As deepfakes proliferate, OpenAI is refining the tech used to clone voices — but the company insists it’s doing so responsibly. Today marks the preview debut of OpenAI’s Voice Engine, an expansion of the company’s existing text-to-speech API. Under development for about two years, Voice Engine allows users to upload any 15-second voice sample to generate a synthetic copy of that voice.
Source: OpenAI built a voice cloning tool, but you can’t use it… yet | TechCrunch
Just as in film and TV, only more so, AI represents a gathering storm for video game actors. Some studios are experimenting with tools that can clone voices, alter voices and generate audio from text. In interactive, multi-choice games, this can generate a potentially endless number of characters and conversations – and is far more efficient than asking performers to record huge quantities of dialogue.
Source: ‘It’s very easy to steal someone’s voice’: how AI is affecting video game actors
The responsible utilization of data in AI development is a critical issue, and a select group of visual AI platforms are leading by example, demonstrating that AI can flourish while respecting intellectual property rights and fairly compensating creators. At the forefront of this movement are four pioneering platforms that have taken proactive steps to ensure their AI models are trained exclusively on properly licensed content.
Source: 4 Visual AI Platforms Proving That AI Can Thrive with Properly Licensed Content – Kaptur
With the breakneck speed at which artificial intelligence has been progressing, there have undoubtedly been some stumbles. One of the biggest issues has been the use of copyrighted materials to train AI models as well as images that may be inappropriate or raise privacy issues. But a technique referred to as “machine unlearning” developed by researchers at the University of Texas Austin may offer a solution to those concerns.
Source: ‘Machine Unlearning’ May Be the Solution to Problematic AI Data
Shot Tower Capital has engaged in an independent audit of Hipgnosis Songs Fund’s assets, lowering the valuation of its assets by 26%. This follows a report on March 18 that revealed a double-counting error in Hipgnosis’ accrued revenue, with the board announcing a 7.6% reduction to its operative net asset value (NAV). The report also finds that Hipgnosis Songs Fund’s acquisition underwriting was “below music industry standards.”
Source: Hipgnosis Songs Fund Audit Says Fund ‘Failed to Perform’
SoundExchange, the entity tasked with collecting and distributing royalties for the use of recordings on non-interactive digital services, just recently unveiled this latest milestone via a brief release. SoundExchange towards the beginning of 2024 indicated that it had distributed over $1 billion during 2023, up from about $959 million in 2022. Bearing in mind this growth the $12 billion milestone will presumably arrive within the next year or so.
Source: SoundExchange Announces $11 Billion Distribution Milestone
Last year was a very good one for music streaming: Audio streams increased 22.3% year over year, according to analytics firm Luminate, accounting for more than two-thirds of global music revenue. It was also the year when the music industry finally started to address the fact that the streaming model doesn’t benefit artists.
Source: Music streaming royalties are terrible—these companies may have a fix
Amazon said Wednesday it is pouring an additional $2.75 billion into Anthropic, bringing its total investment in the artificial intelligence startup to $4 billion. Amazon will maintain a minority stake in San Francisco-based Anthropic, a rival of ChatGPT maker OpenAI. The Seattle-based tech giant made an initial investment of $1.25 billion in Anthropic in September and indicated then it had plans to invest up to $4 billion.
Source: Amazon pours an additional $2.75 billion into AI startup Anthropic
Fresh off scaring the bejeezus out of many in Hollywood with demos of its text-to-video generator Sora, OpenAI now wants in. According to Bloomberg, top executives at the generative AI developer will hold a round of meetings this week with a number of film studios and Hollywood honchos to discuss what Sora can do for them.
We’ve discussed here before why everything that can be made with AI, will be in Hollywood. So it is no great surprise that studio folks would take the meetings. But appearing to get cozy with Sora right now carries significant risk for the studios.
Generative AI was recently at the center of extensive labor unrest in Hollywood that cost the studios the better part of a year’s worth of production. As a result of that unrest, they are also now bound by collective bargaining agreements with writers and actors that circumscribe what they can do unilaterally with tools like Sora.
Overall, 54% of Americans say artificial intelligence programs that generate text and images, like ChatGPT and DALL-E, need to credit the sources they rely on to produce their responses. A much smaller share (14%) says the programs don’t need to credit sources, according to a new Pew Research Center survey. About a third say they’re not sure on this question.
Source: Many Americans think generative AI programs should credit the sources they rely on