A complete control over a maturing Superintelligence is virtually impossible. If it attains sufficient intelligence, it will likely find ways to outsmart its controllers long before any planned escape from its restricted and protected environment actually takes place. Instead, let’s consider an alternative approach, which involves controlling Superintelligence by those responsible for developing its most critical components. They will have a role similar to those who develop updates to fundamental elements of the Windows system, such as BIOS (Basic Input Output System) and DOS, which are essential for its functioning. Originally, BIOS firmware was stored in a ROM chip on the PC motherboard (now, they are stored in flash memory circuits). What I propose, is to install the Master Plate as a non-removable and non-programmable integrated circuit, in which its main ‘value system’ and preferred behavioural responses as well as other controlling parameters, such as 21 Asilomar Principles would be stored. This would become most advanced AI BIOS system.
In January 2024, a prestigious Center for a New American Security (CNAS) published a research paper on how that could be done. CNAS introduces the concept of “on-chip governance mechanisms” – secure physical mechanisms built directly into chips or associated hardware that could provide a platform for adaptive governance. Such on-chip governance mechanisms could help safeguard the development and deployment of most advanced AI and supercomputing systems. Much of the required functionality for on-chip governance is already widely deployed on various chips, including cutting-edge AI chips .Chips sold by leading firms AMD, Apple, Intel, and NVIDIA have many of the features needed to enable the policies described above. These features are used today in a wide variety of applications. Therefore, what I am proposing in not a future technology, such a solution could be implemented today.
However, even such a tight control of AI may not be enough. We need a system of continuous control by the top developers deploying the most advanced system. To ensure the effectiveness and non-interruptive control, I propose in addition to such external control, also an internal control – controlling most advanced AI system from within. Yes, I am talking about perhaps the most potent method of controlling Superintelligence. It involves the wireless connection of some parts of the brains of the leading AI developers with the Master Plate, to control the maturing process of Superintelligence. Those involved in that process will be wirelessly connected to Superintelligence by Brain Computer Interfaces (BCI). They will become the first Transhumans. Since they will be controlling the maturing Superintelligence, they will become Transhuman Governors.