OpenAI's Sora has left a lot of people impressed with its capabilities, but the AI startup says it's still very early days in the testing phase. OpenAI also has no "final plans" to release Sora officially as a product to consumers yet.
What Happened: Speaking at the Berkeley AI Research Lab, lead engineers who worked on OpenAI's Sora revealed a few key tidbits explaining where the video-generating AI Model currently stands.
According to them, OpenAI is not yet ready to release Sora, a product that end users can take for a spin. The Microsoft Corp.-backed MSFT AI startup has been talking about "red teaming" Sora for a while now, back when it announced the AI model in February.
More recently, OpenAI CTO Mira Murati also reinforced this in an interview on Wednesday.
"Right now, we’re in the discovery mode and we haven’t figured out exactly where all the limitations are."
Some of the challenges OpenAI currently faces with Sora is testing the edge cases. This can involve everything from simple videos with errors like inaccurate features like an extra finger or the mouth shaped weirdly to some egregious ones involving illicit or harmful content.
In other cases, Sora could generate historically inaccurate videos – rival Alphabet Inc.'s GOOG GOOGL Google Gemini has been embroiled in a similar controversy involving its founders as well as U.S.'s founding fathers.
OpenAI also doesn't have clarity on issues like nudity – Murati says the company is “working with artists” to understand “what’s useful and what level of flexibility the tool should provide.”
Beyond this, OpenAI expects to use the same guardrails for Sora that it used for Dall-E, which should make the process somewhat easier.
Why It Matters: OpenAI's Sora has impressive capabilities, allowing it to generate 60-second videos based on simple text prompts. It can also enhance existing videos by adding or removing frames based on its understanding of the contextual situation of a given scene in the physical world.
However, its misuse also presents a greater problem at a time when AI-generated deepfake images are wreaking havoc, with victims being minor girls to pop star Taylor Swift. Microsoft CEO Satya Nadella called it “alarming and terrible,” promising “quick action” against the culprits.
Check out more of Benzinga’s Consumer Tech coverage by following this link.
Photo courtesy: Shutterstock
© 2024 Benzinga.com. Benzinga does not provide investment advice. All rights reserved.
Trade confidently with insights and alerts from analyst ratings, free reports and breaking news that affects the stocks you care about.
