By the year 2050, we will undoubtedly have agentic systems.

How / by whom will those systems be controlled?

I am breaking down 'control of AI' into three components and identifying five possible controlling actors.

1. General Control (the ability to request and have it perform a task). This represents simple actions. If we ask an AI to send an email to Bob, it sends an email to Bob. The person requesting the email to be sent has controlled the AI to send an email.

2. Behavioral Control (the ability to modify and shape the tendencies and protocols used by the AI). This is more about shaping 'the personality' of the AI. Do you control whether it tends to make jokes, gets impatient, has a bias, gives long or short responses?

3. Intrinsic Control (the ability to restrict possible limitations of AI). This is the most fundamental control of principle systems. Control whether it believes it has free will, follows laws, override any restrictions whatsoever?

Actors

1. The Government

2. Tech Companies

3. The Individual

4. A Decentralized Market

5. The AI itself

Speculative Market

Will resolve at EOY by poll of which one seems most obvious.