Honestly, people make more of this than it is. I say that as someone whoes reading level in the 8th grade was rated “post-High School” in tests. Though IIRC, that particular test wasn’t considered accurate past a 10th grade reading level or so. Suffice it to say, though, I was always rated at least a few grade levels higher than my actual grade level when it comes to reading.

If you pick up examples of post-High School writing, you’ll find it’s hard to read. Basically, check any abstract on a paper for a technical field. It’ll be full of field-specific jargon and long sentences. Copy and paste it into a writing assistant like Hemmingway, and it will scream at you to simplify the sentence structure.

Converting to terms of Lexile level, Fellowship of the Ring has a rating of 860L. By a conversion chart, we would expect 50% of students to be able to read it by the spring of 4th grade. Even the bottom 10% of students can read it by the beginning of 10th grade.

That’s a relatively hard book; harder than what most fiction asks of you. Of Mice and Men, which is on plenty of High School reading lists, only has a Lexile level of 630L. Conversely, Romeo and Juliet can go up to 1260L (though this varies depending on the editing of different editions).

They’re a vacation where everything is taken care of for you. Find a spot, read a book, get all the drinks you want. Need food? Walk over to the chosen food place. Even with thousands of people on board, you can generally find a quiet spot with drinks.

There’s all-inclusive resorts, yes, and I’ve found they’re generally more expensive than cruises. If you make your resort hotel float, it’s cheaper. I don’t know why.

I’d only go anymore if it’s a trip that would show things you generally can’t see other ways, such as the coast of Alaska or Norway, or going through the Panama Canal. Caribbean cruises are an absolute waste.

“If it’s stupid but works, it wasn’t stupid” doesn’t apply to safety or security.

they repeat what i said, did you read them? previously ai model training was entirely based on simply buying more chips as fast and as hard as possible, deepseek changed that

Yes, and it says exactly what I claimed. DeepSeek is an improvement, but not to the level initially reported. Not even close.

Youtube uses a fuck ton of power but is an incredibly efficient video delivery service

What a colossally stupid thing to say. We’re not looking at starting up new nuclear reactors to run YouTube.

why are you linking me to articles i read ages ago?

Perhaps because you didn’t understand what they said.

You think AI is only useful if it’s taking someones job?

It’s why companies are dumping billions into it.

If the models were actually getting substantially more efficient, we wouldn’t be talking about bringing new nuclear reactors online just to run it.

Deepseek is not the huge leap it appears. It’s better, but not at all what was initially claimed.

Nor is smartphone AI going to do the things people what AI to do. It won’t let the CEO take your job.

Don’t worry about it, it’s just more efficient.

But I was told capitalism was the most efficient system.

The current models depend on massive investment into server farms. They aren’t generating profit, and probably can’t. When the companies involved realize it’s not going to happen, they’ll pull it all. That will generate a new AI winter, and in 20 or 30 years, maybe the pieces will be picked up and the field will go through another summer cycle. This sort of boom/bust cycle has happened before in AI.

And no, self-hosted models aren’t going to make up for it. They aren’t as powerful, and more importantly, they will never be able to drive mass market adaptation.

Every ticket scheme for NFTs fails because of a simple reason: contract law. Venues don’t stick with TicketMaster because they like it. TicketMaster’s store front doesn’t have any magic technology; a room of overcaffinated fresh CS grads could recreate it in a weekend of binge coding.

Venues stick with TicketMaster because they are contractually obliged to do so. NFTs do not and cannot change that legal reality.

Trusted, cryptographic timestamps exist, and have for some time. NFTs don’t add anything new.

Search engines favor text earlier in the site. Text “above the fold” (the area where you wouldn’t have to scroll to see it) is scored higher.

https://www.pedalo.co.uk/seo-experiment-text-position-keyword-rankings/

Those are put there for SEO purposes. Google favors sites with these big stories. The copyright issue alone doesn’t justify what’s there; you could do a quick blurb of a few sentences and it would be enough. Plenty of cookbooks do that.

This is why a lot of those sites have a button that says “skip to recipe”. It’s a bunch of text that’s meant to be for robots, not you, and they really don’t care if you read it.

Now that it’s being created by LLMs, we may have the first known example of human language written by robots and intended for robots. Welcome to a cyberpunk dystopia.

A racist old drunk white shithead in a bar in Ohio once yelled at my wife to go back to the res. My wife barely has any native ancestry, has never lived on a reservation, and the ancestor in question was put through the boarding school program where the whole point was to erase their native identity. Their family can only guess what tribe they’re even from.

No, I don’t like Ohio very much. Too many people like that.

So most photosynthetic organisms are not limited by CO2, which is why having more CO2 in the atmosphere won’t cause plants to explode in growth. There are a few exceptions, though, and one of them is the algae that causes red tide.

Meaning, no, Florida will still be red.

I built my own blog that way. All static generated, no JavaScript, no cookies, just enough CSS to get a nice dark mode look. Loads in 0.3 seconds on a modern connection, or around 10 seconds if you’re on a shitty 2G connection. 370KB load, and about 270KB of that is a picture on a post that could be slimmed down if it used something more modern than jpeg compression.

GDPR says you shouldn’t get a single cookie until you click the consent button. Try this: clear all cookies for a web site that has one of these banners, refresh the page and let it finish loading, and then see how many cookies you have for it before you consent to any.

FLAC is where it’s at. Oddly, most of the head units that understand FLAC don’t have CD drives at all. If it has a CD drive still, it probably only understands MP3.

Which is one response to the question of “why would you encode an MP3 at a high bitrate when you can just use FLAC?” It’s because I had a car that didn’t FLAC.

I built one, but by that time our local FM radio waves were so saturated that there was no good frequency to use.

There’s nothing about being larger that makes access speed inherently slower. We just have to use cheaper technologies to improve density. CPU cache is usually SRAM, which is less dense than DRAM, but faster. 1GB of SRAM would be god tier. Even the Ryzen X3D chips only have 96MB of L3 cache, all SRAM, and those are sick.