Ryan Gosling Deepfake Explains How Ai Works

admin — Thu, 15 Aug 2024 13:44:41 +0000

If you know what to look for, you can easily spot some anomolies in deepfakes. For instance if you look at frame at the 0:40 mark, Ryan suddenly grows an extremely long left arm supporting the blackboard.

But what if you didn’t pay attention as closely? Sure, there are other outwardly signs that this is a deepfake such as odd eye movement as well as softness – not as sharp and clear – around the eyes and mouth to pull this off.

And how about that voice! Very convincing as Ryan Gosling.

Ai-generated video will be very problematic with a fiercely competitive and contested US election looming. Many will believe anything that comes out of their candidate’s mouth (or the opposing candidate) and run with it being the truth. Only it is not.

Going forward, many organizations are crying for some type of Ai digital watermarking system which would label such content as being Ai-generated. There are bills stalled in Congress (not a surprise) from bi-partisan members (also not a surprise) pushing legislation to outlaw, among other things, deepfake porn (also know as “revenge porn).

Until some guardrails are put up, you’re on your own to determine is it real? or is it (deep)fake?

#deepfakes, #AIgeneratedvideo, #digitalwatermarking, #deepfakedetection, #USelection, #deepfakelegislation

Content Summary: Admin – Gerald | Logo: Respective ™ and © owner

The post Ryan Gosling Deepfake Explains How Ai Works first appeared on KLiPS Ai.

AI Giants Use Unauthorized YouTube Subtitles for Training

admin — Tue, 16 Jul 2024 15:36:11 +0000

The article from Proof News reveals that Apple, Nvidia, and Anthropic have controversially utilized subtitles from over 173,000 YouTube videos to train their AI models without obtaining permission from the content creators. This dataset, known as YouTube Subtitles, was compiled by EleutherAI as part of its Pile dataset, which is used to train language models. The dataset includes a wide range of videos, from educational content to videos by well-known YouTubers like MrBeast and PewDiePie.

Content creators have reacted with outrage, describing the unauthorized use of their work as theft and disrespectful. Many creators argue that this practice underscores a significant issue in the AI industry: the use of copyrighted material without proper authorization or compensation. This incident has sparked a broader discussion about ethical AI training practices and the rights of content creators.

The dataset was created by scraping publicly available subtitles from YouTube, which some argue falls into a gray area of legality. While these subtitles are publicly accessible, their use for training commercial AI models without permission raises ethical and legal concerns. This situation highlights the need for clearer guidelines and regulations regarding the use of online content for AI training.

EleutherAI, the group behind the Pile dataset, has faced criticism for including these subtitles. The Pile is a 825-gigabyte dataset designed to train language models, and it includes data from a variety of sources, including books, websites, and now YouTube subtitles. The inclusion of these subtitles was intended to improve the AI’s understanding of diverse and informal language.

In response to the backlash, some AI experts argue that the industry needs to develop more transparent and fair methods for acquiring training data. They suggest that companies should seek explicit permission from content creators and provide compensation when using their work. This incident serves as a wake-up call for the AI community to address the ethical implications of using publicly available data without proper consent.

Overall, the controversy surrounding the use of YouTube subtitles for AI training highlights the ongoing tension between technological advancement and ethical practices in the AI industry. As AI continues to evolve, finding a balance between innovation and respect for creators’ rights will be crucial.

For more details, you can read the full article here.

@ProofNews

The post AI Giants Use Unauthorized YouTube Subtitles for Training first appeared on KLiPS Ai.

unauthorized - KLiPS Ai

Ryan Gosling Deepfake Explains How Ai Works

AI Giants Use Unauthorized YouTube Subtitles for Training