I swear, most companies are just trying to take Dall-E and ChatGPT and trying to shoehorn them into being products that are useful beyond their original scope. They aren’t modifying the technology behind those products to make something novel because they don’t actually know how they work.
Generating fake recipes and fake images of recipes isn’t anything that anyone wants. It’s not a product that solves a problem or provides a new convenience. A recipe that hasn’t been tested by a human has no value.
Companies don't see AI as a technology to help improve things, they see it as another way to cut employees and raise profit margins. It's just a part of the greater issue in publicly traded companies of quarterly profits above all else. It's unsustainable.
I think it's résumé-driven development taken to the extreme. Someone at Instacart just added a line to their résumé: "Integrated generative AI into core product, increased media coverage of company by 300%." Now they just have to cross their fingers and hope their prospective employers don't read Business Insider. (Or use Instacart, I guess.)
I hope that our industry eventually revisits (or arrives at) the idea of doing worthwhile things and doing them well. All the software job descriptions these days say that you should know Docker and Kubernetes and AWS Lambda and Terraform and MongoDB and Cassandra and Cloudflare and GraphQL and so on.... Meanwhile, their Facebook-for-dogs site gets ten visitors a day, and it runs dog-slow (pun intended) because their ex-FAANG dev lead wrote code that is bubble-sorting a billion items in main memory.
The media seems to have decided that ai generated stuff is bad, but can't seem to decide if it's because the outputs are so good it will destroy creative jobs or if it's because the outputs are so bad and unappealing.
You just make it sound like automatically generating content by digesting previously made content and creating a predictive model inherently leads to bad results. And I agree.
18 comments
[ 2.4 ms ] story [ 56.0 ms ] threadhttps://www.reddit.com/r/montreal/comments/18n0657/maison_du...
The menu with the fake food is still up if you want judge for yourself:
https://maisonduroti.com/collections/pret-a-manger?page=1
I swear, most companies are just trying to take Dall-E and ChatGPT and trying to shoehorn them into being products that are useful beyond their original scope. They aren’t modifying the technology behind those products to make something novel because they don’t actually know how they work.
Generating fake recipes and fake images of recipes isn’t anything that anyone wants. It’s not a product that solves a problem or provides a new convenience. A recipe that hasn’t been tested by a human has no value.
But i do agree with you
It seems to me that most companies using AI are raising costs and adding trivial value to their offering.
I hope that our industry eventually revisits (or arrives at) the idea of doing worthwhile things and doing them well. All the software job descriptions these days say that you should know Docker and Kubernetes and AWS Lambda and Terraform and MongoDB and Cassandra and Cloudflare and GraphQL and so on.... Meanwhile, their Facebook-for-dogs site gets ten visitors a day, and it runs dog-slow (pun intended) because their ex-FAANG dev lead wrote code that is bubble-sorting a billion items in main memory.