Hi there.

I am currently on vacation at the moment visiting family. While I am away, I hope you enjoy this newsletter on synthetic voice which follows Schibsted’s Lena Pedersen’s excellent Webinar. This comes as part of the research I have been doing in audio, which will culminate in a report coming out later this month.

I’m back next week and would love to hear any comments or feedback you have. I’m at Jodie.hopperton@INMA.org or you can get me on Slack.

Artifact has quietly launched audio articles, some read by Snoop Dogg

As you probably know by now, I am following Artifact — the low-key news product from the founders of instagram — with huge interest. 

I say low-key because I have not seen a single piece of paid marketing from them, and their updates are clean, simple, and always fun. I have signed up to be a beta tester, and a few weeks ago I got an update in TestFlight. What I found was an introduction to their synthetic voices. And, in Artifacts style, it is thoughtful, low-key, and immensely fun to play around with. 

Here is the e-mail announcement:

Below is the user flow. In most articles, but not all, a play button appears on the bottom navigation. Once that is clicked in, a small bar appears at the bottom playing the article in any settings you have preselected (such as voice and speed). In this half window, you are able to skip forward and back, as well as change individual settings.  

There are two settings that can be changed. 

First, let’s look at the voice itself. Looking at the screenshots below, once you click on the small head, you have the option to choose which voice you would like. There are five featured voices and 30 voices total.

I assume that there is some kind of deal that has been done with both Snoop Dogg and Gwyneth Paltrow to have these names. But what I also think is interesting is the range of voices and accents that Artifact presents. You’ll notice on the screenshots that there are offerings from the UK, USA, Nigeria, Australia, and South Africa.

Last but not least is the speed. 

As you can see from the screenshot, Artifact has decided to use a sliding scale rather than a standard presets. It goes much higher than usual: to 4.5X. Also note that they have named the speeds and added productivity levels — an interesting quirk.

Date for the dairy: October 23, INMA Silicon Valley study tour

Silicon Valley is back, with the Bay Area topping the charts, by far, for the concentration of AI. 

And that is why we are heading there in October, looking at how AI will likely change consumer habits and, therefore, our business. In addition, our study tour will look at the tools we may want to use to improve our own business efficiency. 

I hope you can join us. More here.

I created a synthetic version of my own voice in 15 minutes — and it’s not bad

I’m signed up to Apple’s public beta for iOS 17. I don’t usually sign up to public betas because I often find them buggy and more hassle than they are worth. But this update was too good to miss. I won’t bore you with the many new things, except for one which, rather strangely is hidden away. 

This one thing is the ability to create a synthetic version of your own voice. Obviously I had to try it.  

I say that it’s tucked away because it’s not a feature I have seen promoted and I actually struggled to find it. It’s tucked into “accessibility” under settings, where I had to scroll down to “personal voice” under speech. Once there I tapped on “Create a Personal Voice” to find the instructions. See the series of screenshots below:

I won’t lie: Even for someone who speaks as much as I do, reading 150 lines was a touch tedious. The device records continuously and recognises a line when you start speaking. You can pause anytime and/or redo a phrase.

And then voila, it starts processing. And you wait, which is not one of my virtues. 

The phone needs to be locked and connected to power. Overnight didn’t quite do it, so I kept my phone on as much as I could the next day, checking intermittently. The following morning it was done. 

And I am pretty impressed.  It’s clearly not the same and a little robotic, but this is a beta without much training and I think the intonation is pretty good. Take a listen for yourself. 

I should point that the purpose of this synthetic voice isn’t to help people with ALS. It’s not — at least at the moment — meant for commercial use in any way. All data is kept locally on the iPhone. 

Here’s the announcement:

I’ll be keeping an eye on how this develops. Please let me know how your voice pans out as you make your own.

About this newsletter 

Today’s newsletter is written by Jodie Hopperton, based in Los Angeles and lead for the INMA Product Initiative. Jodie will share research, case studies, and thought leadership on the topic of global news media product.

This newsletter is a public face of the Product Initiative by INMA, outlined here. E-mail Jodie at jodie.hopperton@inma.org with thoughts, suggestions, and questions. Sign up to our Slack channel.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .