Really, keeping a tech company on a mission statement is almost unfair. From Google's “not evil” to WeWork's “raising global awareness,” mission statements are usually written in corporate adolescents. In a troublesome moment when their dreams stretch to the horizon, all venture capitalists are smiling, and no one hears the term “fiduciary responsibility.” It's like judging someone based on the feelings expressed behind the scenes in a high school yearbook.
But Openai, you're pushing it.
You will read these words when you go to the About Page for your company. It first appeared in the 2018 Charter and appeared three years after its founding. To put it simply, as demonstrated by some of the coverage of the perfect company in the future, it's not something they've always lived through. (Disclosure: Vox Media is one of several publishers that have signed a partnership agreement with Openai. Our report is editorially independent.)
Look, if you asked me what my mission statement was when I was 3, it would have probably been “becoming the first NBA player to land on Mars.” We don't always achieve what we're trying to do. Change your priorities, you won't grow to 7 feet 2. I realized that space is scary.
However, the latest product, AI-generated video social network SORA 2-, may have set the highest ever record of the largest distance between mission statements and actual work.
Infinite serving of AI slops
The best way to understand SORA 2 is probably to marry the worst aspect of a large language model like ChatGpt: the powerful ability to enthralle the user.
It's like taking heroin and mixing it… I don't know, is there a drug that is very addictive and makes a loose render in front of the screen and subtracts dozens of IQ points? Heroin, more, more heroin? I'm not sure if you have any drug experience to answer this question.
The fundamental issues posed by video generated by amazingly authentic AI are clear when SORA 2 was launched earlier this week, and are almost instantly realized.
Sign up here to explore the big complex problems the world is facing and the most efficient ways to solve them. It was sent twice a week.
Be copyright infringed. One of the first Sora 2 videos I met was a visit to Rick and Morty's SpongeBob Squareporgy, which was fully rendered. Openai Set SORA 2 allows copyrighted materials by default and holds intellectual property owners responsible for actively asking Openai. It would – but not before they grabbed the buzz of all of Rick and Morty clone social networks, as they did during the short craze for gifting your photos. (I know you still have them somewhere.)
Next is the deep fake. One of the killer features in SORA 2 is to upload an image to the app and then pop it into a video generated by the desired AI, or allow a friend to do so, or if your personal alignment is mixed and neutral. Any Sora users will take advantage of it. As reported by the Washington Post, it took about five seconds for the clip to begin generating fake police bodycam footage. Real people dressed up to the Nazi general. Very realistic, but fake footage of historical events. And yes, Openai CEO Sam Altman Shallifting.
What this means is, at the moment when the US president was posting a deepfake video generated to AI-generated Democrat minority leader with sombreros and mustaches, Open had just handed over someone who had at least a SORA 2 access code, and then passed on a completely realistic fake video with a push of buttons. and We've created a social network like Tiktok that can be shared. It's good that it includes rules to ban guardrails to stop fraud, fraud, nude and graphic violence, but speaking of a safe automatic weapon is completely harmless.
For some reason, I doubt that either of them “benefits all humanity.” But that will almost certainly benefit Openai's revenues. This is when the company is rated $500 billion (which beat SpaceX equality), when the noise about the AI bubble is no longer possible to ignore. (Quick, Sora 2 generates a video of what will be left in the US economy when the only industry driving it becomes Ka-blooey!)
So. What does Sam Altman think about all of this? Luckily, Sam has been blogging since at least 2013, coming back when he was pondering the possibility of alien existence. Sora 2, he wrote, is a moment of “Chat for Creativity,” which could lead to a “Cambrian explosion,” and “may greatly improve the quality of art and entertainment.”
That's… do I guess? Certainly, the post-Cambrian period led to rather strange creatures like Talimontherum, Or “Tully Monster” is a stem-eyed creature with a grabber hose for the mouth that looks like what God would get if he could get the Sora 2 access code. If the Mindless Deepfake Remix Machine, which is SORA 2, is something that is considered creativity in the future, then please get PaperClip Maximizer AI.
But perhaps the worst part of SORA 2, and a similar AI slop generator in Meta – is to mask AI works that can actually benefit all humanity. The same week Openai unleashed Sora 2, and many exiles from major AI companies announced the launch of a regular lab, a startup aimed at using artificial intelligence to accelerate discoveries in physics, chemistry and other science fields. You know what we can actually use.
Maybe I'm too simple and struggling with AI companies to the highest highly compensated mind of my generation, and can do this rather than do this. After all, Openai is business. (Or nonprofit organization? Or a public benefits corporation? Honestly, it's a bit in the air at the moment.) That follows what the market dictates. So, the final line of defense is to stand up and say, “No, I won't eat your AI slops” to us, as users of the world.
Of course, in the last check, Sora was third on the iPhone app chart. We are all screwed.
