88e282102ae2e5b

Does scikit-image have the algorithms you need? If so, I'd highly recommend it for image processing - the API and documentation are well done.

I think it might be better to say that contemporary psychology is frequently not being performed scientifically.

This is exactly like every other linear algebra book I've ever seen, other than the title. I was hoping this had been improved over v1 but it seems to be unchanged in any significant way. What exactly does the author think the bullshit is that they're leaving out?

Well, it's not really a problem now. Spring came a month early, but who cares? Islands aren't underwater yet, there's plenty of food in first world countries (who would be paying for this) and there's no serious drought. From a politician's point of view, you're proposing to spend billions or trillions to solve a problem that we're still only seeing hints of (with an extremely first-world centric view, of course).

If the technology works, why not wait until the people demand it, instead of telling them they have to feel the pain of paying for it now to solve a problem no one can see?

I went back to school for a biochemistry PhD after a few years as a web developer - no regrets, though I do miss not being broke. Computers are still vitally important to my career though and HN serves as a convenient aggregator of programming techniques and new technology that I occasionally find useful. Also it's nice to be able to respond with a smidgen of insight to the odd biology article that finds its way here.

> Every time I see an article like this I wonder why it hasn't already happened

Evolution, unfortunately, ruins all the fun. Producing all those proteins to make your drug of choice puts a large burden on the cells - any of them that lose those genes to mutations in the DNA will be able to grow slightly faster and eventually take over the entire culture.

Honestly, an introductory biology textbook is your best bet. But other response is correct: we can't even simulate a single protein well, much less an entire organism.

Sure, but I'm less worried about absolute rank and more interested in whether the language I'm using is on the order of C/C++/Fortran.

I can sell a 2x slowdown to my boss if I can show that that's the only cost of a significantly more elegant and productive language, but at 100x that's a much harder sell.

What DO you do if someone mails you heroin? Is the only reason he's not in jail the fact that he snuck into the forum and uncovered the plot before it happened?

You're very much correct, but I'm almost glad I got hoodwinked. I've debated whether I'm handicapping myself by sticking with Python and Rust, but reading through this list I was just aghast at how many problems I didn't realize I was avoiding.

This reminds me a lot of some of the first home movies that were made, where people would just stand completely still in front of the camera, as they didn't really understand what the new device was capable of.

That's true, but my argument is that we can basically guarantee that a simple neural net won't become self-aware, in the same way that we know that a linear regression library won't become self-aware. The complexity is bounded, just like it is in the experiments described in this paper.

Well, the actual names of the base pairs they used are dNaM, d5SICS and dTPT3. It doesn't seem to me there are any good choices.

It can replicate them, but the XY pairs are just in a plasmid and not in the genome of the bacteria. They were chosen because natural polymerases could use them.

However, there are no corresponding RNA unnatural bases, and there's no tRNA that would recognize that even if there were. So it's going to take a lot more work before you can use these for coding for noncanonical amino acids.

This really is just a proof of concept.

The paper specifically states that these unnatural base pairs were chosen because they can be handled by natural polymerases. Also, the only way to maintain a plasmid over time is to replicate it. In fact, the whole logic of the strategy is to have Cas9 target mutations to the plasmid so that loss of the XY base pair during replication would lead to loss of the plasmid, and thus loss of the antibiotic resistance gene that it also carried.

Definitely interesting work, and some very clever solutions to difficult problems, but to be clear, they only had a single unnatural base pair, and it was only in a plasmid.

Addressing the other comments about this leading to some crazy scifi scenario, imagine telling someone you used machine learning to identify spam emails and they started freaking out about how you're going to start Skynet. If you understand how things really work you can see it's not just implausible, it's that things fundamentally don't work that way.

It's like I accidentally stuck my hand in a running blender and someone asks whether I'm angry because of the noise of the motor or because the buttons aren't arranged in a reasonable order.

I tried learning about Haskell just so I could understand a flurry of articles I came across that were trying to explain monads. But mostly I found super abstract notation, something about bind and unit and >>=, and I came away just feeling dumb. I gave up pretty quickly, but to be fair I suppose I wasn't super motivated to work through it.

Later, I learned Rust just for fun. Options? Sounds like a good way to express uncertainty about a function call. Oh, and you can compose it with other Options using this function called and_then(), and things will short-circuit if they fail. Neat!

A year later I find out I had been productively using monads the whole time without even knowing it. It's so frustrating knowing how easy it could have been to learn this concept explicitly the first time around.

Once I had learned the syntax, it took me a few days while working on a real project with well-defined goals to internalize the borrow-checker and such. Clippy has also been a surprisingly useful tool for learning - it pointed out some things that weren't idiomatic and that also solidified some concepts for me.

Xubuntu - beautiful, clean and simple, and none of the bloat and weirdness of vanilla Ubuntu.

Comforting shareholders is one potential reason.

It might just be as simple as $200 being a lot of money to the attacker.

Ah yes, I should have stated that the object is immutable.

I think a problem with this is that when you bring an actual database into the test, it ceases to be a unit test and becomes an integration test (not that integration tests aren't useful). I would prefer for the application to be separated enough from the database that I can test the business logic entirely on its own.

If I have some object I want to test in multiple ways, I find it more convenient to create a single instance during setup and then testing various aspects of just that single instance. Since the constructor of the object in question needs a bunch of values passed in, and each unit test only checks the effects of one at a time, I'm just going to copy and paste a bunch of object initialization code if I don't have a setup step first.

Yes, though I break it up into different (labelled) sections, so I don't have a large number of files. It may just work for me since I tend to write longish overviews of things (papers, concepts, experiments, etc.).

There is at least some value in filtering low-quality submissions, wrangling reviewers, and making editorial decisions when reviewers disagree or are just being assholes (e.g. attempting to hinder a competitor).

I use Jupyter notebooks and write in markdown.

For the first idea, you'd have to have some lethal protein or something that's specifically inhibited by the chemical of interest. Coming up with that might be even harder than the original problem. And even if you could, evolution wouldn't favor cells that produced it, since it confers no advantage. There's an energetic cost to making that protein.

For the second idea, again, evolution ruins everything. The competing organism would do just that - compete.

This looks like every other linear algebra book I've ever seen. Other than the provocative title I see nothing that distinguishes it.

Not mentioned in this article is Cpf1, another CRISPR enzyme which is unambiguously owned by the Broad Institute and may have less off-target activity compared to Cas9. It could easily be that this entire disagreement is over nothing.