recursive

Not caring is the middle ground.

Nothing.

To the extent that we have to live with AI and LLM, I believe we have to live with distilling also.

The responses will continue until distillation-posting improves.

If we keep hearing accusations about how someone distilled something from someone, it seems like one of the few reasonable responses.

If my friend keeps complaining about how the inside of his car is wet, I will probably keep telling him to close the windows when it's raining, even if he thinks that is a tiresome take, lacking in insight.

Training a model is objectively easier than generating the sum total of human creative output prior to 2020. That's why the big labs are doing it. What's the difference here?

Missing something? You're missing the boat! Have some AI-prepared koolaid before you get left behind.

I've subscribed for years. I've never hit a limit once.

People who say "clanker" really want to say other words with a "hard R".

It sounds like you know a lot about my internal motivations. Evidently a lot more than I do. I've heard this take, and I don't get it. I'm a human supremacist. If that's worthy of cancellation, go ahead. But it just seems like intentional confounding of issues.

Yeah. "Misleading" is maybe better.

Surely if the machine you go to for answers regularly makes things up you would just stop using it? Perhaps people have to be burned by something really bad personally before they realise the limitations? LLMs are very convincing and persuasive.

Psychics are still in business. Although to be fair they probably don't have as much revenue. I think people really enjoy being told how smart and insightful they are, and how much they've really cut to the crux of the issue. This isn't the whole thing, but I think it counts for a lot.

Calling it a "few dollars" is probably disingenuous, although we don't know the personal circumstances. For many, this could be a "life-changing" quantity of money.

This is a far cry from the gravity of the post, but here's my personal anecdote about trying to navigate health care and insurance.

A month or two ago, I was trying to set up a pediatric dentist visit for my son. Not a medical emergency, but a matter of some urgency. I spent about three hours on a total of a dozen phone calls or so. Some to the insurer, and some to various dentist offices in the region to inquire about scheduling and logistics. I learned that I needed to switch my son's primary office before he would be eligible to receive care. This could only be done once per month. Fortunately, by luck, this would take effect at the close of business on that very day. Next I needed to find a covered provider. Of course, the insurer could not tell me which ones would have openings at short notice, so I had to call several. Then I learned that in order for a pediatric visit to be covered, it needed a referral from a general dentist. The pediatric office helped me find a dentist who could fill out this paperwork, which felt lucky. I verified that my policy worked with this dentist office. Several hours on the phone, but I had a sense of accomplishment for sorting it all out.

The day of the appointment came. The visit went smoothly. Then I found out that it wouldn't be covered because the insurance will not cover pediatric dentistry above the age of 6, but my son is 7.

I think I'm going to opt out of dental insurance altogether in the next enrollment. All the individual people seemed to be doing their best to help me, but the end result is a system that is impossible for me to use.

Clarity is a hallmark of good writing.

The anti AI wave still has so much work to do. Today I saw a duckduckgo ad selling how you can opt out of AI. This is the first ad I've seen suggesting the possibility that someone might not want to be maxing out AI in every possible way. This after seeing probably thousands of ads reference AI.

Check out some of Trump's positions on renewable energy.

It doesn't matter if you can switch easily. It might matter if there are barriers to switching.

Interesting. One of my biggest surprises from all this was that the timing of the dancing seemed significantly wrong. I would have thought that would be one of the easiest things.

The upsides for social media were there too.

But accurate next-token prediction has to model functional affect too, it's a part of natural language

There are a lot of reasons I dislike LLMs. You just found a way of expressing one of them more clearly than I've ever been able to. LLMs use natural language. They do it persuasively and rhetorically. Most humans are just not equipped to defend against this type of simulacrum.

for some reason

Usage is metered/billed by the token. This suggests a few possible hypotheses for why they might tend to be verbose.

I recall having a conversation with someone many moons ago. They asked me a very weighty and significant question, and I answered it. Then they asked me to "promise". This was really thought-provoking for me.

To this day, it's the only part I remember. I told them I would not promise, as everything I said was true. Making a specific promise would create an implication that I'm generally untruthful, unless I "promise".

Un-blocking is still easier than un-building.

The royal "we"! You know, the editorial "we"!

You don't always have to do division by hand. I would make the argument that it's probably good to do it occasionally. At least enough to remember how.

Sounds like something you'd need to evaluate before adopting then.

I always thought bootstrap was pretty good. All the gradients and sparkles don't do much for me.

If it wasn't accidental, that bite represents an attempt to bite.

Why should these developers rebuild a core piece of their stack every few months?

That's what they signed up for when established a hard dependency on an subscription online-only LLM model.

It ain't just google.

These are not analogous. If you paid for an offline model, and you were somehow revoked access because it was phoning home, that would be closer.

If you were paying for an ongoing subscription for a service, that would be something different.