> If there's nothing deceptive about this, the app should append "This message was sent by a bot" to what it sends.
I agree it's deceptive if the bot is making the final decision to send the message, and the message fails to make clear that it's been sent by a fully automated process.
Obvious alternative: the app could offer the user a pre-populated message. That way the system is doing more than a calendar would, but the action of sending the message is still up to the user. Looks like the app does support a variation of this:
> CommuniqAI will cleverly send text messages of your choosing to, and smartly prompt you to call and email, those that mean the most to you.
The intent of the app does appear to be to enable 'emotional fraud' though, as it seems they do aim to mislead the person being messaged (about it being an automated message). They would even want to support deceptively impersonating you for the purpose of replying, if only it were possible with current technology:
> Warning: CommuniqAI works well when used sparingly with simple messages—perhaps too well. While it’ll keep you in regular contact with those that mean the most to you, they’ll be inclined to reply back and, at this time, CommuniqAI is not intelligent enough to reply in context—so you’ll need to reply back yourself.
Here, works well is clearly referring to its ability to deceive someone into thinking the message was not sent automatically.
> Imagine getting a "How's life?" text from your friend when they're taking a nap in the same room.
A possibility, but more likely, imagine getting a How's life? text from someone you met up with yesterday.