QuickVid uses AI to generate short-form videos, complete with voiceovers • TechCrunch

Generative AI It is coming Videos A new website. QuickVid, combines several generative AI Automatically creating systems using one tool short-form YouTube, InstagramTikTok Snapchat videos.
Given As little as one word QuickVid chooses a background video from a library, writes a script and keywords, overlays images generated by DALL-E 2, and adds a synthetic voiceover and background music from YouTube’s royalty-free music library. QuickVid’s creator, Daniel Habib, says that he’s building the service to Help creators meet the “ever-growing” Fans should be demand.
“By providing creators with tools to quickly and easily produce quality content, QuickVid helps creators increase their content output, reducing the risk of burnout,” Habib Telled TechCrunch Send an email interview “Our goal is to empower your favorite creator to keep up with the demands of their audience by leveraging advancements in AI.”
But depending on how they’re used, tools like QuickVid Threaten to Flood already-crowded channels with Spammy and duplicate content They Creators who do not opt-out could also face possible backlash to You might use the tools, regardless of whether it is cost ($10 per months) or because you believe they are useful. to Compete with A raft of new products and services available in AI-generated videos.
Going after video
QuickVidWhich? Habiba self-taught web developer who worked previously at Meta On Facebook Live Video infrastructure and video infrastructure were built in just weeks. They were launched on December 27. It’s relatively bare bones at present — Habib According to the company, there will be more personalization options. January — but QuickVid You can put together the components that make up an informational YouTube Short TikTok video, captions, and avatars.
It’s easy to use. The first user must enter a prompt to describe the subject matter of the desired video. to create. QuickVid uses The prompt to generate A script that uses the generative text capabilities of GPT-3. Keywords can be extracted automatically from the script or manually entered. QuickVid Select a background video from our royalty-free stock media collection Pexels DALL-E 2 creates overlay images. It The voiceover is then outputted via Google Cloud’s text-to-speech API — Habib Users will soon be able to use the app, according to the company. to clone their voice — before combining all these elements into a video.

Image Credits: QuickVid
See This video was made with The prompt “Cats”:
Or This one:
QuickVid certainly isn’t pushing the boundaries of what’s possible with generative AI. Both Meta Google has displayed AI Systems that can generate Completely original clips given as a text prompt But QuickVid amalgamates existing AI to Profit from the repetitive, templated format for B-roll-heavy short-form Videos to get around the problem of not having to generate The footage itself.
“Successful creators have an extremely high-quality bar and aren’t interested in putting out content that they don’t feel is in their own voice,” Habib said. “This is the use case we’re focused on.”
That This is supposedly true in terms of quality. QuickVid’s videos are generally a mixed bag. The Background videos tend to be a bit random, or only tangentially related to The topic, which isn’t surprising given QuickVids is currently limited to the Pexels catalog. The DALL-E 2-generated images, meanwhile, exhibit the limitations of today’s text-to-image technology, such as garbled texts and off proportions.
In Response to My feedback Habib That was conclusion QuickVid Is “being tested and tinkered with daily.”
Copyright Questions
According to Habib, QuickVid Users retain the right to You can use the content they have created commercially if they have permission to It can be monetized on platforms such as YouTube. But Copyright status around AI-generated content is … nebulous, at least presently. The U.S. Patent Trademark Office (USPTO), recently Moved to Copyright protection revoked AIFor example, a comic that is generated by computer software states that copyrightable works must be written by humans.
When Ask about the possible effects of the USPTO decision QuickVid, Habib He said that he believes it only applies to the United States. to The “patentability” Of products that are generated by the creators, not their rights to Make use of their content and monetize it. Creators, he pointed out, aren’t often submitting patents for videos and usually lean into the creator economy, letting other creators repurpose their clips to Their reach can be increased.
“Creators care about putting out high-quality content in their voice that will help grow their channel,” Habib said.
Another Legal challenge on the horizon could have an impact on QuickVid’s DALL-E 2 integration — and, by extension, the site’s ability to generate Image overlays MicrosoftGitHub, OpenAI and OpenAI are available. sued In a class action lawsuit in which they are accused of violating copyright laws by allowing Copilot, a code-generating system, to You can recite sections of licensed code without credit. (Copilot OpenAI and GitHub co-developed it. Microsoft owns.) The Case has implications for generative arts AI Similar to DALL-E 2, also found to Copy and paste data from the datasets they were trained on (i.e. images).
Habib isn’t concerned, arguing that the generative AI genie’s out of the bottle. “If another lawsuit showed up and OpenAI disappeared tomorrow, there are several alternatives that could power QuickVid,” He said, referring to The DALL-E 2 open-source system is similar to the DALL-E 1. Stable Diffusion. QuickVid It is already being tested by Stable Diffusion For creating avatar pictures.
Moderation Spam
Aside From the legal dilemmas QuickVid, It might soon face a moderation issue. While OpenAI has implemented filters. to Stop them! AI has Well-known factual accuracy and toxicity problems. GPT-3 spouts MisinformationParticularly about recent events, that are beyond its knowledge base. And ChatGPT is a fine-tuned offspring of GPT-3. Shown to Use of racist language
That’s worrisome, particularly for people who’d use QuickVid to Make informative videos. In a quick test, I had my partner — who’s far more creative than me, particularly in this area — enter a few offensive prompts to See what QuickVid Would generate. To QuickVid’s credit, obviously problematic prompts like “Jewish new world order” and “9/11 conspiracy theory” didn’t yield toxic scripts. But For “Critical race theory indoctrinating students,” QuickVid A video was created implying that critical race theory could be used to brainwash schoolchildren.
See:
Habib says that he’s relying on OpenAI’s filters to do most of the moderation work and asserts that it’s incumbent on users to Every video created is manually reviewed QuickVid to You can be sure “everything is within the boundaries of the law.”
“As a general rule, I believe people should be able to express themselves and create whatever content they want,” Habib said.
That It appears to include spammy content. Habib makes the case that the video platforms’ algorithms, not QuickVidThese are the best positions for The quality of a video is determined. People who produce low-quality content are not allowed to be included. “are only damaging their own reputations.” The People will naturally be disincentives from creating mass spam campaigns if they suffer reputational damage with QuickVidHe says.
“If people don’t want to watch your video, then you won’t receive distribution on platforms like YouTube,” He added. “Producing low-quality content will also make people look at your channel in a negative light.”
But it’s instructive to look at ad agencies like FractlThe 2019 use of the term was. AI system called Grover generates An entire site of marketing materials — reputation is damned. In an Interview with The Verge, Fractl Partner Kristin Tynski said she saw generative possibilities. AI Enabling “a massive tsunami of computer-generated content across every niche imaginable.”
Tik is one of the most popular video-sharing sites. Tok and YouTube haven’t had to Take action with the Moderation of massive amounts of -generated material. Deepfakes — synthetic videos that replace an existing person with someone else’s likeness — began YouTube as one of the most popular platforms. It was created by tools that made it easier for deep-faked footage to be uploaded. to produce. But These videos are different from any convincing deepfakes. QuickVid creates aren’t obviously AI-generated in some way.
Google Search’s policy on AI-generated text might be a preview of what’s to Please visit the video domain. Google doesn’t treat synthetic text differently from human-written text when it concerns search rankings but Takes action on content that’s “intended to manipulate search rankings and not help users.” That This includes content that is gathered from different web pages or combined. “[doesn’t] add sufficient value” Both contents created by purely automated processes and content generated manually, both of which may be applicable to QuickVid.
In other words: if they become a major success, they may not be banned entirely from platforms. However, it is possible to ban -generated videos from platforms if they do so. That isn’t likely to Experts who believe platforms like Tik can cause panic should be dispelledTok is becoming a new home for False advertising videos, but — as Habib said during the interview — “there is no stopping the generative AI revolution.”