yannyu

> Yes, but isn't a big driver of layer costs the cost of the machines to build those layers?

Let's say the time spent in lithography step is linear the way you're describing. Even with that, the deposition step beforehand is surface area independent and would be applied across the entire wafer, and takes just as long if not longer than the lithography.

Additionally, if you were going to build a fab ground up for some specific purpose, then you might optimize the fab for those specific devices as you lay out. But most of these companies are not doing that and are simply going through TSMC or a similar subcontractor. So you've got an additional question of how far TSMC will go to accommodate customers who only want to use half a wafer, and whether that's the kind of project they could profitably cater to.

The number of exposure steps would be unrelated to the (surface area) size of die/device that you're making. In fact, in semiconductor manufacturing you're typically trying to maximize the number of devices per wafer because it costs the same to manufacture 1 device with 10 layers vs 100 devices with 10 layers on the same wafer. This goes so far as to have companies or business units share wafers for prototyping runs so as to minimize cost per device (by maximizing output per wafer).

Also, etching, moving, etc is all done on the entire wafer at the same time generally, via masks and baths. It's less of a pencil/stylus process, and more of a t-shirt silk-screening process.

The cost driver for fabbing out wafers is the number of layers and the number of usable devices per wafer. Higher layer count increases cost and tends to decrease yield, and more robust designs with higher yields increase usable devices per wafer. If circles or other shapes could help with either of those, they would likely be used. Generally the end goal is to have the most usable devices per wafer, so they'll be packed as tightly as possible on the wafer so as to have the highest potential output.

> I assume they don't pattern the unused area, so the process should be quicker?

The primary driver of time and cost in the fabrication process is the number of layers for the wafers, not the surface area, since all wafers going through a given process are the same size. So you generally want to maximize the number of devices per wafer, because a large part of your costs will be calculated at the per-wafer level, not a per-device level.

Utilitarian analysis inevitably is coopted by those in power to frame "good" as what the powerful want.

For example, some say it's more important to keep people's stock-invested retirement accounts solvent by increasing stock prices rather than take corporate profits and pay people who can't even afford groceries or rent, let alone invest in a retirement account.

Sometimes you'll have commercial/digital reps that handle high volume, low revenue customers. The company may also have a self-service/credit card swipe product option, or a free tier, or a support-only option.

I get what you're getting at, but LLMs aren't thinking machines. They literally just rearrange and regurgitate text that they've been trained on or have contextualized. How would you propose building a general purpose LLM that accomplishes what you're saying? How do we build a machine that is able to divine scientific truth from human outputs?

Is it just me or was anybody else not sure what Obsidian was while reading this article?

The article doesn't explain it at all, so I'm left wondering if I'm incredibly out of the loop or whether this is a very niche thing that I haven't run into yet.

This seems like an overly simplistic take and reads like someone who hasn't worked closely with board members or been on the board of a company with a lot of money behind it.

First you had logs. Everyone uses logs because it's easy. Logs are great, but suddenly you're spending a crapton of time or money maintaining terabytes or petabytes of storage and ingest of logs. And even worse, in some cases for these logs, you don't actually care about 99% of the log line and simply want a single number, such as CPU utilization or the value of the shopping cart or latency.

So, someone says, "let's make something smaller and more portable than logs. We need to track numerical data over time more easily, so that we can see pretty charts of when these values are outside of where they should be." This ends up being metrics and a time-series database (TSDB), built to handle not arbitrary lines of text but instead meant to parse out metadata and append numerical data to existing time-series based on that metadata.

Between metrics and logs, you end up with a good idea of what's going on with your infrastructure, but logs are still too verbose to understand what's happening with your applications past a certain point. If you have an application crashing repeatedly, or if you've got applications running slowly, metrics and logs can't really help you there. So companies built out Application Performance Monitoring, meant to tap directly into the processes running on the box and spit out all sorts of interesting runtime metrics and events about not just the applications, but the specific methods and calls those applications are utilizing within their stack/code.

Initially, this works great if you're running these APM tools on a single box within monolithic stacks, but as the world moved toward Cloud Service Providers and containerized/ephemeral infrastructure, APM stopped being as effective. When a transaction starts to go through multiple machines and microservices, APM deployed on those boxes individually can't give you the context of how these disparate calls relate to a holistic transaction.

So someone says, "hey, what if we include transaction IDs in these service calls, so that we can post-hoc stitch together these individual transaction lines into a whole transaction, end-to-end?" Which is how you end up with the concept of spans and traces, taking what worked well with Application Performance Monitoring and generalizing that out into the modern microservices architectures that are more common today.

Then we might as well admit that "merit" is heavily influenced by starting conditions instead of pretending that everyone has "equal opportunity". How many times have I heard from people that the USA is about "equality of opportunity" and not "equality of outcome"? Legacy admissions is directly contrary to equality of opportunity, and undermines the idea of meritocracy in university admissions that people have been crowing about in anti-affirmative-action rhetoric.

It's funny that this used to be a killer feature in Opera before they fell off. Between that and mouse gestures, they were really ahead of their time.

Let's not pretend that's driven by anything except protectionism of an existing industry. They simply don't want almond milk anywhere near cow milk in the grocery store.

You're not wrong, but is this the way we want art to be created? Should arts be primarily pursued by those who have the means to defer or lose income, or those who have the reputation and connections to nearly guarantee economic success?

The study may be good, but the headline and the obvious slant of the article are as unscientific as you can get. Pop science journalism is abysmal, and this is a good example of it.

A related short story by Ted Chiang speculating on a world without attractiveness:

"Liking What You See: A Documentary" by Ted Chiang - https://waldyrious.neocities.org/ted_chiang/liking-what-you-...

I studied electrical engineering, but power engineering isn't a course that I took. The most enlightening quote I've heard about this was a response to a question:

"What happens when I turn on an electric device in my house?"

"A turbine in a power plant spins more slowly for just a moment."

Is there any tech-oriented forum that isn't largely cynical and negative?

Is there any social media that isn't largely cynical and negative?

I also notice how my mindset changes when I don't visit, which really just makes me think I should stop engaging in 90% of online discourse and discussion.

Ted Chiang explores this idea in a short story called "The Truth of Fact, The Truth of Feeling" (https://devonzuegel.com/post/the-truth-of-fact-the-truth-of-...), which takes place in a world where commercial, individual, always-on recording exists. Ted Chiang also wrote the short story that the movie Arrival was based on.

This is an interesting hypothesis. The silver fox domestication experiment shows that selecting for behavioral traits can result in significant changes to physical traits as well: https://evolution-outreach.biomedcentral.com/articles/10.118...

The market is not fair. People have to eat, and will look for whatever job that can get that will help them eat and provide for their families.

As seen with the Daryl Morey, the Houston Rockets, and the NBA, the Chinese government will go after anyone they can when incidents like this come up. Morey's single tweet in support of HK resulted in China cutting off their relationship with the Houston Rockets and also calling for Morey to be disciplined or punished by the NBA. In this case, the NBA hasn't done anything specific to Morey. However, many people are expecting some sort of retaliation from China against the NBA in the future.

Correct. Elastic competes more with Splunk or Sumologic than it does with Algolia or open-source Lucene/Solr. It's nominally a traditional ecommerce/enterprise search engine as well, but it's not terribly good at it.

I like the idea of someone tearing down Gmail's confidential email feature, but this article isn't terribly thoughtful. It mostly just repeats the same 2 or 3 points over and over (confidential mode isn't, end-to-end encryption is important, our product does end-to-end encryption). It feels like it was rushed out to get ahead of others who might write the same thing, rather than trying to educate people about what these terms mean to them and why they should be important.

So your experience of "East Asian national parks" is just China?

>and support searches for unavailable (out-of-catalog) entities

This is a subtle killer feature of Netflix's search, and something not often done by others. By including these unavailable entities, it allows Netflix's similarity and recommendation engines to still understand what the user was looking for, and provide things that are relevant and similar, even if they aren't exactly what the user is looking for.

That being said, this is a lot easier for something that has a relatively finite, known set of items (movies) in contrast to things with lots of items that can be differentiated in smaller, subtler ways (e-commerce, web documents).

>By the way, I also find it really interesting that video games are software, a special kind of software that we assign a very particular type of value to that's not really monetary.

Video games may be software, but they're consumed in a way that's much more similar to music or video. Video games aren't a mean to an end like other kinds of software, but instead is an experience to be enjoyed simply through the consumption of it.

Yes you're right, I misspoke, 3000 a day worldwide.

Agreed that we're doing pretty well as far as deaths and injuries worldwide, but there's a lot of room to improve.

Or maybe we hold driverless cars up to an unfair standard. Today, 3000 people die in the US every single day because of car accidents. Many, many more are injured or disabled.

Why would we expect driverless cars to reduce that to 0? Shouldn't it be good enough if they reduce it to 50%? 25%?

EM waves are spherical in nature (analogous to ripples in a pond) rather than directional. So the rotation of the earth wouldn't significantly affect the strength/reception of the signal by a distant planet. The bigger factors would be signal degradation because distance and other kinds of interference.