Harm is based on all risks not just the metrics that make what you care about look better. 60% of American adults drink regularly so extreme edge cases happen, but they aren’t anywhere close to the risks from car accidents etc.
Risk of death is significant, but we’re talking under 0.1% per year.
Some absurd number, like 70% of domestic violence, 25% of sexual assaults, 30% of regular assaults, around ~40% of murders, ~15% of robberies, and ~50% of injurious or fatal car accidents occur when people are under the influence of alcohol.
No, LSD won't kill your body. But what are you? You aren't just your body, you are your mind and mental state and the consistency of that.
Do enough LSD, especially consistent high doses, and the "you" you know will die and be permanently replaced by something else.
I of course agree that LSD won't put you in a coffin. But there are more axes upon which a drug's danger should be evaluated than "what percentage of users are put in coffins?"
In defense of LSD, your roommate did "a shit ton of LSD". Keeping dosage under ~300ug radically decreases the likelihood of any sort of permanent personality shift like that.
That said things can still happen on a bad trip. I posted my own personal experience downthread.
I don't understand how anyone who has ever taken LSD can say that it's safe. Maybe at low doses.
Once you can't seperate fantasy from reality you're just one bad trip from falling out of a window you thought was a door.
edit: The responses here just seem like denialism from people who got lucky to be honest.
You can be aware that you are tripping and still imagine that something is something else.
Roads looked like footpaths to me, windows looked like doors, knives looked like forks, and so on and so forth. I was aware that it wasn't real but couldn't be sure of the safety of existing in the world.
I basically had to sit in a corner for 10 hours, hope that the corner was actually the one I thought it was and that I wasn't going to fall down a staircase instead.
One rumor I've seen is that dealers have been replacing LSD with NBOM which has a significantly higher chance of causing a psychotic break and is harder to safely dose. This gets back to the big problem with illegal drugs isn't the drugs, it the illegal.
Personal anecdote: I used to do a decent amount of LSD (~150ug every 4-5 months for ~3 years), until one time I had an extremely bad trip. Ever since that bad trip, I've found that every time I get high (from LSD, mushrooms, or even marijuana) I start having extreme anxiety to the point where I can't enjoy it anymore and am just waiting for it to end. The bad trip was now 8 years ago and to this day I've not had a great experience on any of the above drugs since then.
I consider myself lucky that the anxiety only shows up when on drugs, which is pretty easy to avoid by, y'know, not doing drugs. But clearly something permanently shifted inside my brain.
The inability to experience something I used to enjoy without a near-panic-attack is a bit more of an effect than a conversation might have.
Memories once missing don’t grow back either. The most extreme cases of LSD use have impacted such basic functions as limiting people’s ability to communicate.
Post lobotomy people generally weren’t brain dead, even if diminished. So for extreme cases, yes it’s a solid comparison.
The problem of course is that so many things are experiential (rather than data driven) so the question “What would convince you otherwise?” could only be answered by “Having lived a different life.”
Take, for example, a software engineering opinion that I hold dearly: “Good commit titles and descriptions, with sufficient description of why a change was made and notes on any non-obvious implementation decisions, are valuable and necessary.” This arises from my experiences trying to debug code in repos filled with commits titles ‘wip’ and ‘address feedback’ and ‘changes’ which inexplicably touch 5k LoC. I simply despise trying to divine what some past developer intended without any clue as to why; it is an epistemological impossibility even when sometimes the past developer is me. More over I am convinced this is necessary because I also have the experience of doing all the work to understand some spaghetti—using old issue tickets and git blame to slowly build a mental model of the codebase—finding the bug, changing it, and having the whole system violently reject my fix because what appears to the naive observer to be a bug is, in fact, intended behavior the system depends on.
What would convince me otherwise that spending time on writing good commits is not worth the time and effort? Either an impossible-to-execute-without-confounding-variables longitudinal study that measures developer happiness over time in code bases that do/do not emphasize commit message quality. Or I could have different experiences that lead me to not care. Like if I never had to maintain legacy code or if everywhere I ever worked had issue trackers filled with explicit technical details and motivations then maybe I wouldn’t care; but that’s not my experience so I do care.
I don’t know how you hack your way past the reality that there is not one perfect way to write maintainable software and even if there was nobody actually has time for that. So you have to chose what to prioritize and some of those priorities may become your personal cargo cult.
That's a great example. My experience is precisely the opposite. I gave up on commit messages years ago in my own projects and it has never had the slightest negative effect. I literally have sequences of hundreds of commit messages that are simply "update".
But my experiences are not yours. my projects are likely simpler, and I don't collaborate much. It would be dumb for me to follow your practices, and you mine.
Exactly. What is important is knowing what to prioritize for the project at hand. If it’s useful it’s not a cargo cult. Of course the less experienced would ask “how do you know if it will be useful?” but truly the answer is you experiment.
> For example, the structure of EEOC Title VII inquisitions pressures companies to search extra hard to hire diverse candidates from a fixed size pipeline (the only way to make this work is to relax qualification requirements while pretending qualification requirements aren't relaxed, and maybe to fire any employees who fail to play along), but they can only go so far before they run up against the letter of the law which prohibits them from, say, openly firing people for their race.
This assumes worker pipelines are fixed which, practically, they aren't. You can increase flows to the pipeline by increasing outreach or incentivizing attainable qualifications. You can also seek trainees rather than day 1 contributors, which admittedly _is_ "relax[ing] qualification requirements" but has the trade-off of providing more purpose-fit employees when they do start contributing.
I think the pipelines are pretty inelastic when you consider that most companies are competing to hire "diverse" candidates from the pipeline and the limits of the sort of training you propose (training might work for candidates who are already pretty well educated and highly interested and motivated to serve in an entry-level position, but you're not likely to have broad success training high school graduates for senior level engineering positions). Consider how much time, money, and energy FAANG companies have spent for so little difference in their demographics.
Moreover, the point of my comment was how the EEOC's notion of discrimination is pretty much "whatever we think is discrimination", so companies have to compete to appear not to discriminate, which manifests as increasingly giving preferential treatment to minorities without actually running afoul of the law as interpreted by the courts. So even if tech companies adopt your suggestions this year, next year they'll be the standard and companies will have to find something else to stay compliant.
Yes, but you still need a module like typedload to do the runtime checking.
TypedDict performs no checking by itself at runtime.
class A(TypedDict):
a: int
A(d=32)
# Returns {'d': 32}
typedload.load({'d': 32}, A)
# TypedloadValueError: Value does not contain fields: {'a'} which are necessary for type A
The NY Times has been covering this lockdown for weeks now and I don’t think you get more mainstream than that. Here’s an article from Saturday for instance:
We use isort[0] for this. It even has a "black" compatible profile that line spits along black's defaults. Additionally we use autoflake[1] to remove unused import statements in place.
isort only sorts imports. ssort will sort all other statements within a module so that they go after any other statements they depend on. The two are complementary and I usually run both.
Thanks for your work on this. Coming back to python from golang, I really missed the auto-formatting. black plus something like ssort seems to bring the same to python. I've had really good results with black so far and look forward to trying ssort.
Not sure what would make using a Gmail account anticompetitive, but it's also not true. You need a chrome web store account (which they amusingly encourage you not to use your personal email for) which can use any email address.
That's what made it kind of an interesting read. All of the 'nicer Larry' projects in the article flopped, from Google+ to wearables to unsupervised learning being useful commercially.
Google's biggest business failing around that era was how far behind they were in cloud computing. GCE only launched in 2013. EC2 launched 2007-ish. In 2014, unless you were paying attention to Amazon, it wasn't obvious how big of a blunder this was, yet.
The fact that AWS doesn't link to the Elastic License is hilarious to me. The plain reading of the license is "APLv2 but AWS Can't Sell a Hosted Version" so of course AWS forks the last APL version and plows ahead.
Note, I'm not trying to side with either AWS or Elastic here and I fully recognize that both Elastic re-licensing and AWS forking are within each org's rights. I really just think it is funny how beside the point AWS's press release is here.
To be fair, it was actually extremely difficult for me to find the license text after Elastic's announcements too. All the announcements and FAQ's etc. seemed to omit a link to the actual text.
Thank you! I had actually tried to find it after posting my comment and couldn't find it through their website. The day of the announcement Github hadn't been updated yet, and I gave up when I saw this didn't say where to look: https://github.com/elastic/elasticsearch/blob/master/LICENSE....
There's a lot of "trust me" in this article that simply isn't backed up by the ACLU's own outline of it's free speech position: https://www.aclu.org/issues/free-speech
It is one thing for an individual in the organization to have a nuanced view of the issue and another thing entirely for the organization to have backed off a maximalist view of the right. I encourage you to read the ACLU's position (which interestingly includes "We’ve called on big social media companies to resist calls for censorship.").
Alcohol is an _extremely_ harmful drug.