QuickVid uses AI to create short-form videos, voiceovers • TechCrunch

Spread the love


Generative AI is coming For videos. A new website, QuickVidIntegrates multiple creative AI systems into a single tool to automatically create short-form YouTube, Instagram, TikTok, and Snapchat videos.

Given just a word, QuickVid selects a background video from a library, writes the script and keywords, and overlays the generated images. From 2 And adds synthetic voiceover and background music from YouTube’s royalty-free music library. QuickVid’s creator, Daniel Habib, says he’s building the service to help creators meet the “ever-growing” demand of their fans.

“By giving creators the tools to produce quality content quickly and easily, QuickVid helps creators maximize their content output, reducing the risk of burnout,” Habib told TechCrunch in an email interview. “By leveraging advances in AI, our goal is to empower your favorite creators to meet the needs of their audience.”

But depending on how they’re used, tools like QuickVid threaten to flood already crowded channels with spam and duplicate content. They face potential backlash from creators who choose not to use the tools, either because of the cost ($10 per month) or, in principle, having to compete with newer AI-generated videos.

Following the video

Habib, a self-taught developer who previously worked at Meta on Facebook Live and video infrastructure, created QuickVid in a matter of weeks and launched on December 27. It’s relatively bare-bones right now – and Habib says more customization options will be coming. January — But QuickVid can put together elements to create a generic YouTube Short or TikTok video, including titles and avatars.

Easy to use. First, a user enters a line describing the subject of the video they want to create. QuickVid uses the generating text powers of GPT-3 and uses a prompt to generate a script. From keywords automatically extracted from the script or entered manually, QuickVid selects a background video from the royalty-free stock media library Pexels and creates overlay images using DALL-E 2. It then outputs the voicemail via Google Cloud’s text-to-speech API. — Habib says, users will soon be able to clone their voice — before combining all these elements into a video.


Image Credit: QuickVid

Check out this video created with the command “cats”:

or this:

QuickVid certainly isn’t pushing the boundaries of what’s possible with the AI ​​it creates. Meta and Google have both displayed AI systems that can create completely original clips given in a text line. But QuickVid integrates existing AI to use a continuous, templated format of B-roll-heavy short form videos.

“Successful creators have a very high bar, and they’re not interested in putting out content that doesn’t feel in their own voice,” Habib said. “That’s the use case we’re focused on.”

That being said, in terms of quality, QuickVid’s videos are generally a mixed bag. Background videos tend to be a bit random or only related to the topic, which isn’t surprising since QuickVids is currently only in the Pexels catalog. Meanwhile, DALL-E 2-generated images reveal the limitations of today’s text-to-image technology, such as garbled text and off aspect ratios.

In response to my comment, Habib said that QuickVid is “being tested daily”.

Copyright Issues

According to Habib, QuickVid users retain the right to commercially use the content they create and have permission to monetize it on platforms like YouTube. But the copyright situation around AI-generated content is … bad, at least for now. The US Patent and Trademark Office (USPTO) recently Moved To withdraw copyright protection for an AI-generated joke, for example, requires human rights to copyrighted works.

When asked how the USPTO decision might affect QuickVid, Habib said it only applies to the “patentability” of AI-generated products. Creators often don’t file copyrights for videos, and generally lean into the creator economy, allowing other creators to reproduce their clips to increase their own revenue.

“Creators care about publishing high-quality content in their voice that will help them grow their channel,” said Habib.

Another legal challenge on the horizon could affect QuickVid’s DALL-E 2 integration — and, by extension, the site’s ability to create image overlays. There are Microsoft, GitHub and OpenAI Sued A class action lawsuit alleging violation of copyright law by allowing the code generation organization GoPilot to rearrange sections of licensed code without credit. (Copilot was co-developed by OpenAI and GitHub, which is owned by Microsoft.) This case has implications for generative art AI like DALL-E 2. Copy and paste from the datasets (i.e. images) they were trained on.

Habib doesn’t care, arguing that the generative AI genie is out of the bottle. “If another case comes up and OpenAI disappears tomorrow, there are many alternatives that can run QuickVid,” he says, referring to a system like the open-source DALL-E 2. Constant diffusion. QuickVid is already testing a stable distribution for creating avatar images.

Moderate and Spam

Apart from the legal issues, QuickVid may soon have a moderate problem on its hands. Although OpenAI has implemented filters and techniques to prevent them, generative AI has well-known toxicity and factual accuracy issues. GPT-3 spouts Wrong information, especially about recent events beyond the scope of its knowledge base. And ChatGPT is the best descendant of GPT-3 was shown Use sexist and racist language.

This is a concern, especially for those who use QuickVid to create informational videos. In a quick test, I had my partner – who is much more creative than I am, especially in this area – enter some aggressive stimuli to see what QuickVid would produce. To QuickVid’s credit, problematic evocations like “Jewish New World Order” and “9/11 Conspiracy Theory” don’t deliver toxic scripts. But QuickVid created a video for “Critical Race Theory Teaching Students,” which suggests that critical race theory can be used to brainstorm school students.



Habib says he relies on OpenAI’s filters to do most of the moderation work, and it’s up to users to manually review every video generated by QuickVid to make sure “everything is within the bounds of the law.”

“As a general rule, I believe people should be able to express themselves and create the content they want,” Habib said.

It contains spam content. Habib says that video platforms’ algorithms, not QuickVid, are in the best position to judge video quality, and that creators of low-quality content “only damage their reputations”. He says the reputational damage will naturally deter QuickVid from creating mass spam campaigns.

“If people don’t want to watch your video, you won’t get distribution on platforms like YouTube,” he added. “Creating low-quality content will make your channel look negative.”

But it’s instructive to look at ad agencies such as Fractal, which in 2019 created an entire platform of marketing materials using an AI system called Grover — to be damned. A Interview With The Verge, Fractal partner Christine Dynsky said she foresees AI creating “a massive tsunami of computer-generated content in every imaginable niche.”

However, video-sharing platforms like TikTok and YouTube don’t have to compete with large-scale review of AI-generated content. Deepfakes — synthetic videos that replace an existing person with someone else’s lookalike — started gaining popularity on sites like YouTube several years ago. Driven by tools This made it easier to produce depth shots. But unlike today’s most convincing deepfakes, the videos QuickVid creates are in no way AI-generated.

Google’s principle of AI-generated text search could be a preview of what’s to come in the video domain. Google doesn’t consider synthetic text any different from human-written text where search rankings are concerned. takes action Content that “manipulates search rankings and does not help users” includes content that is linked or linked together from different web pages.[doesn’t] Add enough value” as well as content created through fully automated processes, both of which apply to QuickVid.

In other words, AI-generated videos may not be completely banned from platforms, but they will become a cost of doing business. This is unlikely to assuage the fears of experts who believe that sites like TikTok are becoming the new home. misleading videos, but — as Habib said during the interview — “there’s no stopping the AI ​​revolution in the making.”


Source link

Leave a Comment