Apple: AI Narrated Audio Books

Good Morning AI Runners

Here's what we've got for you today:

• Apple: AI Narrated Audio Books

• GoogleAI: MUSE

Apple: AI Narrated Audio Books

Apple has created a collection of audiobooks narrated by artificial intelligence, which could potentially signal the end of human narrators in the industry.

This move could have significant implications for the audiobook market, which is valued at over $35 billion and saw sales increase by 25% last year, reaching over $1.5 billion.

The AI-narrated audiobooks can be found in the Books app by searching for "AI narration."

Apple worked with independent publishers, who were required to sign NDAs, on the project and agreed to cover production costs, while authors would receive royalties from sales.

Narration is currently offered in four voices in soprano and baritone categories: Madison and Jackson for romance and fiction, and Helena and Mitchell for self-development and non-fiction.

Apple claims that these voices have been trained in specific genres, but has not disclosed the training data used.

And apple are yet to announce how much they are charging publishers for the conversion of a book into an audiobook.

It remains to be seen how consumers will react to AI-narrated audiobooks and whether they will prefer them over those narrated by humans.

Some may appreciate the consistency and lack of mistakes in AI narrations, while others may prefer the inflection and emotion that a human narrator brings to a story.

This might be unrelated but it reminds me of an interesting point on the appeal of human narrators. My favorite podcast, the Founders Podcast, features David Senra reading biographies and autobiographies of successful founders and the thing I enjoy the most is the human-like summary and lessons as well as his reading style.

I wouldn't find the podcast as enjoyable if it were narrated by Apple's "Helena and Mitchell." This goes to show that the appeal of human narrators and the way they can bring inflection and emotion to a story is hard to beat.

P.S. If you are interested in listening to the podcast don't miss the John D. Rockefeller and Jay Gould episodes.

GoogleAI: Muse

Google has introduced a new AI tool called Muse, which is a text-to-image generator that is able to create high-quality images at record speed.

The tool, developed by GoogleAI, is faster and more efficient than many current models, including Stable Diffusion, Dalle-E 2, and even Google's own Parti.

In testing, Muse was able to produce images that matched competitors in quality, variety, and text alignment, but significantly faster, with a generation time of 1.3 seconds compared to Stable Diffusion's 3.7 seconds.

Muse is based on transformer technology while stable diffusion is based on UNet.

In simple terms, it is ChatGPT but for images.

Pic of the day:

The first day of OpenAI, seven years ago.

That's it from RuntheAI for today.

THANK YOU FOR READING AND SEE YOU TOMORROW, SUBSCRIBE TO STAY UPDATED!