Hasnain says:

I know this is a Reddit thread, but... this is basically an English translation of something that gave me the "right in the feels" sucker punch more than most things I've read in the last few years.

It's a real life story of a boy with muscular dystrophy; whose parents chided him for his habits and seeming addiction to gaming, only to learn - after his death - that he had a close knit group of friends online, who were basically like family - and came from around the world to attend the funeral.

Hit me right in the feels. People do a terrible job of understanding, nay, even going towards dismissing that which doesn't feel "normal" or out of their own experience; and it often comes down to disproportionately affecting those who are already suffering and need an escape mechanism.

So, yeah - go and read this.

Hasnain says:

This was a very technical read, about NetHack - while I don't know the game, I appreciated learning more about it.

The authors built a bot to play NetHack online, on a public server, and gain a high score.

They did this in a pretty novel way: predicting the RNG, and then using that to sort of play out a picture perfect run.

The devil was in the details though, and they talk about the challenges of how to get enough data to determine the seed fast enough to be relevant. It also goes into some nice quirky implementation details.

Worth reading if you want to nerd out.

Hasnain says:

The authors(*) talk about how one large company takes a defense in depth approach to secure both its products and code.

It walks through how multiple layers are needed: From frameworks that eliminate classes of bugs, to analysis tools that catch hard-to-find flaws, to human attention and review.

*: Disclaimer - I contributed to this writeup.

---

On a more serious note though, I'm glad to be able to share more about the work we do on the Product Security team and the effort we invest in security - and, if this sounds interesting, we're hiring!

Hasnain says:

This was a very interesting read.

The authors first explain Starcraft and the complexities of the game, showing why it’s a meaningful research goal. They then go into quite some detail about how AI architectures are designed, including novel training strategies where each bot in the league is given different objectives.

It ends with some analysis of the games it played against real professionals. Worth a read!

Hasnain says:

This was an interesting read. The author talks about the lack of price (and general) transparency in the vision and broader medical industry, and unearths the monopoly which you may already have heard of.

It's pretty much summed up by the end quote: "Because this industry has been getting away with fleecing people for decades."

Hasnain says:

Bookmarking both this and the software rasterizer for future reading

Hasnain has not yet written a summary for this.
Posted on 2019-01-19T14:31:25+0000

Hasnain says:

I definitely judged this harshly from the title - and the first section piqued my interest but it still wasn't clear whether I should pursue it more.

The article talks about the decline of car manufacturing, and how self driving cars will revolutionize the world. The author takes it one step further though and talks about how we can rethink how living spaces work, imagining tall skyscrapers with adjustable rooms.

While I'm all for the futuristic viewpoint, I'm a little cautious because it seems like this might perpetuate inequality still - we still have the capacity to build skyscrapers as it stands, and there are other downsides to society/culture that I don't see being addressed by this.

Hasnain says:

I for one welcome our robot overlords

Hasnain says:

This was a really interesting read. The author interviews a researcher at the MIT media lab who studies the phenomena of cultural and collective memories.

The article talks about how there are two decay curves, one rapid and one smooth - and how they have varied over time as technology has innovated. There’s also some side discussion over trends, fads, and famous people throughout history.

Choice quote: “At the same time, scientists and the general scientific community have not been great at adapting their ideas to new mediums. Scientists are the first ones to bring down another scientist who tries to popularize content in a way that would not be traditional. So scientists are their own worst enemies in this battle. They have lagged behind in their ability to learn how to use these mediums. Sometimes they focus too much on the content without paying attention on how to adapt it to the medium that will best help it get out.”

Hasnain says:

In this article the author talks about personal experiences mentoring teenagers, and how the modern society (and technology) are contributing to an extremely high stress daily environment for these kids.

It starts with a bang, mentioning technological innovations (which I hadn't heard about) like apps sending push notifications about new assignments and past test results, so kids are unable to disconnect and are always worrying about school.

The article talks about the parallels between this and working lives for adults (specifically the current millennial generation); with an aside into the constant pressure to be set up for the best education, then the best career - while they are already hearing from millennials that there is no light at the end of the tunnel.

Choice quote: "We owe it to the kids in our country to at least diagnose their disease, which is a society that turns children into stressed, anxious, competitive, indebted consumers. We do this to prepare them for their grown-up lives in a society that turns all people into stressed, anxious, competitive, indebted consumers."

Hasnain says:

The speaker covers some of the novel material behind Noria, a fast, concurrent database which uses materialized views to handle performance.

The slides available here do not talk about the systems research (left for the paper); rather they go into Rust and an implementation of a neat little crate called `evmap` which uses a pair of hashmaps and some pointer swizzling to achieve fast, concurrent read/write access to a shared structure in memory.

While I've seen similar approaches before in code, this is the best explanation I've found of this technique and it stuck - all from just the slides.

Hasnain says:

This was a good perspective on engineering leadership and people management.

The author talks about a bunch of things, including the transition(s) from senior engineer -> tech lead -> manager; the pitfalls of being a tech lead + line manager; avoiding bit-rot as a pure manager, and making the transition from line manager to manager of managers (and beyond).

The novel perspective I've seen here that is rare from most other articles/sources is that it talks a lot about the downsides/negatives of being a manager, which are often left unsaid.

The article is not all negative, though - it gives a lot of useful advice on how to improve one's management skills.

Choice quotes below.

On becoming a manager:

"Hopefully you have already gathered that management is a career change, not a promotion, and you’re aware that nobody is very good at it when they first start."

On becoming a manager of managers:

"Sure, there are compensating rewards. Money, power, impact. But I’m pointing out the negatives because most people don’t stop to consider them when they start saying they want to try managing managers. As almost every manager says after becoming a manager.

As though the mere existence of a ladder compels us all to climb."

Hasnain says:

This was a refreshing take on monorepos. Most arguments for/against them keep on the technical side, talking about how the repo gets large and slow (bad!) or the benefits of being able to make codebase wide changes in an atomic setting (good!), and debate the usefulness/commonality of each.

This instead focuses on the organizational cost, and how monorepos force you to make certain decisions/costs explicit, like the cost of maintaining separate branches/forks of internal code, and of having to coordinate releases of older code.

This is critical for larger organizations where it's very easy to sweep these costs under a rug unless there is a forcing function.

Worth a read.

Hasnain says:

This was an interesting read. The author goes into problems AI is solving, bias issues with AI, and the case that we should scale back our investments in AI, and push a lot more resources towards thinking bigger: rethinking how societal structures are set up in the first place.

While I disagree with the premise (there are enough researchers and resources to do both!); the point in the middle about AI (and, if I stretch the argument, the Valley) mostly focusing on problems with a more singular viewpoint, divorced from problems faced by a large portion of the world, rings true to me.

Paraphrasing what a friend once said, Silicon Valley engineers create problems for themselves and then make startups to solve those problems, without ever venturing outside that bubble.

Hasnain says:

App Store monopolies are real and showing