Step aside: Let AI do the talking

On Android, the takeover of SMS is almost complete. For the longest time, I resisted letting Google manage my messages. I don’t recall exactly when I gave in. Maybe it was during one of those countless pop-ups interrupting my conversations, prompting me to agree just so I could continue texting. Now, I get slick animations when someone “likes” a message, see real-time indicators when people are typing, and enjoy the perks of using the internet to send messages. Along with that, however, I get auto-generated replies.

These AI-suggested replies feel increasingly invasive. No matter what message I receive, there’s a ready-made response waiting for me to tap: a cheerful “Sounds good!” a quick laugh, or even a sympathetic “Oh no!” accompanied by a frowny face emoji. At this point, the AI can handle entire conversations if I let it.

When these features first appeared on Gmail, they were met with backlash. People weren’t just creeped out by the idea of Google “reading” their emails to generate responses; some suggestions were outright inappropriate. For example, “I love you” occasionally popped up in professional work emails. Despite these early hiccups, auto-suggestions have now become the norm, or at least something we’ve grudgingly accepted as “the future.”

Today, even keyboard autocomplete is powered by AI, offering to complete entire sentences instead of just predicting the next word. Sometimes, my fat fingers accidentally select a suggestion, and before I know it, I’ve sent a reply I didn’t even craft myself. At work, I’m encouraged to let AI generate my email responses. On Slack, I get nudged to send a laughing emoji at my coworkers’ jokes. How long until the system skips the suggestion step altogether and just sends the most probable response without asking me?

Here’s the kicker: on the receiving end, another AI will likely read that automated message, process it, and send back a reply of its own. Before you know it, we’ll have two machines talking to each other, stuck in an endless loop of emojis and canned responses. No human intervention. No real conversation. Just AI bouncing meaningless messages back and forth.

If two AIs are exchanging messages and no one is there to read them, is any information transferred?

It’s akin to the old thought experiment: if a tree falls in the forest and no one is around to hear it, does it make a sound? When machines communicate in place of humans, is there any real communication happening?

The goal of conversation is to connect people, but it feels like we’re being written out of the equation. Companies are steadily replacing human interaction with algorithmic efficiency, leaving us as mere spectators of our own messages. We’re at risk of losing something essential. A fundamental part of what makes communication meaningful: us.