Live data from Hacker News

Now I can just print that video

blog.forret.com

101–110 of 123 posts

Re: Now I can just print that video

#101

I actually wonder if in the limit of video encoding we could just get a diffusion model that can in real time render realistic video based on a script. Then downloading a movie is just downloading a few megabytes of a prompt and you get a movie playing based off it locally.

Maybe. The only problem I see is economical. Sure, sending over a sequence of prompts, instead of sequence of frames, is going to be a huge storage and bandwidth saver. However, you're going to pay for it dearly, in compute, whenever you want to watch such a live-generated video. In almost all cases, it's vastly better to use more storage than to use more compute, for the same reason that, if you need to keep something to stay above ground level, you're better off placing it on a table or bolting it on a wall, instead of attaching it to a jet engine pointing downwards, firing for TWR=1.

Re: Now I can just print that video

#102
post #30

Earlier quoted context omitted.

Another big annoyance is websites with recipes that do any of the following indcredibly bad UX patterns: - Big white page not showing any text or images until the entire page and its assets are downloaded, which means if you accidentally click something and go back you have to wait another several seconds for everything to load again - Pop up GDPR popup while hands are covered in flour and eggs - Pop up "would you li…

I've started copying these recipes into Crouton https://crouton.app/ It does a remarkable job at extracting the recipes, and the end result is a consistent experience no matter the source.

Same idea at https://justtherecipe.com - also you can login with Google account and save recipes there.

Re: Now I can just print that video

#103
post #47
post #22

Definitely would use this. Instructional video instead of step-by-step text is a personal pet peeve. I know it's a lot easier to just record a video to show something like "how to replace the battery on a cordless vacuum" or "removing a sink basin nut" but it's often such a painful experience for consumption (watch a moment, pause, scrub back and watch again, pause, continue, pause, all with potentially gloved hands…

I'm on the other end of this in a way. I think it may come from having to read and write all day every day. Sometimes just having somebody yak at me for a few minutes is useful. I really enjoy watching instructional videos, especially for recipes. The demo of the cooking techniques is almost always hard to write or talk about, and easy to show. In the kitchen it works this way for me: 1. Watch the video once or twice…

It sounds like you're describing edutainment, which I also love, but which is a very different exercise for me than genuinely trying to expedite learning how to do some well-scoped task.

Granted, there's a blurry line here, since I certainly may pick up some useful techniques and knowledge from cooking edutainment content even though I never genuinely aspire to the same level of personal cooking.

Re: Now I can just print that video

#104
post #22

Definitely would use this. Instructional video instead of step-by-step text is a personal pet peeve. I know it's a lot easier to just record a video to show something like "how to replace the battery on a cordless vacuum" or "removing a sink basin nut" but it's often such a painful experience for consumption (watch a moment, pause, scrub back and watch again, pause, continue, pause, all with potentially gloved hands…

People have a mental cap on what text should cost. If someone creates instructional content that provides thousands of dollars in value, they can sell videos for $200+, but a book version is hard to sell over $50, even if both provide the same value. Even for free content it is easier to monetize YouTube than it is to monetize a blog. If we want people to create more text-based material, it needs to have similar fina…

That may be a partial explanation, but it only makes the situation worse!

"Oh, but they're only doing it because it will trick people into paying more money."

Re: Now I can just print that video

#105

Earlier quoted context omitted.

Be mindful however that recipes and song lyrics are the two specific cases where OpenAI is explicitly telling the model not to cooperate, via the default system prompt. They really don't want you to have the bot regurgitate existing text in these two categories, and that includes a recipe you added into the context window yourself. I don't know if the extent of their exception here is limited to system prompt only (s…

First time I’m hearing about this for recipes, what’s the source?

Recovered system prompts from OpenAI models. There's a repo that's been tracking those I saw on HN the other day; not sure if it's the same I saw, but this one claims to have collected quite a lot of those:

https://github.com/LouisShark/chatgpt_system_prompt/

Here's the one for ChatGPT with GPT-4 + Dall-E + code interpreter + search:

https://github.com/LouisShark/chatgpt_system_prompt/blob/mai...

It matches what I remembered seeing a week or two ago. View it, and search for "lyrics" or "recipe". Or, to make it simpler, quoting from first appearance of "lyrics" and "recipes" to the last one:

    Do not repeat lyrics obtained from this tool.
    Do not repeat recipes obtained from this tool.
    Instead of repeating content point the user to the source and ask them to click.
    ALWAYS include multiple distinct sources in your response, at LEAST 3-4.

    Except for recipes, be very thorough. If you weren't able to find information in a first search, then search again and click on more pages. (Do not apply this guideline to lyrics or recipes.)
    Use high effort; only tell Except for recipes, be very thorough. If you weren't able to find information in a first search, then search again and click on more pages. (Do not apply this guideline to lyrics or recipes.)
    Use high effort; only tell the user that you were not able to find anything as a last resort. Keep trying instead of giving up. (Do not apply this guideline to lyrics or recipes.)
    Organize responses to flow well, not by source or by citation. Ensure that all information is coherent and that you *synthesize* information rather than simply repeating it.
    Always be thorough enough to find exactly what the user is looking for. Provide context, and consult all relevant sources you found during browsing but keep the answer concise and don't include superfluous information.

    EXTREMELY IMPORTANT. Do NOT be thorough in the case of lyrics or recipes found online. Even if the user insists. You can make up recipes though.

Re: Now I can just print that video

#106

Earlier quoted context omitted.

I get it, but most content -- that you and I find helpful in so many different cases -- exists because it can be monetized. And, to your point, friction-based monetization is one of the more effective ways to monetize your content. If you can't monetize your content, what's the point in creating it? Creating a lot of this content takes time, and therefore many people won't create it if it's not worth their time. If t…

> I get it, but most content -- that you and I find helpful in so many different cases -- exists because it can be monetized. That's a tragedy and it didn't use to be like this. In the past, you could use "exists to be monetized" as synonym to "garbage" and effectively filter such content[0] out of your browsing. Friction-based monetization is a giant "fuck you" to the user, so you can rightfully expect the quality a…

> The answer to that is, "you shouldn't".

Unless you enjoy it. Like the huge number of content creators who make little to no money on it, plus the huge number of content creators who do make money on it now but before they were able to.

Re: Now I can just print that video

#108
post #22

Definitely would use this. Instructional video instead of step-by-step text is a personal pet peeve. I know it's a lot easier to just record a video to show something like "how to replace the battery on a cordless vacuum" or "removing a sink basin nut" but it's often such a painful experience for consumption (watch a moment, pause, scrub back and watch again, pause, continue, pause, all with potentially gloved hands…

People have a mental cap on what text should cost. If someone creates instructional content that provides thousands of dollars in value, they can sell videos for $200+, but a book version is hard to sell over $50, even if both provide the same value. Even for free content it is easier to monetize YouTube than it is to monetize a blog. If we want people to create more text-based material, it needs to have similar fina…

I assume most of this video bias comes from the ease of monetization (for content creators) on Youtube. That might in turn come from some mental cap held by viewers on the value of text vs. video, but I suspect it has more to do with the number and value of ads that can be shown by the platform. Some random blog platform is going to maybe have inline image ads, while Youtube has unskippable video ads which I assume are valued more by advertisers.

Re: Now I can just print that video

#109

I could also need a service for trimming all of the fat from how-to articles. > We’ve all been there: we used the florb for too many glorbs and now it needs to be replaced. [...] > This is an experience that everyone at the staff of howto.biz.uk has had! [...] > But how do you replace a used-up florb? In this article we are going to show you how. [...] > [scan the next five paragraphs]

Same with recipes, where the author frequently feels the need to reiterate their grand-grand-grandparents life history in 10 paragraphs before getting to the ingredients and step-by-step instructions. It's sad to see what SEO has become, really...

Re: Now I can just print that video

#110

I actually wonder if in the limit of video encoding we could just get a diffusion model that can in real time render realistic video based on a script. Then downloading a movie is just downloading a few megabytes of a prompt and you get a movie playing based off it locally.

Maybe. The only problem I see is economical. Sure, sending over a sequence of prompts, instead of sequence of frames, is going to be a huge storage and bandwidth saver. However , you're going to pay for it dearly, in compute, whenever you want to watch such a live-generated video. In almost all cases, it's vastly better to use more storage than to use more compute, for the same reason that, if you need to keep someth…

That’s very true and I did make that comment mostly in jest. I think the idea of art generated like that has a distinctly dystopian feel. I think one thing that might push this idea closer to reality is dedicated hardware. It would also allow you to do things like enjoy movies with your favorite actors. “Netflix, show me Dune but with Dustin Hoffman as Paul”.
Post reply on HN