zfrenchee

Webcola is great -- but no docs or support.

Seems like rigorous data science on the statistics of people graduating with humanities degrees. Doesn't that seem just a little bit ironic?

This is beautiful work!

I'm really perplexed by this refrain that CDB is "not psychoactive"

A psychoactive drug, psychopharmaceutical, or psychotropic is a chemical substance that changes brain function and results in alterations in perception, mood, consciousness, cognition, or behavior. source: https://en.wikipedia.org/wiki/Psychoactive_drug

Isn't the point of taking it that it's psychoactive?

Base editors can make all 12 conversions as of 4ish months ago I think.

Cas9 does not edit DNA, it cleaves DNA. What happens afterwards is up to the cell. The hype machine has ignored this fact.

One of my favorite George Church quotes: "CRISPR: some call it genome editing. I call it genome vandalism".

Base Editors can actually edit DNA, but only single bases at a time. http://www.sciencemag.org/news/2017/10/novel-crispr-derived-...

There's a big prize waiting for the person who can harness DNA repair pathways in conjunction with Cas9 to make precise, multi-base DNA edits. Lots of folks are working on that now.

The original paper is https://www.cell.com/cell/fulltext/S0092-8674(17)30629-3 There was a nice follow up last week https://www.cell.com/cell/abstract/S0092-8674(18)30714-1 There's a lot of good theoretical biology to be done here. I don't think any of us systems biologists are surprised about this result, but pinning down exactly the structure of the genetic basis for complex traits is going to be an interesting enterprise.

There are slots devoted specifically to folks who get workshop papers in. I registered a couple days ago.

To be fair, I agree NIPS registration is getting a little out of hand, but we should get the facts straight.

My qualm with this article is disappointingly poorly backed up. The author makes claims, but does not justify those claims well enough to convince anyone but people who already agree with him. In that sense, this piece is an opinion piece, masquerading as a science.

This is because a deep learning model is "just" a chain of simple, continuous geometric transformations mapping one vector space into another. All it can do is map one data manifold X into another manifold Y, assuming the existence of a learnable continuous transform from X to Y, and the availability of a dense sampling of X:Y to use as training data. So even though a deep learning model can be interpreted as a kind of program, inversely most programs cannot be expressed as deep learning models [why?]—for most tasks, either there exists no corresponding practically-sized deep neural network that solves the task [why?], or even if there exists one, it may not be learnable, i.e. the corresponding geometric transform may be far too complex [???], or there may not be appropriate data available to learn it [like what?].

Scaling up current deep learning techniques by stacking more layers and using more training data can only superficially palliate some of these issues [why?]. It will not solve the more fundamental problem that deep learning models are very limited in what they can represent, and that most of the programs that one may wish to learn cannot be expressed as a continuous geometric morphing of a data manifold. [really? why?]

I tend to disagree with these opinions, but I think the authors opinions aren't unreasonable, I just wish he would explain them rather than re-iterating them.

Who is behind this?

Put Fuchsia and Android green on a color wheel. I dare you.

Do you know of any good alternatives? Any way to write MapReduces in python?

I wonder if Github is working on something like this.

I'm sure they're familiar with the idea and they seem like the best team to bring it online. They're still a startup looking for growth opportunities: this one seems obvious.

It seems no one commenting here is actually from Palo Alto.