apparent

This is an interesting goal. I would agree in general, but in the age of AI I'm feeling like much of the 'labor' being done is done by AI.

I was waiting for someone to make a glueball joke, TBH.

Imagine there's a best-selling toaster. Amazon can recommend it and many people will be happy buying it. Then, one day a new toaster comes out, and it might be even better than the old best-seller. How would it get sufficient traction to be recommended by the algorithm? To send a strong enough signal, it would probably need a level of sales that could only be achieved with some level of advertising.

Separately, I'd note that Amazon doesn't want high levels of returns (which it has to pay for in many cases), so they still try to herd people toward products that they will find acceptable. Might they not love them as much as some others? Perhaps, though this is a fine line to try to walk, and I'm not sure that Amazon is actually trying to walk it.

Oh cool, will look into this. Have only used it to make podcasts in the past. I'm a little hesitant to build in a walled garden though. I'd rather store things in an open format that can be used with many platforms, not just the one.

How do you plan to make this newfound knowledge 'stick'? Do you have enough relevant/related knowledge that you'll remember it? Or is there a way to have Gemini make a bunch of cards for use with spaced repetition software?

In my mind, the externalities of reckless driving are much greater than getting credit on bad terms.

I respect that position, and you're right that not everyone is smart enough to make good choices for themselves. I don't carry a balance on my credit cards, which makes me somewhat smart, but I also don't follow all the subreddits about maximixing credit card points, moving balances from here to there, etc. Some people might look down on me for "leaving money on the table" in that regard. I could probably make a thousand dollars a year if I were more informed/aggressive.

I don't want to look down my nose at people who make decisions that I wouldn't make. They also face very different circumstances. What if someone needs to pay for an unexpected medical bill (which they cannot BNPL) and needs to choose between that and rent? What might look like a dumb choice from the outside might be an exercise in constrained optimization.

I've said that I'm open to BNPL being unavailable for certain transaction types, but I wonder if someone with your opinions would rather have them outlawed entirely?

Payday loan places don't know what you're using the cash for. They just give you a cash advance that you can spend on groceries, rent, drugs, or whatever else you want.

BNPL platforms know what you're spending on, and whether it's a one-time expense that you're spreading out, or a rent/utility payment that's going to come due next month as well.

I can sort of understand using BNPL for seasonal purchases like Christmas gifts, to spread the payments over time. But using them to pay for utilities or rent seems like a colossally bad idea.

In general, I don't favor regulation of loan products, on the notion that people should be free to get loans that fit their circumstances. But I wouldn't be opposed to limiting BNPL loans so they are not available for purchases that are (1) large dollar value and (2) recurring. Or there could be PSAs to warn people about them.

And regardless, they should be teaching kids about money management and the consequences of BNPL in school. There should be plenty of time now that kids aren't learning about "balancing their checkbook".

It says it only applies to passages over 200 words, so you could use a different LLM to rephrase every other paragraph and undermine the watermarking.

His complaint seems to center on conversations between him and the LLM, not copy being written for publication elsewhere. That is, if Claude is going to teach him a new skill, he wants it to pick the most accurate words possible, not "pretty accurate words, subject to watermarking techniques".

And I'm sure students will use tools to have every other paragraph written in the style of a different AI, in an attempt to defeat this fingerprinting.

I appreciate the new AI-slop-reporting button, and I imagine LI will use it to improve their own algorithms so they can detect and down-weight AI on their own.

At the same time, LI's top level algorithm wants to push a bunch of uninteresting crap at me, like posts that someone I know "liked". What's the point of removing AI slop if you're going to keep pushing irrelevant garbage at your users?

I am so grateful for the HNer who recently recommended this extension, [1] which lets you block posts by type. Now I only see posts made by my actual connections (which was the original idea of LI, IMO).

Interestingly, one of the side effects of this is I can now see which of my contacts are annoying over-posters, and can unfollow them. Previously their junk was buried under piles of other junk, so it wasn't so readily apparent.

1: https://news.ycombinator.com/item?id=49224161

Replying to your edit:

Amazing that you again allege insincerity, with zero evidence. You then go on to make false claims about what I said and avoid the obvious point that my problem is not with the historical figure that you quoted, but the fact that you injected race into a conversation that was about food.

As for whether you violate the HN guidelines, the number of your recent comments that are greyed out tends to indicate that I'm not the only one who thinks you're out of step here.

How does any of this make my comment insincere? Even if I agreed with it, I wouldn't understand how it proves your original accusation.

I also don't love your injection of flame-bait racial comments, which probably violate HN guidelines.

Feel free to elaborate on my alleged insincerity. Was I insincere in noticing that there are partisan patterns at play, or insincere when pointing out that the govt should have more control over govt-run/taxpayer funded programs, as opposed to transactions between private parties?

I am sincerely curious to know your opinion.

I thought the half-life was long enough that skipping a pill or taking a lower dose wouldn't make much of a difference?

Also, do you cut pills? I had looked into this and only found info saying that it ruins the delivery mechanism. I was skeptical of this because it would benefit the pill makers to convince people that these special pills cannot be cut, even though they do not require extended delivery (quite the opposite).

What's interesting is that it's different parts of the electorate that get upset. When Republicans wanted to make junk food/soda ineligible for food stamps, Dems got upset. And when Dems wanted to tax large sodas, Republicans/libertarians got upset.

I think the former makes much more sense than the latter, since it is literally being funded by the govt. That's a reasonable first step, or possibly only step, especially considering that most people on food stamps are also on govt-funded/subsidized healthcare (meaning taxpayers will end up footing the bill when they require expensive treatments resulting in part from an unhealthy diet).

They did fail him...but he was also the initial instigator of the fraud.

Seeing the title, I expected this to be a blog post trying to convince me what to blog about, written by a person who does not yet understand if his perspective is correct.

You're surprised that someone had sex with one of the richest people on the planet?

Ice cream is shrinkflated by whipping air into it. They can justify using the same size container, and allegedly selling the same volume, but much of it is just air now.

A side benefit: they can also sell it as "lower calorie" than it was before, since it's diluted with air.

EDIT: interesting to see this upvoted and then downvoted so much. Can someone who is downvoting weigh in? Seems relevant to the topic, and no one is saying it doesn't happen (or that it is happening and they like it).

A fairly well-educated victim:

> “I’m a rule-follower,” said Fudge, who majored in mechanical engineering and minored in math.

I would be more likely to teach in a private school than a public one, for sure. But I have a different job for now so am not pursuing it. That doesn't negate the point I made (and which a sibling commenter has said also applies to his/her case).

I'm not sure the point of your last comment. Of course everyone wants to be paid fairly for what they accomplish, which is not just based on intelligence, but also hard work. I have done a very good job educating my own children and have seen enough laziness in our public schools that I'm certain I would be more effective than a great many of the teachers.

But more importantly, if the incentives were aligned then even better teachers would enter, perhaps leaving me in the bottom part of the distribution. Nothing would make me happier than if schools were filled with teachers who were so effective!

Why do you think it is not relevant to the problem of teacher recruiting? I would be more likely to go into teaching if I thought it could be a good place to work and I would be paid fairly. My impression, from seeing my kids go through public schools, is that the lockstep pay (which can be bumped by getting mostly-useless online degrees) creates bad incentives. There is no way to move up quickly by being an awesome teacher, and mediocre teachers are impossible to get rid of. This hurts morale and drains districts of resources that could be used to pay better-performing teachers.

I thought the same thing regarding professors not being unionized. However, in top universities they are paid as much based on research (or outside offers based on research) as they are on teaching capability.

One thing that is truly bonkers about K12 education is that they can get tenure in just a few years. This is much faster (and easier) than a professor getting tenure. But they both result in the near-inability to be fired.

Teacher recruiting wouldn't be so hard if pay weren't lockstep, as it is in so many places. Unions demand this, and it keeps some potentially-effective teachers out of the profession (and keeps old lousy ones in, unfortunately).

Or allow pay that is not lockstep/tied to mostly-useless degree completion. Unions fight this stuff like hell because the people in charge have gone through the current system and therefore benefit from it.

It is not that hard to do outcome-based pay over a multi-year period, but this is the main criticism. You can either track kids over a number of years and track their relative growth in each year, or you can use bargaining up-front whereby teachers can vote on a "multiplier" for each class before they know who will end up with which class.

Just tried it, and it's great! I like that the blocked posts are retained and minimized/greyed out. That way I can see that it's blocking 90+ percent of the posts.

I'd love to be able to filter the feed so it only shows me actual posts by my connections. I do not care about comments my connections made on strangers' posts, and I surely do not care about posts by strangers that my connections 'liked'.

Anyone know if such a tool exists?