lexpar

The (relevant part of the) 5th amendment does not say "you never have to cooperate with the courts" it says you can not be compelled to incriminate yourself. People and companies are compelled to cooperate with courts all the time - that's literally what a subpoena is.

I thought it was clear in the post that Meta is being compelled by the federal government to prove they are not a monopoly and as part of this they need to present a case where they establish they have legitimate competitors. This requires information about the competitors.

The federal government is compelling action.

That sounds mostly right to me.

I think many constitutional protections do apply in case the USA is prosecuting an individual, eg even as a non American you could take the 5th if an American court was trying to convict you.

However, when you're asking to enter the country as a non US citizen your options are essentially to do whatever the border services agents ask you to do or turn around and go home.

Are you actually saying that Facebook is working to prevent _future generations of humanity_ from existing? I can't imagine how you expect people to take you seriously.

This is really not a high quality comment.

I think they announced that they're doing exactly that (or at least working on it).

I'm afraid that's just not true. I've made sourdough bread for years, and never added any sweeteners.

There are plenty of bread recipes that require no sweetener, and it's definitely not needed for yeast to develop.

Interesting comment in the header of this site.

https://github.com/prithugoswami/personal-website/blob/maste...

It's important to understand that this was already a very violent time in Montreal. Separatist terrorists were running rampant, and had a non-zero degree of public support.

I don't think she's a bad person, and I appreciate her response to the situation.

But building a year (years?) around a project, attending multiple conferences for it, bringing up a couple students on a research idea, then finding it's all a software bug? That's a nightmare.

What a nightmare. I bet this raises the hair on the back of the necks of a lot of other researchers. So much time and momentum can be invested into a research track like this.

I like this idea.

I don't know if I would say I'm 'hoping' for this, but I'm certainly interested to see what kind of incredible projects will come out of the increased necessity brought on by climate change.

It seems like for a long time there's been resistance to really try to do 'big' things in NA. We hardly even build subways since the 60's -- too complicated, too expensive, too disruptive, too big! Maybe we'll increase our ambitions as a society when we're less comfortable.

Yeah, I guess this is fine if by "Transformers are Graph Neural Networks" we mean Transformers < GNN, rather than Tranformers == GNN.

"Sentences are a fully connected graph". Ok fine, but that's a graph with basically no information embedded in its structure. GNNs are supposed to be useful for graphs that have interesting structure, right?

This is a great article.

I've been conscious of and annoyed by the symptoms of these spirals for years. I was aware that this was the effect of one-upping and a desire to demonstrate personal virtue, but I didn't have a name for this phenomenon or a very complete model of why people were interested in these types of communities.

The purity spiral! A nice name, and a model that seems to describe the situation very nicely.

By "taught this" do you mean kill a kid?

Coffee might move the needle, although I'm sure the big chains (Starbucks, McDonalds, Dunkin, Tims) would quickly move to some ersatz blend of roasted plants and lab synthesized caffeine. If they transition the formula slowly enough, lots of us may not even notice.

As an aside, I'm not 100% sure this isn't already happening with Tim Horton's.

Pretty sure the Python developers have stated no more backwards compatibility breaking changes...

Certainly with Python being one of the top languages in the wold today, the pressure to support backwards compatibility is much higher than it was back in 2007.

This may have a kernel of truth to it, but it's difficult to treat as objective when the article is essentially an advertisement for the language the author is developing.

You have a good citation for your first sentence, but not for your second.

There is an obvious equivalence between unary numbers and binary numbers. Should be quite easy to support inductive proofs with unary numbers while supporting computation with binary numbers, and simply translate between them depending on what context one is in. In fact, I believe this is what Coq does behind the scenes.

I took a graduate class where we worked though the first and second volumes of Software Foundations (https://softwarefoundations.cis.upenn.edu/). These are very nice introductions to the Coq environment and are being actively improved. The textbook is available online, with the source file of each chapter being a valid coq script, making it easy to experiment. There are also a large amount of interesting and challenging exercises.

Here's a link to the chapter on the Curry-Howard correspondence: https://softwarefoundations.cis.upenn.edu/lf-current/ProofOb...

It's not really a great analogy.

In machine learning the cost function is static and is defined by a smart person working hard to come up with something sensible. In life, the cost function changes constantly and is developed consciously as well as subconsciously. Indeed, the training examples (previous experiences) will influence the dynamically evolving cost function.

Moving along the gradient in a ML framework means that every refinement of the framework will influence cost non-positively, while in life we often act in ways that decrease utility.

Also, to put it bluntly, life is just a lot more complicated than that.

Thanks!

Does anyone have a nice resource they recommend on what BERT does? I've gathered it was trained by trying to predict missing words in a sentence, but I don't have an intuition on how this is useful for downstream prediction (like, say, learning a word embedding is).

Oh, I was not aware the US is not the only country on Earth.

Snark aside, here's a source that says globally coal power production accounts for 30% of global CO2 emissions. That's a lot, but not quite the whole problem.

https://www.iea.org/geco/emissions/

Actually I agree with you that getting upset over meat eating and plastic straws is a silly way to avoid tackling the much larger issues.

Also, I was genuinely asking for more information, and am still interested.

I'm interested in the justification for this. The following link says about 25% of US GHG emissions are from coal, but coal burning makes up the vast majority GHG emissions for electricity production.

https://www.eia.gov/energyexplained/index.php?page=environme...

Interesting stuff.

I remember reading about this tongue posture idea (keeping the tongue pressed against the roof of the mouth), but nearly all the discussions I found about it online were on incel and bodybuilding forums (supposedly maintaining this tongue posture increases the "sharpness" of the jaw and is more attractive), so I wrote it off as silliness.

This article seems to be making the case that there were on to something.

"The worst-case scenario of 6.5-7.0°C of warming assumed continued economic expansion driven by growth of fossil fuel production"

I assume this means "doing nothing, all future growth based on burning coal", which hopefully is a lot less likely than the headline makes it sound.

You must be getting much less frustrating team mates than me haha :)

Does anyone have a read on the rigor and ideological slant of the blog? I've flipped through a couple posts and find the content easy to follow (I have no economics background). However, I haven't found any meta commentary on the blog or information about the author (Joshua Konstantinos).