Just days after OpenAI unveiled its latest AI video tool, Sora 2, the accompanying social app has ignited a frenzy over invite codes. Users are scrambling for access, with some reselling spots on eBay for upwards of $50 a pop. The app, designed for quick AI-generated clips complete with audio, promises to let anyone create and share short videos—like a TikTok but powered entirely by artificial intelligence. Indeed, the hype is palpable, as early adopters flood social media with demos ranging from whimsical animations to eerily realistic scenes.
One standout example? A flurry of Pikachu-themed videos that have taken off online. Imagine the yellow Pokémon zipping through urban streets or battling in fantastical realms, all whipped up from simple text prompts in under a minute. These clips showcase Sora 2's improved physics and instruction-following, making movements feel almost lifelike. Yet, not all creations are so innocent. Reports are emerging of deepfake videos featuring OpenAI CEO Sam Altman himself—depicting him in absurd situations, like shoplifting gadgets or dancing awkwardly at events. Such content raises immediate red flags about misinformation and ethical boundaries in AI-driven media.
Moreover, the invite-only rollout has created bottlenecks. OpenAI's help center outlines the process: download the iOS app, sign in with your ChatGPT credentials, and punch in a code if you snag one. However, with demand surging—Reddit threads alone boast thousands of comments begging for shares—many are left waiting. Critics point out that this exclusivity, while building buzz, exacerbates access issues, especially in regions facing rollout delays. OpenAI insists the phased approach ensures safety, with built-in age checks and content filters during onboarding.
Still, the rapid proliferation of these tools prompts questions. How soon before viral fakes blur the line between fun and fabrication? As more users gain entry, the app's potential to reshape social sharing hangs in the balance.