“I tested Sinceerly by cold emailing 5 Fortune 500 CEOs. 4 CEOs replied. Of those replies, each was under 10 words. 2 replies had typos. One reply called me Larry (my name is Ben).”
“I tested Sinceerly by cold emailing 5 Fortune 500 CEOs. 4 CEOs replied. Of those replies, each was under 10 words. 2 replies had typos. One reply called me Larry (my name is Ben).”
Because it’s a service using cloud AI. Processing text uses tokens, tokens cost money, hence the subscription fee.
It’s a dumb app - deliberately so - but it costs money to run and a one-off payment won’t cut it if people start using it to modify hundreds or thousands of emails.
If it really needs AI, something this basic should be able to use a tiny AI model that can run locally. Google are working on building small models into Chrome for example (https://developer.chrome.com/docs/ai/built-in). It really doesn’t need a huge model.
Several mobile apps bundle an AI model with them. Samsung phones do a bunch of AI things on-device, including object and face recognition in the Gallery app for photos. There’s no reason an extension couldn’t do that.
does it need to use AI though?
That’s a good point.
Imagine just a script that randomly replacing various homophones like their, they’re and there.
Or has a percentage chance of swapping two letters because you types too fast.
Or just a table of common misspellings would be enough.
i was also thinking of replacing letters with ones close on the keyboard or removing letters