Preventing AI Plagiarism With .ASS Subtitling

Around two years ago, the world was inundated with news about how generative AI or large language models would revolutionize the world. At the time it was easy to get caught up in the hype, but in the intervening months these tools have done little in the way of productive work outside of a few edge cases, and mostly serve to burn tons of cash while turning the Internet into even more of a desolate wasteland than it was before. They do this largely by regurgitating human creations like text, audio, and video into inferior simulacrums and, if you still want to exist on the Internet, there’s basically nothing you can do to prevent this sort of plagiarism. Except feed the AI models garbage data like this YouTuber has started doing.

At least as far as YouTube is concerned, the worst offenders of AI plagiarism work by downloading the video’s subtitles, passing them through some sort of AI model, and then generating another YouTube video based off of the original creator’s work. Most subtitle files are the fairly straightfoward .srt filetype which only allows for timing and text information. But a more obscure subtitle filetype known as Advanced SubStation Alpha, or .ass, allows for all kinds of subtitle customization like orientation, formatting, font types, colors, shadowing, and many others. YouTuber [f4mi] realized that using this subtitle system, extra garbage text could be placed in the subtitle filetype but set out of view of the video itself, either by placing the text outside the viewable area or increasing its transparency. So now when an AI crawler downloads the subtitle file it can’t distinguish real subtitles from the garbage placed into it.

[f4mi] created a few scripts to do this automatically so that it doesn’t have to be done by hand for each one. It also doesn’t impact the actual subtitles on the screen for people who need them for accessibility reasons. It’s a great way to “poison” AI models and make it at least harder for them to rip off the creations of original artists, and [f4mi]’s tests show that it does work. We’ve actually seen a similar method for poisoning data sets used for emails long ago, back when we were all collectively much more concerned about groups like the NSA using automated snooping tools in our emails than we were that machines were going to steal our creative endeavors.

Thanks to [www2] for the tip!

Continue reading “Preventing AI Plagiarism With .ASS Subtitling”

The 1987 Videonics Editing System

Videonics: The Dawn Of Home Video Editing, Revisited

Here’s a slice of history that will make any retro-tech fan grin: before TikTok and iMovie, there was a beast called the Videonics DirectED Plus. This early attempt at democratizing video editing saved enthusiasts from six-figure pro setups—but only barely. Popular Science recently brought this retro marvel back to life in a video made using the very system that inspired it. Picture it: 1987, VHS at its peak, where editing your kid’s jazz recital video required not just love but the patience of a saint, eight VCRs, three Videonics units, two camcorders, and enough remotes to operate a space shuttle.

The Videonics DirectED Plus held promise with a twist. It offered a way to bypass monstrous editing rigs, yet mastering its panel of buttons felt like deciphering hieroglyphs. The ‘Getting Started’ tape was the analog era’s lifeline, often missing and leaving owners hunting through second-hand stores, forgotten basements, and enthusiast forums. Fast forward to today, and recreating this rig isn’t just retro fever—it’s a scavenger hunt.

The 1987 Videonics Editing SystemOnce assembled, the system resembled a spaghetti junction of cables and clunky commands. One wrong button press could erase precious minutes of hard-won footage. Still, the determination of DIY pioneers drove the machine’s success, setting the stage for the plug-and-play ease we now take for granted.

These adventures into retro tech remind us of the grit behind today’s seamless content creation. Curious for more? Watch the full journey by Popular Science here.

Continue reading “Videonics: The Dawn Of Home Video Editing, Revisited”

3D Printer Streaming Solution Unlocks Webcam Features

While 3D printer hardware has come along way in the past decade and a half, the real development has been in the software. Open source slicers are constantly improving, and OctoPrint can turn even the most basic of printers into a network-connected powerhouse. But despite all these improvements, there’s still certain combinations of hardware that require a bit of manual work.

[Reticulated] wanted an easy way to monitor his prints over streaming video, but didn’t have any of the cameras that are supported by OctoPrint. Of course he could just point a cheap network-connected camera at the printer and be done with it, but he was looking for a bit better integration than that. In the process, he demonstrates how to unlock some features hidden in inexpensive webcams.

He set about building something that wouldn’t require buying more equipment or overloading the limited hardware responsible for the actual printing. A few of his existing cameras have RTMP support, which allows a fairly straightforward setup with YouTube Live once Monaserver is set up to handle the RTMP feeds from the cameras and OBS Studio is configured to stream it out to YouTube. Using the OctoPrint API, he was able to pull data such as the current extruder temperature and overlay it on the video.

One of the other interesting parts of this build is that not all of [Reticulated]’s cameras have built-in RTMP support but following this guide he was able to get more of them working with this setup than otherwise would have had this capability by default. Even beyond 3D printing, this is an excellent guide (and tip) for getting a quick live stream going for whatever reason. For anything more mobile than a working 3D printer, though, you might want to look at taking your streaming setup mobile instead.

DIY Tachistoscope Feeds Your Hunger For Popcorn And Propaganda

You’ve probably heard of subliminal advertising — the idea is that behaviors can be elicited by flashing extremely brief messages on a movie or TV screen. “BUY POPCORN NOW” is the canonical example, with movies containing such subconscious messaging supposedly experiencing dramatic increases in popcorn sales.

Did it work? Maybe, maybe not, but the idea is intriguing enough to at least explore using this subliminal tachistoscope. [Roni Bandini] seems to have taken this project on as a sort of cautionary tale about brainwashing techniques, not only in motion pictures and TV but in printed media too; he goes pretty hard on the Peronistas’ use of not-so-subliminal messages to mold young Argentinian minds back in the 1940s and 50s.

The tachistoscope [Roni] presents is a little more sophisticated than those ham-fisted propaganda attempts. The Raspberry Pi-powered device downloads a video from YouTube and automatically replaces random frames with a propaganda message inspired by those used by the Peronistas, with the modified video piped to a composite video output for display on a TV.

A digital counter on the tachistoscope keeps track of the total time viewers have been propagandized. For extra fun, the machine has a switch to enable ChatGPT-created political messages to be inserted into the stream; we shudder to think what those might look like. Watch the video below for a sample of the brainwashing, but don’t blame us if you fall in love with [Evita].

We understand that this is more of a statement on the power of propaganda than an actual tool for mind control, but if [Roni] is serious about his brainwashing, some small mods might make it more effective. Thanks to the full frame of text on a black background, the subliminal messages aren’t very subliminal; they might be more subtle if the text was overlaid on the target frame rather than replaced completely. Seems like that should be possible with ffmpeg or something similar.

Continue reading “DIY Tachistoscope Feeds Your Hunger For Popcorn And Propaganda”

Going To Extremes To Block YouTube Ads

Many users of YouTube feel that the quality of the service has been decreasing in recent years — the platform offers up bizarre recommendations, fails to provide relevant search results, and continues to shove an increasing amount of ads into the videos themselves. For shareholders of Google’s parent company, though, this is a feature and not a bug; and since shareholder opinion is valued much more highly than user opinion, the user experience will likely continue to decline. But if you’re willing to put a bit of effort in you can stop a large chunk of YouTube ads from making it to your own computers and smartphones.

[Eric] is setting up this adblocking system on his entire network, so running something like Pi-hole on a single-board computer wouldn’t have the performance needed. Instead, he’s installing the pfSense router software on a mini PC. To start, [Eric] sets up a pretty effective generic adblocker in pfSense to replace his Pi-hole, which does an excellent job, but YouTube is a different beast when it comes to serving ads especially on Android and iOS apps. One initial attempt to at least reduce ads was to subtly send YouTube traffic through a VPN to a country with fewer ads, in this case Italy, but this solution didn’t pan out long-term.

A few other false starts later, all of which are documented in detail by [Eric] for those following along, and eventually he settled on a solution which is effectively a man-in-the-middle attack between any device on his network and the Google ad servers. His router is still not powerful enough to decode this information on the fly but his trick to get around that is to effectively corrupt the incoming advertising data with a few bad bytes so they aren’t able to be displayed on any devices on the network. It’s an effective and unique solution, and one that Google hopefully won’t be able to patch anytime soon. There are some other ways to improve the miserable stock YouTube experience that we have seen as well, like bringing back the dislike button.

Thanks to [Jack] for the tip!

Prompt Injection: An AI-Targeted Attack

For a brief window of time in the mid-2010s, a fairly common joke was to send voice commands to Alexa or other assistant devices over video. Late-night hosts and others would purposefully attempt to activate voice assistants like these en masse and get them to do ridiculous things. This isn’t quite as common of a gag anymore and was relatively harmless unless the voice assistant was set up to do something like automatically place Amazon orders, but now that much more powerful AI tools are coming online we’re seeing that joke taken to its logical conclusion: prompt-injection attacks. Continue reading “Prompt Injection: An AI-Targeted Attack”

Screenshot of ImHex hex editor, with the MOC3 file structure being reverse-engineered inside of it

Live2D: Silently Subverting Threat Models

In online spaces, VTubers have been steadily growing in popularity in the past few years – they are entertainers using motion capture tech to animate a special-sauce 2D or 3D model, typically livestreaming it as their avatar to an audience. The tech in question is pretty fun, lively communities tend to form around the entertainers and artists involved, and there’s loads of room for creativity in the VTuber format; as for viewers, there’s a VTuber for anyone’s taste out there – what’s not to like? On the tech side of making everything work, most creators in the VTubing space currently go with a software suite from a company called Live2D – which is where today’s investigation comes in. Continue reading “Live2D: Silently Subverting Threat Models”