csantini

I see what you mean. The "existence" part I suspect is a trick. I think "That thing going on for the ride", is actually just a feeling evolved for a bodily purpose.

The body is doing the ride, with all its chemical gradients pulling the levers, and the thing we call Consciousness it's just "the feeling of the ride" which has evolved to keep some temporal/spatial unity. You could have legs, and memories, without being able to connect them to you.

For example, without that feeling of unity, the brain wouldn't not know which subject all the things are related to.

Something would be thirsty or hungry, but it wouldn't know that is the same thing with those legs and memories that it was referring to a just a moment ago.

Funny, I've been repeating around "Consciousness is just a feeling" for a while. My question is also "why has it evolved at all" ? Keep reading for a hint.

Hunger, for example, is just a feeling. Thirstiness is just a feeling. But we don't build complex theories of the world around Thirstiness. Somehow we do with Consciousness, because I suspect it tricks us, playing recursively with our thinking.

Consciousness, which nobody really ever defines clearly, it's probably just a name around a bunch of feelings we have.

It's clear why Thirstiness has evolved: to get us to find water. Not salty one.

Probably Consciousness does something like that. For example, it might be just a feeling of oneness to keep us intact, across peripherals (legs, arms) and time (now is really a continuation or whatever we were doing before, you need a feeling to enforce that).

If you want it today, Pampy does most of it:

https://github.com/santinic/pampy

Even match on Point(x, y, _)

Author here. Pampy and Pampy.js (https://github.com/santinic/pampy.js) are more similar to Lisp-style Pattern Matching. Of course they cannot provide static analysis, but they can improve understandability, and they are useful with nested structures which are typical of how you solve problems in dynamic languages (less types, more nested dicts-and-lists).

An example:

pampy.matchAll(json, {_: {age: Number}}, (key, age) => age)

Finds any object with a Numeric age, that is value of any key in a json object we don't know the structure of.

You are right! I'm thinking to backport it with some minor syntax change for python2.

Same issue here! Mid 2012 slowing down drastically since High Sierra. It's a common issue apparently! [1] I'm guessing it's a bug/feature in the WindowServer [2].

[1] https://www.youtube.com/watch?v=nBCPtcfL_Fs [2] http://www.openradar.me/35122177

This is well done, but it proves to me that we are not there yet with javascript really replacing mobile apps, even on a such a simple app:

When you click on a news and then you go back to the ng2 app it takes more than a second to load back the js code when compared to any native HN app.

This is a big limitation for a news reader: of course the javascript app cannot stay in memory when you leave the page and then come back to it.

This is also one reason why Tensor Flow is open. Yes, Google wants it to become the standard, but it's also not a competitive advantage.

The advantage is not the Deep Learning algorithm.

Some theoretical progress has been made on Neural Networks lately, but largely it's the same stuff from the 90s, with much more GPUs and data.

The competitive advantage is the cloud, and the software mess that keeps it alive.

I think Deep Learning experts will be like Linux Kernel experts. You need 1000 kernel experts in the world, but you need 10 million javascript monkeys that code what dialog message appears when the user does something stupid in some app.

TL;DR: Deep Learning will become a commodity. Software will eat Deep Learning too.

I'd like to clean up a bit the air from the hype fog:

DL is giving amazing results only when you have big sets of labelled data. Hence it will be much cheaper for companies to buy Google/Microsoft Vision/Audio REST APIs rather than paying the costs of: cloud + find data + deep learning experts. So, I don't think we will see a massive growth of DL gigs.

e.g. Google Vision API: https://cloud.google.com/vision/

Except those areas where your own CNN implementation is needed (automotive, industrial automation), Deep Learning will be another "library" in the ever increasing Software Engineering mess of gluing many open source libraries and REST apis to get something useful done. You need 1 guy training a Neural Network for every 100 software monkeys maintaining the infrastructure complexity. There are now many Software Engineering jobs because it's hard to glue and maintain publicly-available code to solve some specific business problem.

I think the the same applies for many Data Scientist jobs, which are these days more about fetching/cleaning/visualizing data than making machine learning on it.

The EU report on TTIP "State of Play": http://trade.ec.europa.eu/doclib/docs/2016/april/tradoc_1544...

EU page on the negotiations: http://trade.ec.europa.eu/doclib/press/index.cfm?id=1230

Actually that answer would be on tex.stackexchange.com.

http://tex.stackexchange.com/questions/2275/keeping-tables-f...

I'm not supporting this yet, let's see if I can make it work so that it uses also other stackexchange.com domains.

Thanks for the suggestion!

I'm not currently using tags, thank you for the suggestion. Not sure is available in the SO api, I'll check.

The keyboard shortcut is not hard to add. You could do it yourself :P

Not yet.

it uses Google. Because stackoverflow doesn't give you what you want.

howdoi beahaves in a very different way. It doesn't give you unix.stackexchange.com answers by default:

$ howdoi unzip bz2

readcsvbz2file <- read.csv(bzfile("file.csv.bz2"))

how2 switches API depending on what language you are querying.

Try Stackexchange search, it never gets you what you are looking for, they are just using some kind of Elastic Search.

I think Google instead indexes SO pages as any html, and does a much better job at figuring out what "unzip .bz2" means.

He's trying to say that Social-Democratic policies can only work in Europe because it has a more homogeneous culture ?

this is not even false, it makes no sense :D

TLDR: Given any Turing Machine you can build a Recurrent Neural Network that does the same thing.

You can then apply known facts about Touring machines, e.g. Halting Problem, so you can't generally predict when the neural network is gonna get to the final state.

Me neather, short term. But in the long run Capitalism will not want humans, we're too expensive. It also really makes no sense to work, I think this will become increasingly more evident.

Note: Marx wouldn't agree with me. He would say that the rich will still have the means of production (automation), so the poor will still do stupid jobs to get paid that little to afford the products made by automation. It's basically like it now, but I hope this will change.

There is no startup that can fix this. We need to collectively own the automation to not work.

Not working at all is the inevitable (awesome) future of humanity.

To get a grasp, nice talk by Dr.Ingraffea (Cornell) about fracking issues/absurdities:

https://www.youtube.com/watch?v=mSWmXpEkEPg

yes!

eheh it's really cool, but the user list goes a bit crazy with too many users

ehehe, it's cool :D though the user list goes crazy with too many users

Jesus, how much I waited for this! :O

Literally just deployed, absolutely love it:

http://hackurls.com/#&togetherjs=global

One liner copy-paste for community on your website. I can finally talk in real time with my users and understand why the use my website.

+1 London

I'm an italian erasmus student in London, I live in the east end (good place for startups) for 300£ per month in a nice double. Totally I need ~900£ to live well (good food, pubs and discos).

Bit of self publicity: I'm looking for young startups in the area. I'm the author of hackurls.com and I'm following other web startup-like projects. Honestly I'm a competent and passionate programmer (git pyhthon django mysql mongodb java..). Feel free to contact me at vuotomeccanico on gmail.

"Finding a person who gives a shit about Haskell is an inherently NP-complete computer science problem. It's similar in scope and complexity to the problem of trying to find a tenured academic who didn't have the bulk of his or her work done by uncredited graduate students."

Are you using hackurls.com source code ?

Thank you, there is a tooltip only if it's a AskHN news (news without a link).