Artificial intelligence does not stop growing in uses and, from what we can see, Apple has not wanted to be left out of this new technological revolution. Just yesterday, for example, we told you about Microsoft’s ambitious plans to integrate ChatGPT into its Bing search engine, and recently we have seen many other examples of common activities for users in which, in one way or another, artificial intelligence is gaining ground. .
Many of these new uses are directly related to creating content from scratch, based solely on user-generated input, which the algorithm must interpret to provide the appropriate response. However, we have also seen other AI that do not start from scratch. What they do is start from a broader set of information to interpret it and offer a result related to it. Explained like this, I know it may sound a bit complex, but I’m sure you’ll understand it better if we talk about solutions like GANverse3D and Instant NeRF, both from NVIDIA, which feed on two-dimensional images to create realistic three-dimensional replicas of them. In these cases, you don’t start from scratch, what you do is ask the AI to interpret a specific content to generate a related output.
Such is the case, as we can read in The Verge, of what they have already begun to test at Apple, an AI to generate audiobooks from the original books. I know that many will be thinking of a text-to-speech system similar to the ones we have known for years, such as the more than popular Loquendo, but as we can see in the two samples published by Apple here, establishing said parallelism is like comparing a scooter with a Ferrari. I know it may sound too categorical, but you will understand me better if you listen to some of the samples on the Apple website.
With this new feature, currently only available in English and for a very, very limited selection of literary genres, Apple intends to offer writers and publishers a way to publish their works in audiobook format without incurring the high costs. economic and time that it causes to do it in the conventional way.
Apple poses some limitations, yes. In addition to the already mentioned one about the literary genres with which it can be used (the company is already working on new voices that will be used in other genres), the most logical and foreseeable is that the audiobooks generated with this system can only be distributed to through your audiobook store, with the sole exception of public and/or academic libraries. In other words, we will not be able to see them on Spotify or, even less, on Audible, since Amazon’s service prohibits recordings generated in this way, it only accepts narrations made by humans.
