The Ghost in the Machine

[Directly from Seth Godin’s The Ghost in the Machine]

When a system becomes complex and our knowledge peters out, we’re tempted to assert, in the words of Gilbert Ryle, that there’s a ‘ghost in the machine.’

“How does the stoplight work?” “Well, it knows that there’s a break in the traffic so it switches from green to red.”

Actually, it doesn’t ‘know’ anything.

Professionals can answer questions about how. All the way down.

I’m that professional. I can answer those questions. All the way down.

The Ghost in the Machine, in the case of the stoplight, is an “embedded system.” 

What’s an embedded system? 

They are the tiny, hidden computers everywhere in our world today — your pocket, car, house, and yes, inside the traffic light. In modern Western culture, we come into contact with hundreds, if not thousands, every day. 

The brain of an embedded system is a microcontroller, and the microcontroller plus its human-generated software is special-purpose and usually single-purpose. The ghost in the traffic light only does one thing, as does the ghost in your microwave, your car’s airbags, the airliner’s engine throttle, and the lightbulb that turns on when you tell Alexa you’re home. 

For most of our history, the ghosts in our machines were quite deterministic. Programmed with a decision tree. Single-purpose. If this, then that. 

Now our ghosts are becoming AI-informed. The Small Language Model, if not the Large Language Model, is making its way into your devices. Following its incentives. Learning your patterns and preferences. Making decisions. 

The ghosts are learning.