Inc.
Your AI assistant might soon know you better than your closest friends. But how much personal information is too much?
Something Is Spying on Me. And It’s My Fault
Author: Dave Sokolin
I’ve been demoing a lot of the different AI assistants out there over the past few weeks. Grok Bot, Muse, Instinct. I’d like to get my hands on a few others, but I’ve noticed something disturbing in the process.
Recently, I was sitting outside when I said something and a voice like Siri responded. The problem was, I had no idea which of the bots running on my devices was talking to me. My immediate move was to close everything on my iPhone and iPad, then go through my settings to see what had access to my microphone.
I have so many different conversational agents running simultaneously that I no longer know who’s listening, or when they’re listening. And it’s my own fault. I’ve turned all these things on. I’ve agreed to terms and conditions I’m not even sure I understand.
We’re entering an interesting era. One where the possibilities for productivity are almost endless, but at the same time, we’re trading away our privacy to get there. And we’re doing it voluntarily.
The better these assistants become, the more information we want to give them. The more they know, the more useful they become. And the more useful they become, the harder they are to leave.
The Bug Is the Feature
An intelligent assistant has to know everything about your life. And in some ways, that’s been the goal of surveillance capitalism all along. Instagram, YouTube, the algorithms that learn everything about us.
This is surveillance capitalism’s dream. Yet it’s also my dream to have a full-time assistant that knows everything about me and can actually do things on my behalf. That’s the paradox.
In the end, I suspect privacy wins out, and Apple probably ends up gaining my trust, with me turning everything else off. But for now, Apple doesn’t even have a real assistant. Siri sucks. Even 2.0, powered by Google, doesn’t really do anything different. It just works a little better. It doesn’t have anything close to the capabilities of Grok Bot or Muse.
Apparently, OpenAI has one coming, and I’ll probably trust that one about as much. Although I bet it’ll be more useful, simply because I have so much information historically uploaded to ChatGPT.
The New AOL Address
Are our agents going to be as sticky as email addresses? Some people still have the same AOL and Yahoo addresses they had 20 years ago.
I think a lot of these early agents will have enormous long-term lock-in. Once an assistant knows your life, your preferences, your relationships, your history, how easy is it really going to be to switch?
It’ll be interesting to see what Google and Apple do, now that they’re teaming up, to combat this. They’re losing ground to a new category of assistants that people are adopting incredibly quickly.
Think about it. Muse reportedly had 5 million downloads in its first few weeks. That’s potentially 5 million people beginning to move part of their digital lives outside the traditional Apple and Google ecosystems. And those relationships could last for decades.
The Most Embarrassing Hack in History
Of course, all of this could change with one major hack of a celebrity or government official.
Imagine an always-on assistant in your life. Even if it doesn’t listen to everything, it potentially knows everything. Your psychiatrist appointments. Your proctologist appointments. Your private conversations. Your relationships. The things you search for but would never tell another person.
Imagine a hack of that proportion. Not just your credit card number or your password. Your entire life. And unlike a password, you can’t exactly change your personal history.
It’s Not If. It’s When.
There is so much promise and so much peril in this new era. The news has us completely freaked out about AI, yet when you actually try these things, they can help you so much that you start to see what all the excitement is about.
I’ve seen it firsthand. Right now, I feel like I’m racing past other people in productivity. I’m getting things done that would have taken me hours, or that I probably wouldn’t have bothered doing at all.
But eventually these tools will proliferate, and we’ll all be marching at a different speed. That’s what makes this moment so interesting. The productivity gains are real, and so are the risks.
Who Do You Trust With Your Life?
What’s the cost versus the benefit when it comes to privacy? I think this is going to become one of the biggest questions of the AI era.
And as I finish writing this, I keep coming back to Apple. I think Apple has been a great steward of my information. I’ve trusted them with my digital life for years.
Now, if they could only get their act together and build something worthy of it.
Credits: TCA, LLC.