If the signal is very weak and the noise is large, it’s easy to imagine that there’s no signal at all.
AI and computers can be used as lenses now, which means we can strip away the noise and see things that we certainly didn’t expect.
Dina Katabi at MIT can point a radio antenna at you while you’re sleeping (even from across the room or through drywall) and determine whether you’re dreaming or not.
We can see patterns in how people type, surf the web or interact with others. We can sort through huge volumes of data and find connections that others were sure were invisible. Not just in whether we’re dreaming, but in how medicine works, the forces driving our culture and more.
Systems work because each element of the system sends a signal to the other elements. Sometimes those signals are obvious and the system works in predictable ways. Other times, we rely on false proxies because we simply can’t tell the signals from the noise. New diagnostics in every field are changing this.
Typical consumer AI is a clever trick that makes us think the computer is a pretty good writer. But machine learning aimed at patterns and signals shows us things that we never knew were there.
For a long time, we’ve assumed that complex systems (like the brain, or the weather) were a sort of magic. As we are able to decode the signals, though, we have a chance to understand how they work.