Google Brain Residency

I enjoyed the section on negative examples, and the things that don't work out. Makes me feel a lot better about the ton of things i've tried which didn't quite pan out the way I hoped.

Interesting to see that the Google Brain Residency isn&#x27;t that much more advanced than other graduate work or fellowship programs out there.<p>I disagree with him about needing a &quot;Rails for deep learning,&quot; I think it&#x27;s quite fine now in practice, especially since DL code is typically significantly less than a web app. A file for data pipeline, a file for your model, and a file for training and&#x2F;or inference. Not sure it really needs to be much more complicated than that.

I don&#x27;t really see the point in prime classification, I feel like you&#x27;re never gonna get better than doing a deterministic sieve algorithm. Even if it did do quite well in terms of accuracy, probabilistic models aren&#x27;t 100% correct so you still have to check for primality deterministically.<p>However the twin prime factorizer idea was interesting and could potentially lead to some factorization speedups (based on heuristics) if done correctly.<p>Either way though, prime number distribution is quite tricky - we should probably be looking in other number bases as well rather than just base 10.

&gt; I believe my motivating demo will be achieved some day soon—you will watch Charlie Chaplin in 4K resolution and it will be indistinguishable from a modern movie.<p>Wow, where does it stop, then? Can we upsample modern 4k films? Or deeper dynamic ranges? Stereo projections from mono images?<p>I suppose that the limiting factor is the resolution&#x2F;color depth&#x2F;perspective channels of the training images?

All the &quot;what happened to ryan dahl&quot; questions have been answered.

For context - the author of this article, Ryan Dahl, is also the creator of Node.js.

This is interesting:<p><i>The program invites two dozen people, with varying backgrounds in ML</i><p>... &quot;with varying backgrounds in ML&quot; -- does that mean that relative beginners have a chance of being accepted?<p>(I have a personal stake in this! I&#x27;m an academic researcher who is getting into ML and working on legal applications, and I&#x27;d totally apply for this if I thought you didn&#x27;t have to basically be Andrew Ng to get in...)

The colouring algorithm is very interesting, what if we give a dinosaur image as an input after learning the network with images from all the 10,000 reptile species?

I admit I&#x27;m a little disappointed that &quot;Google Brain Residency&quot; isn&#x27;t a prototype for uploading your consciousness to &quot;the Net&quot;.<p>Loved the article though. Dashed cyberpunks dreams notwithstanding.

yup, the very best machine learning models are still made from distilled postdoc tears...

This is just incredibly impressive on the surface. I wonder if it could be used for compression of videos and pictures?

Great article! I especially like how its understandable and relatable to folks not wholly within the ML space.

<a href="https:&#x2F;&#x2F;how-old.net&#x2F;" rel="nofollow">https:&#x2F;&#x2F;how-old.net&#x2F;</a> has improved a lot

<a href="http:&#x2F;&#x2F;how-old.net&#x2F;" rel="nofollow">http:&#x2F;&#x2F;how-old.net&#x2F;</a>

That XKCD feels spot on... <a href="https:&#x2F;&#x2F;xkcd.com&#x2F;1838&#x2F;" rel="nofollow">https:&#x2F;&#x2F;xkcd.com&#x2F;1838&#x2F;</a>