AI audio firm ElevenLabs‘ co-founder and chief govt Mati Staniszewski believes that AI fashions shall be commoditized over time, a revealing remark for a corporation targeted at the moment on constructing them.
Talking onstage on the iinfoai Disrupt 2025 convention on Tuesday, the ElevenLabs founder was discussing each his short-term and long-term views of the AI audio house.
Staniszewski mentioned that his firm’s researchers have been in a position to crack a number of the mannequin structure challenges, and this focus will proceed within the audio house for the subsequent 12 months or two.
“Over the long run, it’ll commoditize — over the subsequent couple of years,” Staniszewski mentioned. “Even when there’s variations — which I feel would be the reality for some voices, some languages — by itself, the variations shall be smaller.”
Requested why ElevenLabs would deal with constructing fashions if he believed they might be commoditized in time, Staniszewski defined that, within the brief time period, they had been nonetheless the “greatest benefit and the most important step change you possibly can have at the moment.”
As an example, if the AI voices or interactions don’t sound good, that’s nonetheless an issue that must be solved.
“The one strategy to clear up it’s… constructing the fashions your self, after which, over the long run, there shall be different gamers that may clear up that, too,” mentioned Staniszewski.
He additionally famous that these in search of dependable, scalable use circumstances would nonetheless probably use totally different fashions for various use circumstances.
Nonetheless, within the subsequent 12 months or two, Staniszewski mentioned that an growing variety of fashions will transfer into multi-modal or fused approaches.
“So, you’ll create audio and video on the identical time, or audio and LLMs on the identical time in a conversational setting,” he mentioned, pointing to Google’s Veo 3 for instance of what might be achieved when combining fashions collectively.
The founder mentioned that ElevenLabs plans to launch partnerships with different corporations and work with open supply applied sciences to see if the corporate can mix its audio experience with a number of the experience of different fashions.
For ElevenLabs, the aim is to deal with each mannequin constructing and functions to create long-term worth, he mentioned.
“The identical approach software program and {hardware} was the magic for Apple, we expect the product and AI would be the magic for the era of the perfect use circumstances,” he added.
