blueblob

Yeah, how could you hate on anyone who uses the standard text editor?[0]

[0] https://www.gnu.org/fun/jokes/ed-msg.html

It's either a lot less common than you're making it out to be or it should not have made international news. Anyone who followed this story from the beginning realized it was not a good deal for Wisconsinites.

I find Zelle super frustrating. I needed to transfer money from a local credit union to Citizen's bank because there isn't a Citizen's near me. Both support Zelle, except that you can only have one account set up in Zelle. So I literally couldn't even transfer money to my other account through Zelle. Instead, I have to mobile deposit a check to myself.

It's not in line with what people want in the app though. It's like using a text editor that all of the sudden implements a wiki on gardening preinstalled in the app. It's not relevant to the features you want in a chat app and there's no reason it couldn't be a separate app.

I agree with the command discoverability. Not only does prefix+? show you the commands, it is also searchable, and if you use vi-mode, it uses the same search commands.

Just because it's ubiquitous doesn't mean that it's good. Also, to be clear, x86 then became x86_64/amd64 which isn't the same architecture either. There will always be iterations and oddball architectures where something new can be learned and even reapplied to update x86. POWER, Sparc, etc. all taught new lessons.

Apple isn't scattering anything by running processors that run the same architecture as android and ios on phones. Most open source software can already be compiled for x86, arm, sparc, etc.

Boolean interactions are piecewise functions as well. He seems to be arguing to use functions for modeling and considering the whole process or interactions between the functions but has not described an actual approach to doing so in this article. How is picking a set of booleans really all that functionally different from from drawing a surface over the intersection of all of his functions and just treating them as booleans on each surface of that hypercube?

I don't find it particularly groundbreaking to use a set of ad-hoc functions as interactions. He seems so focused on the idea of functions and complexity that he's missing much of the point of why people play games and solely focused on strategy games. In my opinion, a lot of the best games are the simplest games because I play them to take my mind off of other things. It's not about having a lot of complexity, it's about having interesting complexity.

This was written with such self-aggrandizement that it is not super enjoyable to read. It is easy to see why the ideas of someone who projects themselves as a genius is having trouble relating to people who may not believe him to be one. The usefulness of ideas seems to have some respect to how well they fit into the way other people want to use them. He's not focusing on how other people would use them but rather on how he wants them to be used. He seems to have missed the transition from systems-oriented design to human-oriented design.

Fair enough

In my mind this comparison isn't exactly the right way of looking at it. Most of the early buyers who held on bought full shares of bitcoin before fractional shares were an option. The people buying it at 48,000 are likely buying fractional shares. So they're not necessarily invested at the full risk of losing such a significant amount of money. I think this is part of the reason it's going up.

To be clear, you think that going out of your way to go on a self-righteous tirade against a guy because he's proud of his kid graduating college is on par with racism? You're being intentionally obtuse and it's readily apparent by the fact that you have to attack me and "my understanding." Your critiques ignore all of the arguments that I made about the substance of the article in favor of attacks. You complain that we can't have discussions and yet you're the person who makes it so we can't have discussions.

http://www.paulgraham.com/disagree.html

> It falsely signals to employers that you’re better than the average Jane just because you have a piece of paper

This is a reductionist argument stating that everything you learn in college is just a piece of paper. It's a bad faith argument. If you believe it, you weren't convinced and if you don't believe it, you still weren't convinced. The whole intent was to be polarizing.

> I preceded to tell him that it doesn’t matter if you have a college degree anymore to do today’s jobs and that he was basically bragging about his family’s socioeconomic status, not his son’s actual achievements. I tried to elaborate about the inequalities of higher education and why it’s an outdated metric of future success. I explained why employers are absolutely wrong for thinking his son is better than someone without a degree or a degree from a public university.

This is a clear example of hypocrisy and an opinion that is regurgitated over and over in the article. The man being discussed never asserted that his son was better than everyone else, that was the author's excessive bias. Believe it or not, many fathers are proud of their childrens' achievements. The whole argument being peddled is that nothing of value was gained and that this man needed to be put in his place by the author because she's better than his son.

The "data" that the author provided was mostly correlation (not causation) with wealth and people doing well in college. That doesn't imply in any way that people that have strong support systems but not wealth cannot do well. The argument is fallacious.

The tone that the author set for the article in the very beginning is that people are supposed to be enraged, like the author is, that employers want some indication that the candidate is a good quality candidate. In the current system, that often means getting a college degree. Employers want the closest thing they can get to a guarantee that the candidate knows something. Without proposing an alternative this was just a poorly supported rant by the author. Your devil's advocate that I must stimulate discussion because the author did not is again, fallacious.

>> anecdotal evidence misapplied as representative

> Such as? Did you happen to skip over ALL the other sources they used to prove their point?

Such as using specific schools when talking about corruption but then using all schools in the later points about performance. Presumably, people with the prestigious degrees would have better job performance but no stratified data was used in that discussion. To me, it appears as though the author had a point they wanted to make and did the bare minimum research to make that point without questioning whether it was right.

With this in mind, I don't find that the article had all that much substance other than cherry-picked statistics and made no effort to to have any nuance. The takeaway was supposed to be that all people with college degrees got no value because the author had no nuance.

>You're taking an article with enough substance to have a valuable discussion around it and poo-pooing it away because the author is inferior relative to your subjective standards. That's what I'm getting from your comment.

You're asserting that I am assessing that the author is inferior because she doesn't have a college degree? I actually don't know whether the author has a college degree, but don't find the points particularly convincing either way.

I am stating that I don't believe the article to have much valuable substance beyond the clickbait title. You're asking me to invent substance for the article in the comments because of your opinion that the article hits some critical point.

The argument wasn't strong in any way, in my opinion. The article itself limited the discussion by taking a demeanor that is effectively an attack on all college graduates because of anecdotal evidence misapplied as representative. There was no discussion as to what alternatives are to a college degree that employers might find attractive. The writing highlighted the insecurity and hypocrisy of the author. Mostly, the article was a waste of time to read.

The author comes across as someone who didn't try applying for college and hates everyone that did. The hypocrisy of the entitlement displayed in her first narrative definitely weakens the argument from the get-go.

None of the points are particularly convincing, mostly because she picks the base for each of the statistics as whatever best supports her argument. She starts out only talking about elite colleges because that's where she can find corruption and proceeds to pretend that it reflects the entire degree-granting community. "Case studies" 2 and 3 are both the same argument.

The last point about degrees being an indicator of on-the-job success is actually saying that 2/3 of people with degrees were considered successful by their bosses. Whether 2/3 is good or bad depends upon how it compares to employees without degrees, a statistic conveniently left out.

A degree isn't supposed to be an indicator of on-the-job success; it's supposed to be an indicator of ability to learn, adapt, and manage your time when you are on your own.

He was appointed by the Obama/Biden administration, so I'm quite ambivalent on the question. I feel like almost everyone in tech wants him to go but I may be just projecting my own bias.

Second paragraph... glad we can all count.

In this false equivalence you're trying to make, did the people in the women's march come with zip ties to kidnap congressmen? Did they come to interfere with the electoral process? Did they steal? Were they violent?

You bring up the Democratic Party as if everyone who disagrees with you must always be a part of the exact opposition and you come across as a conspiracy theorist because of it. It really doesn't seem to be the world's most unsurprising thing that among two possible conclusions people could be on opposite sides without being brainwashed. It's almost like the binomial distribution exists for something like this.

It feels like the more aggressive the estimate is for the upper bound the less accurate it sounds, whether that's true or not. Somehow it just feels more believable that it's 2x instead of 10x without considering evidence either way. I think epidemiology is a tough field because you never actually have any verifiable ground truth.

If I understood correctly, CentOS Stream is a beta or a nightly for RedHat now?

So you're saying everyone should always be staring at their webcam to make sure the light is not on? It isn't like there's a subfield of Human Factors studying Cognitive load[1], saying that people don't do a good job at what you're saying people should do. Also, have you ever heard of a rootkit?

[1] https://en.wikipedia.org/wiki/Cognitive_load

Hasn't FreeBSD 12 been out for a while now? This is just a LTS release?

Agreed. Shopper buys yellow cilantro with brown spots and bag is full of water. I plan a meal around protein and shopper gets everything except the protein. Shopper substitutes regular soda for diet soda. Everything at the store is not listed on the site.

"toronto weather" gives me the weather now and the 7 day forecast in a widget on the top of the search page. If I search "packers" it gives me the latest packers scores on the top (and if a game is ongoing an indicator of the time, who has the ball, and the score), if I search "patriots", there is an nfl tab on the top that I can click and it gives me the scores. I'm not saying you're wrong, but DDG has been working on some of those issues. You can always add !imdb or !rt in front of your shows/movies for rottentomatoes or imdb.

Agreed. Just because you don't compare your model to another model doesn't mean you're hiding something and it's irresponsible to pretend that's the case. Perhaps the author should have run char-cnn against data instead of making baseless accusations.

On the whole, a lot of the arguments are realistic of ML in general. There is far too much focus on individual problems and the accuracy measure without enough focus on the interpretation of the accuracy, how well it will really generalize, etc.

At $20/mo in DigitalOcean services that are only required if you're running on a chromebook, why not just run on an inexpensive PC? It's seemingly long-term more expensive to do it this way.

This is an area that fades in and out of focus with such venues as the Workshop on Human Interpretability in Machine Learning (WHI) [1]. It's becoming increasingly important when it comes to auditability and understanding of what is actually learned by algorithm. Avoiding classifiers from learning to discriminate based on age, race, etc [2] or in domains where it's important to know what the algorithm is doing such as medicine. Work in understanding DL is not really interpretable in any domain, typically they train another (simpler, less accurate) model and use that to explain what the model is doing or use perturbation analysis to try to tease out what it is learning. If all you care about is getting the right answer and not why you get that answer maybe it doesn't matter.

I wouldn't say I'm hating on DL nor that I hate on random forests, or ensembles, etc., but when you have very little data fitting an uninterpretable, high dimensional model might not be the right answer, in my opinion, see [3].

[1] https://arxiv.org/html/1607.02531v2 [2] https://arxiv.org/abs/1606.08813 [3] https://arxiv.org/abs/1601.04650

It doesn't matter that it's on the holdout, he's partitioning an already small dataset into 5 partitions and talking about the accuracy in using 80 points to predict 20 points. The whole argument is usually that in the law of large numbers you can now have a statistically significant difference in accuracy. When you're predicting 20 points each with 5 (potentially different) models you likely don't have enough to talk about statistical significance.

Not sure what kind of argument that is. If something overfits it will have less error, does that make it better? It may mean it would generalize a lot less when run on more data. Whether or not something is meaningful depends on what you take the meaning to be.

Define "meaningfully better." Perhaps you mean statistically significantly better? It may have better accuracy, but it has significantly less interpretability. What does it capture that regression couldn't capture? At least with regression you can interpret the relationship between all of the variables and their relative importance by looking at the coefficients of the regression. With deep learning, the best approaches for explanation are to train another model at the same time that you use for explanation. Additionally, it was proven that a perceptron can learn any function, so in some senses the "deep" part of deep learning is because people are being lazy because at least you could get a better interpretation of the perceptron. I don't mean to imply that there's not a place for deep learning, but I think this isn't a great refutation of the argument that fitting a deep model is somewhat inappropriate for a small dataset.

Open Sound System