HN user

fitzwatermellow

4,025 karma

email: fitzwatermellow@gmail.com

Posts500
Comments249
View on HN
www.businessoffashion.com 9y ago

Nasty Gal: What Went Wrong?

fitzwatermellow
1pts0
arxiv.org 9y ago

Neuromorphic Silicon Photonics

fitzwatermellow
6pts0
www.nature.com 9y ago

Drilling into Mexico’s Chicxulub Impact Crater (2016)

fitzwatermellow
2pts0
www.newscientist.com 9y ago

The World in 2076

fitzwatermellow
57pts66
www.youtube.com 9y ago

Chrome Dev Summit 2016 Keynote and Session Videos

fitzwatermellow
2pts0
thetalkhouse.com 9y ago

Arrival Screenwriter Eric Heisserer on Adapting Ted Chiang's Story of Your Life

fitzwatermellow
3pts0
pubs.acs.org 9y ago

Neural Networks for the Prediction of Organic Chemistry Reactions

fitzwatermellow
108pts31
www.whitmanarchive.org 9y ago

Chants Democratic and Native American, by Walt Whitman (1860)

fitzwatermellow
1pts0
newsroom.fb.com 9y ago

Facebook's AI Roadmap: Accelerating Innovation and Powering New Experiences

fitzwatermellow
3pts0
www.popularmechanics.com 9y ago

Inside JPL: NASA's Crazy, Kooky, Legendary Research Facility

fitzwatermellow
1pts0
anthology.rhizome.org 9y ago

Net Art Anthology: Restoration of Historical Works from 1980s to Now

fitzwatermellow
1pts0
www.youtube.com 9y ago

Oculus Connect 2016 – Keynotes and Developer Sessions Playlist

fitzwatermellow
1pts0
arxiv.org 9y ago

Neural Symbolic Machines: Learning Semantic Parsers with Weak Supervision

fitzwatermellow
94pts8
blogs.windows.com 9y ago

Microsoft Joins “JPEG 3D” Group at Khronos to Contribute to GlTF

fitzwatermellow
1pts0
backchannel.com 9y ago

Google: Our Assistant Will Trigger the Next Era of AI

fitzwatermellow
2pts0
ghslsys.jrc.ec.europa.eu 9y ago

Global Human Settlement Layer: Open Dataset of Human Presence on the Planet

fitzwatermellow
1pts0
www.cell.com 9y ago

Bayesian Brains without Probabilities

fitzwatermellow
3pts0
www.thin-ghost.org 9y ago

Collected Ghost Stories of British Horror Writing Master M. R. James

fitzwatermellow
2pts0
ctext.org 9y ago

Chinese Text Project Adds 5M+ Scanned Pages from Harvard Yenching Library

fitzwatermellow
1pts0
www.youtube.com 9y ago

Melding Mind and Math: Eric Kandel, Jim Simons and Alan Alda in Conversation

fitzwatermellow
1pts0
arxiv.org 9y ago

A Learned Representation for Artistic Style

fitzwatermellow
1pts0
www.udacity.com 9y ago

Udacity Artificial Intelligence Engineer Nanodegree Program

fitzwatermellow
3pts0
floooh.github.io 9y ago

Altai: Modern 3D API Wrapper Over WebGL

fitzwatermellow
1pts0
www.openias.org 9y ago

Bayesian Wi‑Fi

fitzwatermellow
3pts0
aging.ai 9y ago

Aging.AI: Deep Learning Age Prediction Using Simple Blood Test Results

fitzwatermellow
1pts0
eyewire.org 9y ago

Eyewire: A Game to Map the Brain

fitzwatermellow
3pts0
dataverse.harvard.edu 9y ago

Harvard Dataverse: 65K+ Research Datasets Available for Sharing

fitzwatermellow
3pts0
sites.google.com 9y ago

Computational Imaging (Berkeley CS294, Fall 2016)

fitzwatermellow
3pts0
www.petuum.com 9y ago

Petuum: Distributed Machine Learning Platform Closes $10M+ Series A

fitzwatermellow
3pts0
reallybadchess.com 9y ago

Really Bad Chess: Just Like the Classic Game, but with Random Opening Pieces

fitzwatermellow
6pts0

Dedicated probabilistic modeling environments are also gaining ground and may become standard in the near future.

The holy grail is something like: feed in some data, or parameters and have an algorithm generate the corresponding correct Bayesian inference and posterior distribution. It's very easy for scientists, even with years of knowledge and experience, to implement things incorrectly ;)

Check out Stan and Figaro:

http://mc-stan.org/

https://www.cra.com/technical-expertise/probabilistic-modeli...

Can also recommend The Crown on Netflix, with American actor John Lithgow masterfully interpreting the Old British Bulldog. There's even an entire episode devoted to the affair of Sutherland's portrait commemorating Churchill's 80th birthday, considered one of the art world's great lost masterpieces:

Secret of Winston Churchill's unpopular Sutherland portrait revealed

http://www.telegraph.co.uk/news/winston-churchill/11730850/S...

Training data? After all, AlphaGo trained on a database of over 30M expert human moves. I suspect one championship round from Team EnVy is worth billions of iterations of random exploration ;)

Kudos to both Blizzard and DeepMind. Anticipating a lot of fun with this. StarCraft 2 could indeed become the AI pedagogy standard.

Something useful, indeed.

Perhaps a teaching-focused site that explicates all the tips and tricks you've gleaned about atomic microscopy. Maybe featuring a WebGL microscope simulator. And extensive Youtube tutorials for beginners.

Or a data bank. Resources that would appeal to researchers rather than students. Modelled after something like the Electron Microscopy Data Bank:

http://www.ebi.ac.uk/pdbe/emdb/

Your goal is simply to convey that when it comes to this particular characterization technique, you're the world's #1 expert. Not so different than the inbound-style, content-rich influencer marketing all of us are seeking to master here ;)

Good luck!

A fine "beat" poet as well ;)

https://www.poetryfoundation.org/poems-and-poets/poets/detai...

Had he come of political consciousness during the Digital Age, rather than the Cold War, his message would be exactly the same.

Any technological advancements that take us further away from the Mysterious Unity, the Mother of All, will necessarily bind us in webs of illusion and ultimately lead to disunity and strife.

Totally 100% well deserved. Just imagine the 13 year old out there discovering "Masters of War" or "Idiot Wind" for the first time today. Their entire point of view is about to change. Mind upon to hitherto unknowable vistas. Ergo, the absolute benchmark of what Great Literature strives to embody.

Also worth checking out is Dylan's controversial speech upon receiving the the MusiCares Person of the Year 2015 award. I wonder what he'll have to say to the World when he gets to Stockholm ;)

Read Bob Dylan's Complete, Riveting MusiCares Speech

http://www.rollingstone.com/music/news/read-bob-dylans-compl...

And of course some essential viewing. D.A. Pennebaker's seminal documentary classic Don't Look Back which follows Dylan on his concert tour of the U.K. in 1965.

Inside Criterion's Incredible Restoration of Dylan Doc 'Don't Look Back'

http://www.rollingstone.com/movies/news/inside-the-restorati...

And from that timeless English tour, a masterful performance ;)

"It Ain't Me Babe" (Live, 5/7/65 - Free Trade Hall, Manchester, England)

https://www.youtube.com/watch?v=4d8o8vNTNao

Ermanno Borra, who first speculated that periodic Extra-terrestrial signals could be detected using Fourier transform analysis on all those old piles of astronomical spectra data we have lying around, also states that "technology now available on Earth could be used to send signals having the required energy to be detected at a target located 1000 light years away"! I vote we try running that experiment for a more conclusive answer to age old conundrum: are we truly alone in the universe?

William Gibson stated his famous maxim about the future already being here way back in 1990. Just last week, Lagos, Nigeria celebrated its own "Startup Week". Currently there are over 2000 technology accelerators spanning the globe. But the "future" you and I are living in barely appears in the halcyon dreams of the vast majority of Earth's citizens!

Stanford economist Paul Romer once theorized that breakthrough technologies don't really become productive for a generation. That lag time has shrunk to about a decade.

So instead of obsessing over the "vertical" invention, think about what it means to solve the "horizontal" problem. That in less than ten years time, something you take for granted today could be at the fingertips of 4B+ people ;)

There are a lot of great "visual" programming tools you can play with, often right in your browser. Typically designed for kids (Aged 6 to 106)!

MIT Scratch: Create shareable stories, games and animations

https://scratch.mit.edu/

Minecraft: Hour of Code

https://code.org/mc

Processing: flexible software sketchbook for learning how to code within the context of the visual arts

https://processing.org/

Google Blocky: the library for building visual programming editors

https://developers.google.com/blockly/

SolveSpace looks exactly like the sort of "user-unfriendly" tools I build for myself and am absolutely personally productive on but would never release to the general public. Maybe I need to re-think this ;)

That said I'm getting back into GUI editors. Anaconda for research computing. And Atom for text and code. Including Markdown.

The advantage is Atom's module system. 5000+ third-party creations. Including really edge cases. That, I believe, is due to the reach of JS/HTML/CSS.

It's a good question. Perhaps the best question anyone can ask of themselves in relation to "why are you doing what you are doing?"

And the way to prove if it is true or not is by "reverse induction." Remove one-by-one all the supposed "fruits" (wealth, glory, honor, respect, celebrity, etc) of your labors. And determine if it is still really important to you. Or to Us.

Alan Watts ~ Egocentricity In Humanity

https://www.youtube.com/watch?v=AkhR9c17ibE

Interesting to package this as a firefox plugin. But the filters are pretty basic. Most can be implemented using an open library:

http://evanw.github.io/webgl-filter/

https://inspirit.github.io/jsfeat/

Next step is definitely trying face detection in the pipeline. So you can do Snapchat style filter art ;)

But its super hard getting it to run in web media stream capture at even 30 fps!

WebCV: https://github.com/jamt9000/webcv

Well. There's enough quality content in this thread to start a dedicated cable television channel, a la Viceland ;)

Not sure if it's my favorite. And the subject is more technology than "tech". But the talk that keeps haunting me is Michael Dearing's lecture from the Reid Hoffman "Blitzscaling" class at Stanford:

Heroes of Capitalism From Beyond The Grave

https://www.youtube.com/watch?v=3vCdfa_aeI8

Dearing draws upon an obscure letter by Daniel McCallum, superintendant of the New York and Erie Railroad, written to his bosses in the 1850s. In the report, McCallum bemoans the stress and frustration of operating a railroad system spanning thousands of miles. All of the joy and magic he used to revel in whilst running a fifty mile stretch back in his home town has long since dissipated. Furthermore, the unit cost per mile seems to be exploding rather counter-intuitively!

Dearing goes on to elucidate the absolute necessity of the railroads ("the thing to know about the railroads is: they were startups once") themselves. As guarantors of civilization and progress. Beacons bringing light and reason to the dark swamps of ignorance and inhumanity. And not just in the physical transport of goods, people and ideas across the continent. But as the wealth created from that creative destruction remains the best cure for all of our other inimical maladies: poverty, injustice, disease and stagnation.

So, no pressure. But civilization depends upon you!

Links to References in the Talk:

Estimates of World GDP: From One Million BC to the Present

http://delong.typepad.com/print/20061012_LRWGDP.pdf

The Process of Creative Destruction by Joseph Schumpeter

http://www.econlib.org/library/Enc/CreativeDestruction.html

The Visible Hand: The Managerial Revolution in American Business by Alfred D. Chandler, Jr.

http://www.hup.harvard.edu/catalog.php?isbn=9780674940529

Report of D. C. McCallum to the stockholders of the New York and Erie Railroad

https://books.google.com/books?id=4Gc9AQAAMAAJ&pg=PA33#v=one...

Things As They Are In America by William Chambers

https://archive.org/details/thingsastheyare04chamgoog

Thanks for the reply!

So what I'm really after is a way to iteratively "perturb" the sample in such a way that the resulting texture will smoothly reflect those changes.

Take a look at this rough illustration:

Op-Art Texture Synthesis

http://imgur.com/a/1IKN3

You mentioned rotation and mirroring type bitmap ops in your algorithm. But how do we precisely apply a series of dynamic filters to the input to "evolve" the patch? Even using overlapping kernels it seems quite intensive! And prone to artefacts...

Great work! The fact that it captures "long-range order" seemingly perfectly is something not many have been able to do before! And the "collapse" visualization is great fun to watch.

But is your algorithm really qualitatively all that different from previous search methods (e.g. Efros and Leung), if you are still (uniform random?) sampling over the input distribution of patches?

I notice also your input textures tend to have sharp boundaries (as is common in pixel art). It would be interesting to see the results when the input has a lot of "noise", such as high def images of rocks or clouds ;)

While I still prefer search methods because they are easy to implement (and give better results), "deep" methods are definitely gaining some ground:

Deep Textures

http://bethgelab.org/deeptextures/

Combining Markov Random Fields and Convolutional Neural Networks for Image Synthesis

https://arxiv.org/abs/1601.04589

Anyone else watching LeBron James' Cleveland Hustles on CNBC? It's pretty great. The lesson: energy and commitment trump geography and privilege every time. And $125K goes a long way in Ohio ;)

http://www.cnbc.com/cleveland-hustles/

Re-formatting mid-size post-industrial cities into thriving locally-focused entrepreneurial hubs provides an alternative to the NYC, LA, DC, SF model. Instead of seeking to re-locate, why not attempt to foster what you value in your own backyard first ;)

Small Tech Hubs: Buffalo, Pittsburgh, Nashville, and New Orleans want your startup

http://venturebeat.com/2016/06/26/small-tech-hubs-buffalo-pi...