Best Amazon Prime Day deals: Mumsnet favourites

Best Amazon Prime Day deals:
Mumsnet favourites

Shop now

Please or to access all these features

Chat

Join the discussion and chat with other Mumsnetters about everyday life, relationships and parenting.

Zoe and CGM and use of AI ethics

1 reply

StampedingWildebeest · 17/06/2026 18:47

I was looking at some random mapping diagram for work while also contemplating future of my work and pondered this.

I have a little more cash than I did before and was having a muse over the Zoe CGM last week, when I looked it no longer exists and it's now an App.

When the CGM was out, I read about people's experiences but ultimately thought I'd wait and see, as some people were saying it's a bit generic and I also had a nagging feeling about them essentially using the data to develop new products. And lo and behold that's what they did do, the data collected was used to develop an AI based product.

Does anyone feel there's something iffy about that ethically? You weren't getting as personalised a service as you thought (probably billed as being in testing). Ultimately they are free to abandon the model and build something else I guess and I presume the app offers more in terms or community. But that was built on the data provided by participants which begs a question over whether people should be compensated for training AI models. And will that be (can see how it's relevant to other things) an area of legal interest in years ahead.

Anyone have more knowledgeable thoughts than my ramblings?

OP posts:
TheSmallAssassin · 17/06/2026 23:16

I get where you are coming from, we always knew they were learning scientifically from us, how different is using our data to train ML models? I did wonder if this was the case when photo logging was pushed, but it does also reduce the friction for logging.

To be fair the price has dropped significantly (from £24.99 a month to £9.99 a month), but also human coaching has been ditched for AI (unsurprisingly personal guidance from nutritionists was expensive)

I read somewhere that when a Captcha asks you to identify motorbikes or traffic lights, you are actually training an ML model (for autonomous vehicles?) - I think it does need to be more transparent.

(The CGM was only ever for the first two weeks, and Zoe only used the data from the first two days, IIRC, to determine your glucose response which fed into your personalised food scores)

New posts on this thread. Refresh page