How you’ll be able to create truly insane videos from text thanks to OpenAI’s Sora tool
Author: Connor Jewiss
Published on: 2024-12-09 22:14:00
Source: Stuff
Introduction
We’ve had a pretty fast-paced year for AI tools, especially from OpenAI. They’re the makers of ChatGPT, DALL-E, and have arguably inspired other AI chatbots to hit the shelves. They’re also the ones working on self-aware AI, but we’ll leave that for another day, as I don’t want any existential dread this weekend. Instead, OpenAI’s latest tool is going to replace the best cameras – it can create entire videos just from your text prompt. Or at least that’s the idea.
So, what’s the deal with Sora? It’s an AI model that’s been trained to understand your text instructions and then, like a digital Spielberg, create them as full video scenes. Want a stylish lady strutting down Tokyo lit by neon glow? Or perhaps a close encounter with woolly mammoths in a snowy landscape? Sora’s got you covered, generating these scenes with a flair that’ll make you question reality.
The examples OpenAI flaunts range from the stunningly realistic to the delightfully whimsical. We’re talking, a cinematic journey from a drone’s view of Big Sur’s rugged cliffs to an animated, fluffy monster pondering the mysteries of a melting candle. Now that it’s rolling out, MKBHD generated a few example videos as well. They’re not perfect – if you look closely, you’ll be able to nitpick mistakes. But it’s impressive to say it’s completely made up.
For example, it can’t handle text very well, and it doesn’t understand what motion different objects or creatures should take. It doesn’t know how humans or animals have to walk. But, it does handle high frame rates very well, with video not blurry at all.
Sora isn’t just playing around, it’s a real thing. OpenAI previously rolled it out to experts to poke, prod, and assess for any potential risks or harms. Visual artists, designers, and filmmakers could also get their creative hands on it. Now, it’s available for anyone in the public to use (unless you’re in Europe or the UK because of regulations). You’ll need to be subscribed to the recently released $200/month ChatGPT Pro tier.
And because we’re living in the age of “pics or it didn’t happen,” safety is a top priority. OpenAI has allegedly geared up safety nets, like detection classifiers to spot Sora-generated videos and robust image classifiers to keep the content in check. It remains to be seen how well they’ll actually work, however.
Sora represents a thrilling new chapter in AI’s evolution. New tools (and products) are coming thick and fast. If you remember the Will Smith eating spaghetti AI video, you can see how far we’ve come. So, whether you’re a filmmaker dreaming, a designer seeking inspiration, or just someone who loves to marvel at the sheer coolness of tech, Sora is a glimpse into a future where the lines between creation, imagination, and AI blur into something truly extraordinary. And I can’t wait to see how it turns out.
Top Features
We’ve had a pretty fast-paced year for AI tools, especially from OpenAI. They’re the makers of ChatGPT, DALL-E, and have arguably inspired other AI chatbots to hit the shelves. They’re also the ones working on self-aware AI, but we’ll leave that for another day, as I don’t want any existential dread this weekend. Instead, OpenAI’s latest tool is going to replace the best cameras – it can create entire videos just from your text prompt. Or at least that’s the idea.
So, what’s the deal with Sora? It’s an AI model that’s been trained to understand your text instructions and then, like a digital Spielberg, create them as full video scenes. Want a stylish lady strutting down Tokyo lit by neon glow? Or perhaps a close encounter with woolly mammoths in a snowy landscape? Sora’s got you covered, generating these scenes with a flair that’ll make you question reality.
The examples OpenAI flaunts range from the stunningly realistic to the delightfully whimsical. We’re talking, a cinematic journey from a drone’s view of Big Sur’s rugged cliffs to an animated, fluffy monster pondering the mysteries of a melting candle. Now that it’s rolling out, MKBHD generated a few example videos as well. They’re not perfect – if you look closely, you’ll be able to nitpick mistakes. But it’s impressive to say it’s completely made up.
For example, it can’t handle text very well, and it doesn’t understand what motion different objects or creatures should take. It doesn’t know how humans or animals have to walk. But, it does handle high frame rates very well, with video not blurry at all.
Sora isn’t just playing around, it’s a real thing. OpenAI previously rolled it out to experts to poke, prod, and assess for any potential risks or harms. Visual artists, designers, and filmmakers could also get their creative hands on it. Now, it’s available for anyone in the public to use (unless you’re in Europe or the UK because of regulations). You’ll need to be subscribed to the recently released $200/month ChatGPT Pro tier.
And because we’re living in the age of “pics or it didn’t happen,” safety is a top priority. OpenAI has allegedly geared up safety nets, like detection classifiers to spot Sora-generated videos and robust image classifiers to keep the content in check. It remains to be seen how well they’ll actually work, however.
Sora represents a thrilling new chapter in AI’s evolution. New tools (and products) are coming thick and fast. If you remember the Will Smith eating spaghetti AI video, you can see how far we’ve come. So, whether you’re a filmmaker dreaming, a designer seeking inspiration, or just someone who loves to marvel at the sheer coolness of tech, Sora is a glimpse into a future where the lines between creation, imagination, and AI blur into something truly extraordinary. And I can’t wait to see how it turns out.
Pros and Cons
We’ve had a pretty fast-paced year for AI tools, especially from OpenAI. They’re the makers of ChatGPT, DALL-E, and have arguably inspired other AI chatbots to hit the shelves. They’re also the ones working on self-aware AI, but we’ll leave that for another day, as I don’t want any existential dread this weekend. Instead, OpenAI’s latest tool is going to replace the best cameras – it can create entire videos just from your text prompt. Or at least that’s the idea.
So, what’s the deal with Sora? It’s an AI model that’s been trained to understand your text instructions and then, like a digital Spielberg, create them as full video scenes. Want a stylish lady strutting down Tokyo lit by neon glow? Or perhaps a close encounter with woolly mammoths in a snowy landscape? Sora’s got you covered, generating these scenes with a flair that’ll make you question reality.
The examples OpenAI flaunts range from the stunningly realistic to the delightfully whimsical. We’re talking, a cinematic journey from a drone’s view of Big Sur’s rugged cliffs to an animated, fluffy monster pondering the mysteries of a melting candle. Now that it’s rolling out, MKBHD generated a few example videos as well. They’re not perfect – if you look closely, you’ll be able to nitpick mistakes. But it’s impressive to say it’s completely made up.
For example, it can’t handle text very well, and it doesn’t understand what motion different objects or creatures should take. It doesn’t know how humans or animals have to walk. But, it does handle high frame rates very well, with video not blurry at all.
Sora isn’t just playing around, it’s a real thing. OpenAI previously rolled it out to experts to poke, prod, and assess for any potential risks or harms. Visual artists, designers, and filmmakers could also get their creative hands on it. Now, it’s available for anyone in the public to use (unless you’re in Europe or the UK because of regulations). You’ll need to be subscribed to the recently released $200/month ChatGPT Pro tier.
And because we’re living in the age of “pics or it didn’t happen,” safety is a top priority. OpenAI has allegedly geared up safety nets, like detection classifiers to spot Sora-generated videos and robust image classifiers to keep the content in check. It remains to be seen how well they’ll actually work, however.
Sora represents a thrilling new chapter in AI’s evolution. New tools (and products) are coming thick and fast. If you remember the Will Smith eating spaghetti AI video, you can see how far we’ve come. So, whether you’re a filmmaker dreaming, a designer seeking inspiration, or just someone who loves to marvel at the sheer coolness of tech, Sora is a glimpse into a future where the lines between creation, imagination, and AI blur into something truly extraordinary. And I can’t wait to see how it turns out.
User Reviews
We’ve had a pretty fast-paced year for AI tools, especially from OpenAI. They’re the makers of ChatGPT, DALL-E, and have arguably inspired other AI chatbots to hit the shelves. They’re also the ones working on self-aware AI, but we’ll leave that for another day, as I don’t want any existential dread this weekend. Instead, OpenAI’s latest tool is going to replace the best cameras – it can create entire videos just from your text prompt. Or at least that’s the idea.
So, what’s the deal with Sora? It’s an AI model that’s been trained to understand your text instructions and then, like a digital Spielberg, create them as full video scenes. Want a stylish lady strutting down Tokyo lit by neon glow? Or perhaps a close encounter with woolly mammoths in a snowy landscape? Sora’s got you covered, generating these scenes with a flair that’ll make you question reality.
The examples OpenAI flaunts range from the stunningly realistic to the delightfully whimsical. We’re talking, a cinematic journey from a drone’s view of Big Sur’s rugged cliffs to an animated, fluffy monster pondering the mysteries of a melting candle. Now that it’s rolling out, MKBHD generated a few example videos as well. They’re not perfect – if you look closely, you’ll be able to nitpick mistakes. But it’s impressive to say it’s completely made up.
For example, it can’t handle text very well, and it doesn’t understand what motion different objects or creatures should take. It doesn’t know how humans or animals have to walk. But, it does handle high frame rates very well, with video not blurry at all.
Sora isn’t just playing around, it’s a real thing. OpenAI previously rolled it out to experts to poke, prod, and assess for any potential risks or harms. Visual artists, designers, and filmmakers could also get their creative hands on it. Now, it’s available for anyone in the public to use (unless you’re in Europe or the UK because of regulations). You’ll need to be subscribed to the recently released $200/month ChatGPT Pro tier.
And because we’re living in the age of “pics or it didn’t happen,” safety is a top priority. OpenAI has allegedly geared up safety nets, like detection classifiers to spot Sora-generated videos and robust image classifiers to keep the content in check. It remains to be seen how well they’ll actually work, however.
Sora represents a thrilling new chapter in AI’s evolution. New tools (and products) are coming thick and fast. If you remember the Will Smith eating spaghetti AI video, you can see how far we’ve come. So, whether you’re a filmmaker dreaming, a designer seeking inspiration, or just someone who loves to marvel at the sheer coolness of tech, Sora is a glimpse into a future where the lines between creation, imagination, and AI blur into something truly extraordinary. And I can’t wait to see how it turns out.