lootsauce

Seems like a lame argument, they are the CDC why can't they simply ask the labs to report back the findings its a global pandemic for goodness sake.

Here is a great collection of maps you might want to add.

https://www.davidrumsey.com/

License there is CC BY-NC-SA 3.0

Ah here is prior discussion on the Met release https://news.ycombinator.com/item?id=7781846

Is there a comprehensive list of all such large collections that have been made available. I recall the Metropolitan Museum of Art released hundreds of thousands of images a while back.

I am one of those poor saps that would use this.

I totally agree with this approach, however it seems like a good idea to use alternatives for future work.

not stirring anything up, its the first headline on the readme. https://github.com/request/request#deprecated

Not identical but I am considering got because of nice docs and solid streaming support https://github.com/sindresorhus/got#comparison

I am now looking for another library that is as expressive with http streaming. I have enjoyed using request for complex things things like piping file downloads while rewriting headers, adding cookies, without buffering the entire response in memory.

I hope the community keeps this project in maintenance mode for a good long time. I have had a hard time finding an http streaming Swiss Army knife like request.

This looks really interesting. I have been digging deep into sync for the past few months and am very interested in how this works. I am currently trying to understand automerge and delta patching trying to implement from scratch. I have looked at Gun and couch db + pouch db but both did not feel like the right answer.

how does one buy this?

Our organization has embarked on a content marketing strategy that looks exactly like this and it seems to be paying off. Our sales pipeline is better than it has been in years. Our sales and marketing people brought in a content marketing company to help us and although it seems obvious and simple having the formula laid out clearly without a wall of text does not translate to low value in my mind.

You seem to be describing strategy writ-large for a company, this is about "content marketing" strategy.

What makes this a low quality article?

It's not about conveying the information, it's about how to get your brain to integrate and hang on to the information. The process of writing it down, and the other steps described are all about cajoling the brain to learn new material quickly.

Personally I think it takes more than a note taking process, there has to be some personal intrinsic motivation to understand and a genuine desire to know. Otherwise its just test cramming and will be forgotton soon after.

A few observations I generally do not see discussed.

Velocity of money has a large impact on the inflationary impact of circulating money.

Bank loans endogenously produce new money in the economy. This can dramatically impact the inflation seen relative to Fed printing.

Official macro stats, especially inflation are tweaked to death and are not a stable metric.

This is basically a very simplified picture of free will described by Aquinas https://www.iep.utm.edu/freewi-m/#H4

Our will power is like a muscle and can get tired. What if we have free will that is not actually "exercised" all that often. Say 85% or more of our actions are basically pre-ordained by decisions we have previously fully internalized and so are basically automatic, but which still alighn with our general will. Thats like an inexpensive (cached) result from the will banks. Then actual "exercise" of free will comes about when the patterns for these cache lookups fail and we must decide. But not only that, it is not exercised unless our desire and our will diverge. If I desire a second slice of pie, but my will is to not over-eat, it requires effort (will-power) to follow through with my will. That is the exercise of free will and does not actually happen as often as we might think.

We have found in database ML to be great to productionalize basic machine-learning methods. We use BigQuery ML and it saved us a ton of fussing with various frameworks, and because we were already in BQ it saved us the moving of data from database to ML cluster and back into database. We simply author some SQL and get great results quickly.

Agree. We have a data config later in our app that stores a manual sorting of attributes. This config is client specific and managed by our data engineers. This is the default sort and is available via this tri-state behavior.

For example, what ANN do you use in production, what are the reasons for using it. Another question is how much data does it take to make these models work, can they be primed effectively with little data if doing transfer learning from a more general model?

This is a fantastic post. Would love to see an even more detailed walk through some code examples, and discussion of development to production of these models. Do you have any other resources you would suggest?

Just read this today! Also this lib is from the maker of the amazing UMAP dimension reduction lib.

https://hdbscan.readthedocs.io/en/latest/performance_and_sca...

Totally agreed, there is a degree of hubris in geo-engineering that is breathtaking. There is of course geo-engineering going on right now, climate change itself is the product of unintended consequences, but to submit that all we need to do is have intentional geo-engineering as a solution is the worst kind of myopathy.

my impl if anyone cares https://github.com/andrewluetgers/flylsh

I was fascinated with this paper when it came out and reached out to the author to get access to the source. I subsequently re-implemented the python in JS for fun and to learn how it works. This is the best post I have found that covers in detail how the algorithm works.

https://medium.com/@jaiyamsharma/efficient-nearest-neighbors...

The really interesting thing from this research imho is that it is an algorithm derived from observed neural activity as opposed to most ANNs that are merely inspired by neural structure.

From Saket Navlakha's page https://snl.salk.edu/~navlakha/

"We work at the interface of theoretical computer science, machine learning, and systems biology. We primarily study "algorithms in nature", i.e., how collections of molecules, cells, and organisms process information and solve computational problems."

The whole range of neural nets feel like they came out of a process of random recipes that got thrown at the wall and we all celebrate the things that stick. It does not feel like this approach is going to lead to AGI. I am far more interested in the "algorithms in nature" kind of research where we will eventually (given the right tools) find some really interesting new advances in neural net based algorithms and potentially completely new architectures.

if that is the case they are not doing it right - from your link: "we make the modest proposal that website makers configure their main sites to be accessible by domain.com as well as www.domain.com."

Seems completely anti-user to not have a redirect. I wonder if it's just an oversight.

Got a good chuckle “that's a job for an unpaid reviewer, not an unpaid commentator.”

My frustration with the arguments for communism and or socialism vs capitalism is that they are surface arguments on ideological basis. I see socialism and communism as dressed up authoritarianism.

A massive hand-over of power to a few in government is not an improvement over capitalism where a much larger group vie for power in a dual reinforcing and opposing system of government and corporate interests. To hand the economy over to the government creates a far more centralized system of power and all the discussion of supposed benefits we could attain by such a system does not nullify the massive threat to freedom by that kind of system.

Capitalism is a horrible system but its the least horrible one we know of. A better discussion is how do we get to a better regulatory regime.