lefnire

By the creator of Habitica. https://github.com/lefnire/gnothi with NLP helpers at https://github.com/lefnire/ml-tools. Of course self-promotion, but I think y'all huggingface lovers might find the NLP code useful.

Read on the treadmill.

Or videos, Udacity courses, etc - hard stuff, not novels. Two birds one stone, blood-pumping and endorphins helping maintain focus. I get a quality hour of education every day, where before it was a constant TODO.

I was prescribed Aderall for ADHD growing up; 14-25yo. Towards the end doctors were loath to continue my prescription, until finally a doctor refused and I couldn't get it since. Their story has always been: it's an amphetamine, and comes with all those risks and health concerns. Particularly around heart health.

When I took it, it felt like that movie "Limitless". Superpower concentration. I took their word on the health bit though, no free lunch, so I stayed away. I developed A-fib (Atrial Fibrillation) at age 30, which is very rare at that young. Could be any number of things, most notably genetics (though I'd be my family's first); but doctors to whom I mention Aderall all have this "ahhhhh" reaction. "Could be something else, but if I were a betting man..."

Frankly, I've always figured the way Aderall abusers abuse - here and there, for finals or work deadlines - couldn't be that dangerous, unless you get into the habit. I (and many others) was prescribed 1x/d for ~10 years. Seems to have caught up to me, but that's some relative heavy usage. I certainly don't condone, just brain-dumping experience.

I'm with the others on this. Never mind the cringe - he's all show, so much so I think he's bluffing (doesn't know ML). He amps up on "character" so much you're excited for the knowledge drop - when it comes, it's so fast and technical there's nothing to gain from it. The adage "if you can't explain something simply you don't understand it" applies. I was hoping he understood ML enough to boil things down; instead he spews equations and jargon so fast (1) you don't catch it, (2) I think he's just reading from a source. He doesn't go for essence, he goes for speed - and that's not helpful.

Again, the cringe isn't the problem directly; but that it's a cover for his bluff. The result is a not-newbie-friendly resource.

* Course: fast.ai (http://course.fast.ai). Practical, to the point, theory + code.

* Book: Hands-On Machine Learning w/ Scikit-Learn & TensorFlow (http://amzn.to/2vPG3Ur). Theory & code, starting from "shallow" learning (eg Linear Regression) on sckikit-learn, pandas, numpy; and moves to deep learning with TF.

* Podcast: Machine Learning Guide (http://ocdevel.com/podcasts/machine-learning). Commute/exercise backdrop to solidify theory. Provides curriculum & resources.

I wouldn't compare AI to mars colonization. AI is coming in strong, we've made tremendous progress - mars colonization is still in its infancy / theoretics. AI's a constant-moving target of a definition; by all accounts, we've "achieved" AI already if you'd ask someone from 50 years ago. Art, music, conversation, research, ... If he wants to say "what if the Singularity never happens," that's fine and good - but it just seems weird to me to say "what if AI never happens." It's like saying "what if self driving cars never happen" just because he's not yet driving one.

False-starts: in this regard, AI is like VR. VR had its own winter too, after Virtual Boy and the like. We're in VRs second stand; same as AI. And in both cases, both are making a very strong case, and making lots of money. I'd put my money on both horses now.

With current VR content, these robots will be damn fine archers.

AI

So much value. Cross-platform compatibility (browser/JS, server/Node, mobile/React-Native, robotics/Johny-Five, etc). In-built asynchronous execution of nodes (a boon in ANN architectures).

Then there's dev mindshare. So many people know JS, empowering them would add bodies to meet rising ML demand. I learned Python specifically for TensorFlow. Python's easy to learn, but like any language takes much time to master. I've mastered JS, so Python was a frustrating little reset.

All that said, this cazala/synaptic project doesn't look promising to me save as showcase. Better to focus on exposing JS APIs on existing computation-graph GPU-runnable frameworks, eg node-tensorflow (https://github.com/node-tensorflow/node-tensorflow).

Or possibly 'extrapolation'. We've seen science explain magic time and again: eg, what was previously an evil spiritual infestation is now a bacterial infection. Think on the analogy, a metaphysical phenomenon became physical (and observed/manipulable). Unless you're a dualist, you buy that the brain (a science-accessible object) equals the mind in a fundamental way (from MRIs, brain damage, etc). Whether it's connected to a separate physical phenomenon yet unobserved, or creates the mind by emergence in a way that's not inductively accessible but only deductively (through information theory and the like, per this article). So yes, 'faith' in science to do what it does best - but 'extrapolation' from prior scientific achievements in explaining magic.

I too hate when people dare to dream out loud about interesting unsolved riddles in the universe. For some, sci-fi psuedoscience is their inspiration into the field, boosting their achieving the impossible - take Musk. I recently landed my first machine learning job, brought here precisely because I think synthesized consciousness is possible, swayed by none other than this community's most hated quack: Ray Kurzweil. Many choose science because they're inspired to achieve incredible (literally "not credible") things. If journalists should shut up, I wouldn't have my rewarding job. I've learned to ignore psuedoscience whistle-blowers, they just sound curmudgeonly to me.

How has it fared for you so far in the job market, have you started seeking yet / talking to recruiters or employers?

I don't have anything certain, nor have I seen anyone answer this certainly - though this question has come up often. Obviously MOOCs will eventually be the way - huge companies are getting behind the movement; courses taught by best-of-best (eg Thrun, Ng); sustainability, etc. But I don't think we're quite there yet - I'd give it 3yrs. A recurring answer by hiring managers and recruiters is that they don't (yet) respect nanodegrees, at the various companies they recruit for. A Masters is much more respected (and looks like the majority minimum required degree for a decent ML job; no need for PhD, good luck with a BS). One option I'm very seriously considering is Georgia Tech's online MS "OMSCS" https://www.omscs.gatech.edu. It's a legitimate accredited MS at $7k (more expensive than Udacity, but _much_ less expensive than most MS programs). TMK it actually uses some Udacity courses in lieu of actual courses - they're partnered (hey, it might actually just be a nanodegree disguised as a university MS). I think it's sort of a transition from academia-proper to MOOCs, and it's respected by employers. So that would be my personal recommendation.

I'm going to be doing a lot more research in coming weeks. I'm going to publish my findings to my podcast http://ocdevel.com/podcasts/machine-learning and maybe drop what I find here too. Hopefully there will be some more answers here to pool from.

Thanks! Right, I aim to be more a syllabus & high-level than deep-dive. I haven't found much of its kind, and with so much commute/chores/exercise seemed like a hole worth fillin'.

Google never gets any credit. I used Google Now before Siri exploded the world. I listened to a machine learning series, which started like: "... applications include Facebook's facial recognition, Amazon's product recommendations, Google's image search, and Apple's self-driving car." Apple's car? Google is the world king of ML, and you gave them image search?

+1k. Don't need anything else. I think the simplicity's a mental hump for many, as I don't get why it's not more popular.

Please DO learn to code - https://jobpigapp.com/blog.html#/4

Functional proof-of-concept; I wanted to gauge developer interest before getting too deep.

A project for building lists of things to be used in developer projects (Creative Commons). Think of those times you need data: locations (countries to cities), professional industries and their skills, insurance companies and their plans, etc. Sourcing these data across the internet lands you gobs of CSVs & XLSXs; JSON, SOAP, XML APIs (some costing an arm and a leg!); copy-pasta from Wikipedia... it's horrible. They're data in the public domain, c'mon.

With CC-Taxonomy, anyone can add a list (say "JavaScript Frameworks" and children). The community can add items, vote on items (aka relevant / appropriate), comment, and suggest edits. Most importantly, at any time you can download any list's latest in various formats (JSON implemented, CSV & YAML pending).

If it's something you're interested in, make an appearance - it's open source, and could use help! Also, the name is bad :) Suggestions?

GTK. Yeah, I think the way location is handled has something to do with this (https://github.com/lefnire/jobpig/issues/1). I'll investigate in coming days, thanks for pointing out

I created Jobpig as a Pandora-like (thumb up/down) to filter jobs based on preferences, increasingly personalized over usage. It works more like Pandora than other learning boards; where jobs match users via hard-coded features (a la Music Genome Project) rather than collaborative filtering. These features include: location, commitment (eg full-time), company, source (eg stackoverflow), skills (eg python), and remote.

Eg I personally seek React, Python, Postgres, Remote, Part-time, Contract. That's combo's a tough ask on most boards; but Jobpig finds the closest match to my criteria, and it's working quite well for me. It scrapes these boards[1] (discuss[2]), and employers can post custom jobs. It's open source[3]. Any feedback would be greatly appreciated!

[1] https://github.com/lefnire/jobpig/tree/master/server/lib/ada...

[2] https://github.com/lefnire/jobpig/issues/5

[3] https://github.com/lefnire/jobpig

>learning materials have never been cheaper and self-taught people are the kings of tech

I've had English & PoliSci coworkers who couldn't find rewarding work in their degrees; so they self-taught tech and landed well-paying jobs in short order. Smart cats. I remember one venting about an English-degree colleague berating him for his privilege. Same degree; A made a choice, and B blames privilege. It's a general sentiment I see towards coders, like we were born with a laptop and capital.

You sound like the kinda person who says "AI will never drive," "AI will never play Go." True there's a lot of hype, which ML experts are concerned may lead to another burst & winter. On the flip-side there's a lot of curmudgeonly nay-sayers such as yourself at which ML experts roll their eyes and forge ahead. What I find is both extremes don't understand ML, they're just repeating their peers. ML is big, and it's gonna do big things. Not "only Go", not "take over the world"; somewhere in between.

Seems like a verbiage quibble; like the whole argument boils down to his original difference of exponential vs true mathematical singularity. He recognizes that we'll see superhuman intelligence (AGI), which will very likely grow and grow; but that said growth may be slow, and even cap. IMO having a growing AGI is the real point, exponential/singularity is a side conversation.

Human intelligence grows slowly (as he pointed out); but our technology grows rapidly, as it builds on previous generations. Few computer programmers can build a computer from scratch (circuits to disks), but their contributions continue to explode the information age. So even if an AGI grows slowly, it's contributions will stack (and possibly exponentially) to _major_ effect.

Fun philosophical debate :) AI devs argue that the man responding in Chinese is consulting a look-up table (input to output rule-book). Table-lookups aren't tractable for complex AI agents, and so must be boiled down to algorithms which yield the same results (and can be represented in code more compactly / efficiently). If the man knew the algorithm to convert input to output, then he knows Chinese.

IBM, Facebook, Nuance, and most importantly: Google Google Google. "Google Search Will Be Your Next Brain"[1] is a good start, plus it's follow-up[2]. Some lay-of-the-land, least-to-most cool IMO:

* Facebook's interest is in AI for social purposes (face recognition, NLP), rather than AGI (reason DeepMind turned them down). But they're still solidly in the field / a worthy target.

* IBM has been interested in AGI for a very long time, hence Watson. NLP, neural nets, machine learning, etc. Dig into Watson, lots of fun here.

* Nuance. Smaller, Dragon Speak. NLP & Markov Models primarily, but their creator Ray Kurzweil (now at Google) is the biggest AGI champion you'll see. He started this Singularity religion based on what AGI will bring to bear, see "The Singularity is Near" and "How to Create a Mind". Super fun stuff; inspiring follow, if not exaggerated. Both Nuance, and a ~similar company Wolfram Alpha (which Siri uses) would be really cool targets.

* Google. G's always been interested in AI for NLP for translating web queries & mobile speech queries; speech synthesis for play-back; computer vision, robotics, & machine-learning (neural nets) for self-driving cars; you name it. If it's AI, G's on it. Importantly, one Geoff Hinton recently tweaked an AI algorithm which unblocked a clog in the field, and now shit is _throwing down_, catalyzing Hinton & Jeff Dean's G-internal "Google Brain" project. Shortly after, G acquired DeepMind[3], and _this_ my friend... this is the grail. Deepmind is working on some benign-seeming game-playing algos presently, feeding some findings back into search. But its CEO is not _at all_ shy about his ultimate goal for human-level AGI. In fact, it's DeepMind (well, and that sophist book "Superintelligence") that got all these hot-shots up in arms[4]. These guys are the SpaceX / Tesla of AGI. If you want to shoot for the moon: DeepMind.

[1] https://medium.com/backchannel/google-search-will-be-your-ne...

[2] https://medium.com/backchannel/the-deep-mind-of-demis-hassab...

[3] http://deepmind.com/

[4] http://time.com/3973500/elon-musk-stephen-hawking-ai-weapons...

Agreed, but it can be balanced. Dedicate 1h/d to learning, no more. Balance that 1h w/ new vs current tech. Some technologies really _are_ a major investment; while others are smaller improvements than is worth sacrificing your expertise, as you pointed out.

I've had "1h learning" as a daily task for many years now, it's how I transitioned from PHP to Node. Shiny and hot, yes, but more importantly it's one language across all the work I do, and that saves me _tremendous_ amounts of time from when I was on PHP. That was a worth-while investment. Now I'm using Angular while React is getting hot. I'm not convinced that'll boost my productivity substantially, but I have that 1h/d to learn so why not? If it turns out to slow down my productivity compared to my Angular expertise, then I'll veer that 1h/d towards something else instead (VR + Unity is sounding super interesting, or maybe I'll explore this functional programming stuff everyone's raving about).

How about in terms of adoption, momentum, team, performance, etc? Any insights / predictions on how Famo.us might stack up _against_ react-canvas?

It's built in - go to https://habitrpg.com/#/options/settings/settings and uncheck "Show Header"

I haven't found anyone else who uses it surprisingly, but Gmail Offline (Chrome App). I use it when online. It has amazing keyboard shortcuts (faster than Gmail proper), and then of course you get the perk that the data is offline once you hit the train. Actually, I think the keyboard shortcuts might be the only reason I use it...

possibility, I'll bounce that around & keep it on my mind grapes