pacala

That's because sooner or later every species enters an equilibrium state with the amount of energy available in the ecosystem. In the process it may out-compete quite a bit of other species, and so we speak of 'invasive species'.

One reason for 'even'. Silicon Valley contains some of the richest zip codes, richest corporations and highest paying jobs in the country. Yet all this monetary wealth doesn't seem to help with a major humanitarian crisis unfolding on its doorstep.

Researchers say there are three possible explanations for the anomalous data. One is mundane, 99.98%. Two would revolutionize physics, 0.01% and 0.01%.

I don't understand. There are dozens, if not hundreds, or even thousands of bits to use to identify any given browser, but providing one extra such bit that politely asks 'Do Not Track' is now a problem because it makes tracking slightly easier?

It's not a black/white situation. It's a 'argue in front of a judge' situation.

With sign: Unassailable evidence that tresspassers are informed they are not welcome.

Without sign. Owner word vs tresspasser word.

The latter is a weaker stance.

Great point. There are 'No Soliciting' signs and laws.

https://banneradviser.com/no-soliciting-signs

> In a nutshell, no, door to door solicitation isn’t illegal. But if you have a no soliciting sign posted on your property, and the salesperson is refusing to vacate the property, they can be assessed trespassing fines and possible legal charges.

If everyone has a lock on their door, then people will just ignore it and break into the house anyways.

1 extra bit is the last of my concerns, there's plenty of bits to uniquely fingerprint a browser anyways. I'd gladly trade one inconsequential bit, which requires malicious intent to misuse, to keep my privacy safe when dealing with honorable entities like, I presume, Google.

https://blog.mozilla.org/internetcitizen/2018/07/26/this-is-...

What goes into one's fingerprint:

1. navigator.userAgent, 2. navigator.language, 3. navigator.doNotTrack, 4. screen.width, 5. screen.height, 6. screen.colorDepth, 7. Intl.DateTimeFormat().resolvedOptions().timeZone, 8. navigator.platform, 9. navigator.hardwareConcurrency, 10. GPU vendor and renderer, 11. isTouch, 12. storage types, 13. font-list, 14. canvas-hash

Does Google Chrome set DoNotTrack by default in incognito mode? If not, why not?

Having used both Python and Typescript recently, I largely prefer the consistency of Typescript. [Hey mom, I like cherries better than strawberries]. Having said that, I'd love to have named function arguments in Typescript. Props workaround is, well, a workaround.

One of the greatest boons of static typing is formal documentation of data structures and interfaces. Catching errors early is a bonus :)

Typescript greatest features are optional types and the extensive type inference. On small projects this can work well, but on larger ones I can see overly rigid style guides requiring typing everything explicitly in triplicate becoming cumbersome in an Enterprisey way. Curious what people's experiences are.

The title right now: 'ZFS Removes References to Slavery'. Probably 95% of the commenters have not read TFA in sufficient detail to argue minutiae details. They see 'language changed in support of the dominant political current' and link it to Newspeak because that's the only frame of reference most people are aware of. It's not like it's an isolated incident, there is plenty of language redefinition, history rewriting and statues toppling as we speak.

I'm somewhat confused about the TV media landscape, as there are cable news channels, Fox / CNN / MSNBC, but there are also the big 3 networks, ABC / CBS / NBC, which also air news and are also available on cable.

https://en.wikipedia.org/wiki/United_States_cable_news

https://en.wikipedia.org/wiki/Big_Three_television_networks

Seattle has 600,000 people. In a given year there are 30 murders and 300 rapes. One can collect 175,200,000 amateur camera footage of 'an hour in the life of the average Seattle-ite' until one encounters an hour with a murder, or 17,520,0000 such peaceful one hour videos until one captures a rape on camera.

Does the presence of overwhelmingly peaceful footage of average people going by their average days make Seattle at large an Utopia with no crime? Is that footage in itself sufficient evidence to abolish the police and the court system, because, look, there are 17,520,000 hours of peaceful footage before something terrible happens to someone?

It is incredibly difficult to build an accurate image of a large scale group of people judging by a few hours of direct experience. 1000 harder if through footage selected by people with their own agendas. Media coverage, especially audio-visual coverage, is wild because media coverage is simply an inappropriate way to depict such phenomena.

That is almost right. Society has decided that, on average [or median if you prefer], every job is worthless. Income inequality is skyrocketing between automation, globalization and capital gains. The only jobs that are not worthless are top 10% jobs that increase the inequality gradient, mainly finance or automation.

Wow. Building and operating a massive unaccountable automated system to deboost and/or silence undesirable speech is now OK because 'it was a mistake'.

Edit. At what point are the users and/or the public at large owed a detailed postmortem for those 'mistakes'?

There can be more than one truth valid at any given point:

* The wealth was not taken from those children. At the limit, those children can just fuck off and die in the street of starvation, it's not like they have too many useful skills to offer to society anyways.

* We have an ingrained tribal expectation that the tribe will take care of all its members to some degree, within the available resources. Driven by the game theoretical proposition that tribes that fail to do so will find their human resources and social cohesion depleted and be easily torn apart by rival tribes. By now integrated in our genetic heritage.

One only need look into the streets [or at Twitter] to see that we are incapable of having any nuanced conversation anymore. It's either 100% A or 100% B. If you dare explore a balance, both sides will rip you apart for lack of ideological purity.

> The low skilled workers are going to get paid very little because we choose to pay them very little.

I wish it were that simple. More reasons:

* Low skilled workers are paid very little because there don't have any higher paid options.

* Low skilled workers are paid very little because there are some many other low skilled workers willing to take their place, domestically, internationally, and through immigration.

* If there are 2 business using low skill labor, one paying low wages and another high wages, the latter goes bankrupt.

Cherry on top: the AWS project is to turn SWE cushy jobs into 'low skill' occupations. Bezos understands economics. He's building the ultimately efficient money making machine, which, by necessity, reduces the human expenses to the absolute minimum.

Another interesting fragment:

> Q: Are the interactions that are happening right now between police and protesters something that you think is predictable? Or is this something new that we haven't seen before?

> A: It's not completely new; it's just the intensity of it compared [with], let's say, five years ago during the Eric Garner and the Mike Brown protests. What we're seeing is really an immediate escalation to very high levels of force, a high degree of confrontation.

> And I think part of it is driven by deep frustration within policing, which is that police feel under assault, and they have no answer. They trotted out all the possible solutions: police-community dialogue sessions, implicit bias training, community policing, body cameras. And it just didn't work. It didn't make any difference. And so they ran out of excuses.

> So the protests today are a much more kind of existential threat to the police. And the police are overreacting as a result.

These statements are not mutually exclusive.

This is the reason why professionals don't engage in politics. In the short term it lends cover to the political actors of the day, and in the long term it destroys the credibility foundation of their own profession.

Sadly, some of us are aware of the mess found in lower impact journals. It doesn't help with the trust issue. At all.

It is an abdication of responsibility from Lancet / NEJM editors. Everyone not living under a rock is well aware hydroxychloroquine has, unfortunately, become a political hot potato. In this context, a paper purporting results counter to decades of clinical practice should warrant a modicum of additional scrutiny.

This incident comes in the middle of the replication crisis, with 50%-70% of studies failing replication, including in the medical sciences field. Furthermore, many suspect that BigPharma pushes inconclusive garbage as peer reviewed science to promote commercial interests. Then there is the response to covid19, which, justified or not, some regard as catastrophic overreaction. The BigScience enterprise is in peril of losing public trust. Which is a terrible loss.

Lancet / NEJM editors failed. There are no excuses. The right course of action is to apologize, not to further antagonize the public. Try humility for a while.

If they are anywhere nearly as sensitive to 'bayesian priors', aka parameters made up of full cloth, as the covid19 models, then I take a hard pass.

Edit: Perhaps that was a bit too much of an 'in jest' reply. I'm frustrated at the whole mathematical rituals schtick, when in reality the whole thing is at best an unfalsifiable qualitative guess. The parameter space is immense, and we only have 1 measurable trajectory through it for validation.

We already know quite a bit about the world. Unclear why should we reach out into the fog of unknown for some cutting edge unproven state-of-the-art system, when there are quite a bit of well established and widely agreed upon known ones that can get us 90% of the way there, with much lower risks of catastrophic failure.

Case in point: covid19 policies inadvertently triggering the mega-depression of 2020 and beyond, for what appears to be relatively minor medical gains in the big picture. How bad can you botch it if it takes 5 months to come up with 'masks are good, perhaps, more research is needed' from a peer reviewed source, when even a 14 year old could have been told you that's rather a good idea all the way back in January.

https://www.thelancet.com/journals/lancet/article/PIIS0140-6...

Corollary: Climate science models should be built by engineers, not by scientists. Task people who 'use what is known and understood to construct systems that can be validated'. Perhaps the whole scientist-as-faux-priest meme should be toned down a notch.

This is pretty mind-boggling, to the point of suspecting an error in the methodology. If it is just a language model, then it has no baked in notion of test time tasks. How on earth does a language model know what is required of it at test-time without fine tuning? How does it know that the test time prompt are examples of the task, and not some story prompt it's supposed to riff off in random ways?

"closure is a poor mans object and an object is a poor mans closure" is true in principle. Devil's in the details. If the language forces the declaration of the object equivalent of a closure in a different scope, or even a different file, then massaging captured variables by hand, it makes objects a whole lot more laborious to use, especially for those using rapid iteration to get a feel of the design space.