Chat GPT’s Voice Cloning AI App Is Too Dangerous to Release. OpenAI’s Voice Engine is an AI model that generates speech mimicking a speaker’s voice from a 15-second audio sample. It can be used in applications like educational aids, translation, and support for non-verbal individuals. What could go wrong? Well… at Schwab, my voice is my password. Until now. Accordingly, OpenAI is employing a cautious approach to deployment due to potential misuse. But they are still allowing a small number of businesses to test the new tool. Why are they testing this if it’s too dangerous to release? What really worries me: if Open AI can do this, others can. And they won’t issue a press release.
Higgsfield AI launches Diffuse. The company is led by the former head of AI at Snap, Alex Mashrabov. The Diffuse mobile app generates videos from text prompts and can use selfies to generate clips starring that person. It’s free for now, but the plan is to charge social media content creators and marketers fees for generating the short form videos. While Sora is being tested by professionals, it is computational expensive, so it will certainly be priced for enterprises, not individuals. Higgfield’s generative models were developed by a 16-person team in less than nine months. Mashrabov sold his previous startup, AI Factory, for $166 million. Funding for Higgfield comes from an $8 million seed round led by Menlo Ventures.
Bitkraft Ventures Raises $275 Million For Third Interactive Fund. This is the Venture Capital firm’s 3rd Fund. Bitkraft Venture Fund 3 will invest in studios, platforms, and technology in the gaming and interactive media categories at the seed and Series A stages. Their portfolio includes Frost Giant, Anzu, Carry1st, InWorld, Voicemod, Immutable, and Karate Combat. Bitkraft operates six venture funds with 130 portfolio companies.
Meta’s Reality Labs Celebrates 10th anniversary of the Oculus VR acquisition. Meta has spent over ten billion dollars on its XR efforts since them. At this point, it’s safe to say it’s cost them over $100 Billion. CNet’s XR editor took a trip back in time with CTO Andrew Bosworth, who oversees the effort. Bos says displays in regular glasses are still a challenge, but adding AI to wearables like smartglasses will open up a new way of thinking about how XR will evolve.
Apple’s Spatial Personas for Vision Pro. Technically, Personas have been available since the headset launched on February 2, but they’ve received a significant upgrade this week that everyone is talking about. Your persona is created by scanning your face in your headset, but previously it existed inside a frame. While the Persona only has a head, shoulders, and hands, because of the way face tracking captures expressions it creates a sense of presence. When a persona in a conversation moves around, the headset translates the perspective as if they are there.
This column, once called “This Week in XR,” is also a podcast hosted by author Charlie Fink, and Ted Schilowitz, former studio executive and co-founder of Red Camera, and Rony Abovitz, founder of Magic Leap. This week our guest is Maureen Fan, CEO of Baobab Studios. We can be found on Spotify, iTunes, and YouTube.
What We’re Reading
How AI May Actually End The World. (Shelly Palmer Blog)
Seven Best Sora Videos (Tom’s Guide)
