AI Governance for a Prosperous Future

Abstract

Artificial intelligence is the biggest invention of humanity, and the quintessence of the fourth industrial revolution. It is the result of our age-old dream to have a loyal, yet very capable and obedient servant, an equal, at times even a superior intelligence that works, protects and inspires and that we still control to secure and advance our own welfare.

And therein lies the seed of contention. Something that is superior to us can ultimately not be controlled by us but could rather perceive us as it’s resource. A truly dystopian future from a human perspective. Even though this future might be decades or a century away, with AIs still in their infancy, they nevertheless have already demonstrated their transformative power, changing our workplaces, our cars and our homes, selecting our partners and our perception of reality.

And as any tool is also a weapon, we must ensure that AI becomes more tool than weapon and works for humanity and not the other way round.

Corporate AI governance is key to safeguarding the transition to a society in which AI is omnipresent and a blessing, not a curse. It must ensure that its intelligized products and services behave ethically responsible. It must consider the well-being of its employees, customers and business partners wherever AI is deployed.

This is a very tall order, and even more so as with AI we are venturing into the unknown. Only this June (2022) Google sanctioned one of their software developers who claimed that one of the company’s most advanced AI, called LaMDA, should have developed consciousness (https://www.gizchina.com/2022/06/14/google-employee-suspended-after-saying-that-ai-has-become-conscious/). AI is a highly sensitive topic.

Therefore, transparency and a well-structured AI governance are imperative to building trust and ensuring that we do not experience a major backlash in this domain by being careless. A backlash that could cost us dearly as AI is the key to a better future, one with less diseases, less suffering and more wealth for all people on the planet.

This article reflects on the many aspects of AI governance and proposes a way to structure it methodically to make it applicable in work processes, products and services always considering the dichotomy between benefits and risks of AI.

Glossary

Definitions of terms may be more expansive and varying than in this glossary and some still are evolving.

AC

Artificial consciousness

Agile

Highly iterative, learning-focused development process

AI

Artificial intelligence

Alexa

Standard name of Amazon’s virtual assistant, used synonymously for the products

Artificial personality

Also juridical personality, is a non-living entity and has a legal name

Asimov, Isaac

Famous twentieth-century science fiction author

Autopilot

Tesla’s autonomous driving software versioned as FSD

Awareness

Ability to understand the environment in context-> understanding

Consciousness

To experience oneself and the environment through feelings

CSR

Corporate social responsibility

CSR-AI

CSR with a comprehensive AI-governance framework

Cyber

Synonym for digital, used in conjunction with attack, defence, risk, system, threat

Cybernetic

Used for (complex) systems with feedback loops for self-regulation or learning

DOJO

Tesla’s massive neural net computer mainly used for FSD

EAI

Edge Artificial Intelligence

Edge intelligence

Similar to NAI but mainly used for company peripheral intelligent IoT devices

ESG

Environmental social governance, part of modern twenty-first-century CSR

FSD

Full Self-Driving software for autonomous Tesla cars, also called autopilot

GAI

General artificial intelligence

GDPR

General Data Protection Regulation; EU law from 2016

IMF

International Monetary Fund, supranational money lending organization

Industry 4.0

Manufacturing sector-driven initiative to fully digitalize and interconnect production

Know-how

Applied, applicable knowledge, e.g. methodologies, procedures, etc..

Knowledge

General knowledge, abstract and applicable, contains applicable knowledge

Legal person

Legal persons or legal entities are not natural human persons but, e.g. companies

Maslow pyramid

Also Maslow hierarchy of need, based on prioritized needs for survival

NAI

Narrow artificial intelligence with a single or simple task range

Natural person

Human person

Neuralink

Company that develops neural human-machine interfaces

Neural Net

Network of artificial or natural information carrying neuronal cells

NI

Natural intelligence

OECD

Organization for Economic Co-operation and Development

Recoupled

Efficient anglicized word for the German expression ‘rückgekoppelt’ meaning ‘feedback looped’, a system with feedback loop (self-regulating, learning)

Self-awareness

To understand oneself, as a system with boundaries, embedded within the environment

Singularity

State of super intelligent AI far beyond human intelligence levels

Social Credit System

Chinese initiative to introduce a data-driven universal trustworthiness and behaviour measurement system that uses AI

Subject

Actor with self-awareness and own free will and thus accountable

Tesla

Maker of electrical vehicles

Understanding

Ontological and causal (episodic and procedural) associative memory