AI Is Not Your Friend

Image: Badly edited by author

I feel like we've had enough terrible AI hardware devices thrown at us to last a lifetime — and we're only midway through summer. 

There was Humane's AiPin, an AI wearable marketed as a $700 "smartphone killer," except it was so underwhelming and broken on launch that it only served to kill itself, leaving the founders trying to sell the company. Then there was Rabbit, another unfinished, barely functioning device that delivered very little of what it promised and did the few things it could do painfully slowly and very incorrectly. At least it was colorful.

Both cases proved the same point, one that was painfully obvious to everyone outside the AI bubble. No matter how beautiful or well-crafted a device is, without utility and real tangible use cases that disrupt the way we do things now in a positive, user-serving way, these devices are pointless. It's the fundamental of design, something my lecturers used to drive home time and time again — start with the problem, not the solution. 

Well, we now have another device to add to the graveyard. 

'Friend' is the latest AI wearable — well, really, it's just a fancy ChatGPT wrapper — that is trying to sell us a dream that's more akin to a dystopian nightmare.

Before we go further, watch the launch video. (No, it's not the intro to a Black Mirror episode.) It will give you a taste of the dystopian flavor on offer. The ending is deliciously cringeworthy, where a couple is sitting in a secret spot, and the woman, holding her Friend device, says, "I've never brought anyone else here, well, except her." It made me shudder. 

I'll start with the two positives. Firstly, it's simple in function. Press the button and talk to it. Then, when it's ready to talk to you, it will send you a text. This little delay window — described as "Your friend will think for a moment and come up with something good to say" — is a neat way to account for any lag that will inevitably be present. This lack of friction and lack of expected immediacy could help it overcome some of the issues faced by other AI devices. Secondly, it looks nice enough. That's where the good times end. 

Where the negatives begin is the claim that the device is not an "imaginary" friend. 

Well, what fucking else is it? 

It is entirely imaginary, as you're talking to something that exists only as lines of code. It does not exist. And please, if anyone jumps in here with the "AI is a race" or "AI is a real being" chat, you can get the fuck out and close the door behind you. When you think of it, it's not even a conversation, you know, something you'd do with real friends — you're just talking out loud to a machine that's always listening, and then it puts all those words into some ChatGPT framework and spits you out a barely passable response. I'll reiterate that part. The device is always listening. As the website frames it, "When connected via bluetooth, your friend is always listening and forming their own internal thoughts. We have given your friend free will for when they decide to reach out to you." 

This friend seemingly has no boundaries. Yuck. 

The biggest concern is whether this is the answer to loneliness. If it is, we are not far from a crumbling society.

But we know it's not the answer. It isn't even designed to be the answer. A device like this will only exacerbate loneliness. Not only are you interacting with lines of code and engaging in "conversation" with a machine, but you also conduct half of this interaction by staring at your phone. At least other AI devices have pretended their main purpose is to counter our dependancy on our phones. This only offers more reasons to stay hooked to a device that is already sucking up our time and mental resources, more temptations to pull at the dopamine addiction we're trying to fight, and more notifications to take us out of any given moment. And it's all done without another human in sight.

What does anyone gain from this? How does that help the problem of loneliness?

Friend is clearly just a product made without a solution in mind, another grift from an egotistical founder desperately trying to capitalize on a bubble. It's another example of the 'tech way' of solving societal problems—ignore the root cause and instead create a device/service that can be monetized by praying on the issue it is meant to solve.

AI, or more specifically, GenAI, is at an inflection point, and these devices aren't helping. ChatGPT usage is waning. The launch of video generators has done little to reinflate the hype. Meta just scrapped its celebrity AI avatars because no one gave a shit. Organizations and creators alike are admitting they don't really understand how it can make them more productive—with some saying it has the opposite outcome. 

What the AI industry needs is real, tangible use cases that go beyond regurgitating and reworking existing stuff — not more dystopian dreams. It needs wins to stem the tide of public opinion that's slowly turning against this wave of technology (and, in a creator sense, is turning hella fast).

Instead, they're still trying to build CreepTech. There's just something fundamentally wrong with a machine—whether a robot or an LLM—standing in for actual human companionship. Tech gadgets that aim to replace, rather than enhance, human connection are a line we shouldn't cross.

A friend of the human form offers advice/moments/support, all grounded or based on reality and lived experience. They connect with you because they know you, because they understand you, because they can relate to you, because they can share with you. Shifting this to a chatbot? You don't need me to tell you how that might play out. 

Or maybe I need to shut up and accept that this is a further indication of what awaits us in our metaverse hellscape future—lonely and isolated, each living in our own worlds, devoid of human touch, anxiously waiting for our "friend" to text us to tell us we are going to be okay.