“AI” long predates LLM bullshit.

Hallucinations aren’t a problem with the actually medically useful tools he’s talking about. Machine learning is being used to draw extra attention to abnormalities that humans may miss.

It’s completely unrelated to LLM nonsense.

Except the summary is almost always literally the content the sites ask the sites linking them to show.

They have “please show this preview instead of a boring plain link” code.

It’s a decent book overall. If you’re interested in the theory behind choice architecture it’s worth a read.

But yeah, read it a couple months ago and remembered it specifically addressed this question.

In fact, the truth is surprisingly simple: much depends merely on what happens if people don’t make a decision, something called a no-action default, or simply a default. The countries on the left of the graph ask you to choose to be an organ donor, and those on the right ask you to choose not to be a donor. If you do not make an active choice, you are, by default, a nondonor in Germany and a donor in Austria.

Dan and I wanted to understand this. We started by asking a sample of Americans whether they would be donors or not by presenting them with a choice on a webpage. One group, the opt-in condition, was told that they had just moved to a new state where the default was not to be an organ donor, and they were given a chance to change that status with a simple click of a mouse. A second group, the opt-out condition, saw an identical scenario, except the default was to be a donor. They could indicate that they did not want to be a donor with a mouse click. The third group was simply required to choose; they needed to check one box or the other to go on to the next page. This neutral ques-tion, with nothing prechecked, is a mandated-choice condi-tion; it’s important, because it shows what people do when they are forced to choose.

The effect of the default was remarkably strong: when they had to opt in, only 42 percent agreed to donate, but when they had to opt out, 82 percent agreed to donate. The most interesting result was from those forced to make a choice: 79 percent said they would be a donor, almost the same percentage of donors as in the opt-out condition. The only difference between the group that was asked to opt out and those who were forced to make a choice was that we forced the respondents in the mandated-choice condition to pick either box before they could go forward. It shows that if forced to make a choice, most participants would become donors. Otherwise, if they were given a default, most simply took it, whatever it was.

From The Elements of Choice by Eric Johnson

It’s more complicated than the one example, and he covers it further, but as a rough guideline, it looks like forced choice and opt out are similar in this case. Which would make sense because the opposition is mostly religious and strict religious people are more motivated to opt out.

Every single thing you’ve said is factually incorrect.

There is no debate about that fact that people historically thought gods would strike people down for words; it’s abundant historical record.

And nobody anywhere near this thread said anything anyone could possibly interpret to mean that words are the same as physical assault.

I will always downvote comments using ridiculous nonsense to justify slurs.

No, they literally believed that using the name of gods could get you struck down, cursed, etc. by those gods.

And nobody is claiming words are physical weapons.

Both sides of your argument are wild mischaracterizations of reality and neither could plausibly be done in good faith.

No it isn’t. You’ve already acknowledged that many more words were historically viewed as damaging.

Acknowledging the harm of hate is more modern, but the evidence behind it is pretty much indisputable.

The context doesn’t matter because the literal only reason to use the words is to cause harm.

Magic damage felt spikier than other classes to me in Elden Ring, to the point early and mid-game where there were segments where I would run out of magic before getting through crowds even with all blue flasks.

Hogwarts Legacy. Combat is fast and brutal.

The side stuff feels kind of bland mechanically and something about the open world doesn’t capture me like I want it to, but it’s pretty good pure magic combat.

As part of its earnings call, Unity revealed that it’ had $1.4 billion’s year-over-year revenue for the quarter fell to $446.5 million from $544.2.

🤷🏼‍♀️

By default it will turn itself off after two days, but it still sleeps pretty completely without a bunch of idle power draw without doing that.

It has a pretty long battery life with no backlight and airplane mode. If you do a bunch of downloads or run heavy apps and have the backlight high, it will drain faster, but it depends how you use it. Boox pretty aggressively limits background behavior by default, though you can change some of it to allow what you want. I don’t have benchmarks or anything to give you a direct comparison, but I rarely think about battery. You’re right to raise it as a question, though.

The one thing with color specifically is that it needs more light than black and white to really shine. In bright sunlight it looks great, but indoors I generally have to raise the backlight higher than I would for other content, and that’s a good chunk of the power draw so makes a dent.

Anything bad about Android is worse on kindle or kobo’s OS. They’re more invasive, give you less privacy options, and make it much more difficult than a decent android app to organize content. I don’t actually particularly like Android, and would be miserable if I had to use it in place of my iPhone. But the device specific software is pretty much all really bad.

Don’t get a device without e-ink as a reader. It will end up in the trash where it belongs. A low resolution backlit display will just discourage actually using it to read.

I use Boox. I don’t really trust them, but Android is just way better than not Android, and their modifications to support e-ink are the best IMO.

I primarily use the go color 7, and the page turn buttons also add a lot.

Of course they aren’t, because they’re not required to, and money is money.

The fun part is that if it actually were restricted to collecting data for law enforcement? It would be a pretty obvious (though probably still not enforced because the courts suck) violation of your rights against searches without due process of law. But because it’s “publicly available”, they can pretend that it’s not really a search.

It’s really weird to include ~500% additional monthly contributions into the math.

They’re struggling because they’re not learning, or learning how to learn.

LLM outputs aren’t reliable. Using one for your research is doing the exact opposite of the steps that are required to make good decisions.

The prerequisite to making a good decision is learning the information relevant to the decision, then you use that information to determine your options and likely outcomes of those paths. The internalization of the problem space is fundamental to the process. You need to actually understand the space you’re making a decision about in order to make a good decision. The effort is the point.

Evaluating sources and consolidating the information they contain into concise, organized structures is how your brain learns. The notes aren’t the goal of note taking. They’re simply the process you use to internalize the information.