twin control, the twitter thread derails into "social responsibility" etc.

I have a political bias towards the left (equality, more democratic control etc), but I would never say scientific truth should be subservient to my ideology. That kind of a view has a huge stench of the postmodernist ideology on how even science is subjective like literary criticism.

Any ethical axiom is cannot be contingent on facts due to the is-ought problem (https://en.wikipedia.org/wiki/Is%E2%80%93ought_problem). But the moment some the moral axioms are used to make decisions in the real world, the decisions have to be consistent with the axioms. The effect of decisions is contingent on facts aka Science. So given a belief in induction, science can help inform what you "ought" to do with respect to your axioms.

Contrary to the subset of people in the left who blindly believe the Blank Slate hypothesis, or fear that genetic studies will change their ethical assumptions, I argue that more knowledge (genetic, sociological etc ) through science would only help you to be more committed to you ethics by allowing you to exercise it properly. I think Steven Pinker's argument against (part of) left's obsession with blank slate rings true https://www.ted.com/talks/steven_pinker_chalks_it_up_to_the_...

I think Peter Singer (of Animal Liberation fame) alludes to the same point in the Darwinian Left https://en.wikipedia.org/wiki/A_Darwinian_Left (haven't read it yet)

This study measures academic performance in non-identical twins with identical twins as the control group. It finds a significant difference in outcomes between non-identical twins than identical twins suggesting the role of genes in academic outcomes.

I am not a genetic researcher, more of a layman in this line of study. Some academics in social media (twitter) handwave and dismiss this research as inaccurate and they also feel its dangerous for such a study into genes (which might have ethical implications when applied to social policy) to be done without adequate interdisciplinary collaborations.

Irrespective of their motivations, I am interested in knowing whether such criticisms hold good for this study.

There is this anime called psycho-pass, that explores how an AI assigns jobs based on psychological profile aka Psycho-Pass, identify criminals for police to execute or jail etc.

Very good take on how things might be.

I have read articles that as incomes rise, people in India a significant percentage of whom were traditionally vegetarian start to consume meat more often. I think this is also the case for other developing countries, as incomes rise meat consumption increases.

I wish this wasn't the case, ie people becoming wealthier, consumerism encourages them to be immoral. (Regardless of what their traditional belief are)

Hope the awareness campaigns against meat reach effectively the segments that see rapid increase in incomes.

Well, the difference is the "actual monopoly" here can be voted out democratically. Give the workers of Amazon and other companies the same power to influence the board through one person one vote or a liquid democracy, then we can talk about the supposed hypocrisy involved in asking government to break up monopolies.

They don't teach Ethics in AI for skynet scenario, but for analyzing ethics in real use cases like Automated targeting in military applications, manipulating people, biases in AI models etc

I think legal contract analysis is an area where machine learning is actually needed for core features to work.

Some products like https://www.contractstandards.com/ are beginning to appear. They need document classification, text sentence level classification, entailment, clustering are all necessary for making contract analysis easier for the user.

And these products need good domain experts to identify pain points where ml can help.

In ml as API side, one potentially valuable product suite I have seen is https://aylien.com/text-api/ . Even they are building additional products like news analysis.

But I am little sceptical on whether ml as API can scale in terms of addressing specific task-specific nuances needed for products building upon them.