That's what the package uses. And if `crypto.randomUUID()` doesn't exist, it falls back to `crypto.getRandomValues()`, which per the documentation isn't AS strong:
I don't understand these blog posts. Unless you're an absolute beginner, what is so hard about doing any of these tasks yourself or reviewing AI output?
AI won't think for you. It just (sort of) does work you delegate to it. It's from this perspective that you might get productivity gains, but the human in the loop is still the bottleneck. This doesn't mean you remove the human. It means the human decides how much they need AI.
There is no AI workflow that removes the human and magically does exactly what was intended. That's wishful thinking.
I think what matters most is that they get out and socialize.
If there is alcohol involved with that, it's a personal choice that I don't think deserves shame unless it's consumed in a shameful way. I would have to assume any adult should know their limits, be resilient against peer pressure, etc.
I would actually argue that the fights you have with yourself regarding addictions are just as much of a rite of passage as the fights you have with others while drunk.
If you so strongly wish you could eliminate negative experiences, you'll eventually have to ask yourself "compared to what?" and realize such a treadmill of misery is your true source of regret, not your actions or the consequences.
> It demonstrates that Tesla’s hardware and software stack is more portable than the company’s licensing struggles would suggest.
Unless I missed something, this is a completely unsupported claim by the article. Passion projects and retrofits are nothing at all like manufacturing.
The article claims that the whole project only cost $40,000, and then compares that to electric conversion offerings that cost $75,000 (and mentions that the global conversion market in 2024 was $5.9 billion). I think the implication is that there could be a large market for FSD conversions that goes beyond passion projects because it is not only possible but affordable.
I would be surprised however if this project only cost $40,000, when you factor in the cost of labor and maintaining a facility to do this work.
It's specifically talking about the "FSD" model under the hood being able to run on this retrofit even though the cameras don't align 100% like they originally would.
Yes, but when all it takes to avoid this chaos is hiring someone with at least 5 or 10 years of experience for a reasonable wage, this entire perspective looks insane.
It's... just... not that hard to write code nor does it cost that much. There are millions of us working silently at places that aren't "big tech". We all shrugged our shoulders, took a sip of coffee, and went back to our Teams meetings where the only LLM usage is still just Copilot.
Well, "better" is relative. Web UIs definitely have the advantage you are mentioning. However, they are also much slower to use, so it is a tradeoff.
I am blind and do rely on accessiblity. But I would choose a TUI over a Web UI every day simply because the web is slow as hell. Feels like running away from a "monster" in a dream. You have rich details, but you are being drowned in unnecessary verbosity and an inherently slow stack of tools working together.
The web might be accessible, but it is barelyuseable in practice.
Its more or less the same with every screen reader and browser I know. If there was a simple fix for this inherent issue, I wouldn't have to talk about it. You might get a bit of stuff out of the way if you configure less verbosity of the screen reader, but it doesn't really help with slow interactions. Problems include:
* Virtual buffer: In most systems, the DOM is rendered to a "virtual buffer" and the screen reader lets you navigate that, because cursor (caret) support was on the map for accessibility, but was never really sufficiently implemented on the browser side. So screen readers had to solve it in a separate step.
* Using the keyboard to do screen reading and navigation on a site conflicts with keyboard shortcuts from that site. So most screen reader have two modes: You switch between entering text in a field and navigating/reading the site. Its basically like vi insert mode, but less simple.
Those are the major issues from my POV. The rest of the slowness likely comes from having to go through the accessibility API of your OS.
Am I the only one who doesn't get either side of this?
AI still sucks pretty bad at writing code. The only people I've ever known to need a "flow" state to write code are junior devs.
Everyone else is used to constant interruptions and has been through every layer of abstraction many many times. This is why those with experience find it so tempting to say this job can be automated away, but they forget how many gotchas there are, how they crop up, and how brittle all this crap always was. AI is actively making this problem worse.
In truth, there is no "correct way" to argue. What convinces people says more about the audience.
For many audiences, it isn't even about reason. That's especially true online where it's just power struggles between incoherent groups.
In the specific case of atheists, they are arguing about something non-falsifiable. Those topics are natural cesspools for grifters and charlatans. It's one thing to study the topic, but quite another to give fiery speeches and sell books to people desperate to find their identity somewhere in that slop.
There is of course nuance as with everything. Potentially arguing in bad faith myself, I don't consider discussing non-falsifiable claims, giving fiery speeches to sell books and engaging in the general cesspool of internet mud slinging to be "argumentation." If we're considering dunk-style quips and counter attack as arguments, then sure, we can step into this highly relative middle ground you've proposed. But those are fights and grandstanding.
A (productive) argument is about arriving at truth through discourse, the other stuff is largely vitriol or unproductive. It's unfortunate that we have the same word to apply to two different concepts.
The OP said, "I was taught early: Examine and, if necessary, attack both, for the credibility of a person..."
"Attack the person" is not a productive / valid way to approach a counter argument. It's even got a fallacy associated with it. If Bibi says, "All people deserve the right to defend themselves." It's not a useful or appropriate response to respond with, "You have zero credibility as the leader of an apartheid state so anything you say is meaningless." It may be true that a person offers arguments in bad faith, but also, the concept introduced by this person can be true in themselves.
It's more difficult to unpack, analyze and develop more sharp and compelling arguments against. It's a lot easier to say, "shut up, bitch."
Both have a place in culture and society, but to say something like, "I was taught to attack the person" with reference to argumentation and then to defend it with, "well... there's no right way it's whoever is listening" is somewhat disingenuous.
I dunno. The truth is usually extremely straightforward to understand if you're willing to approach it without preconditions or any other wishful thinking. It's when people refuse to let all that go that they begin to argue at all.
Debate is usually not about finding the truth, but to enlighten someone on how their beliefs hid the truth from them.
I can very confidently say "you know what fuck you may you rot in hell for all eternity" mid-argument to a televangelist or some quack claiming to cure cancer with snake oil. Very few would disagree with me on here, but if I was at one of their gatherings I'd probably get beaten down. Some people are just easier to attack that way because the insult isn't far from the truth.
If you play the messenger deductively, then its invalid. So much for formal logic. Most arguments in practice are defeasible, and the ad hominem, for better or worse, often plays a significant and essential part in their resolution (addressing reputations, biases, agendas, interest conflicts, and so on). Virtually all interactions between people, even formal ones (e. g. in courts), are not some school exercise in formal logic.
Why? There's a built-in for this.
https://nodejs.org/api/crypto.html#cryptorandomuuidoptions
reply