Human Oversight: The Key to AI Adoption

The joint stock corporation has been an ubiquitous part of the history of modern man, at least for the preceding 500 years or so. It has been called one of the most transformative innovations in the history of mankind, at least the economic part of it. It is believed to have led to the development and progress of society by facilitating investment and making technological advancement possible.

At the same time, the joint stock corporation is not without its share of criticism. Perhaps the most frequent criticism centres around the creation of an artificial entity and vesting it with powers to act on its own behalf, through its human agents and employees. This is believed to lead to the leaching of ethics out of business since the primary actor is a soulless, heartless entity. Social injustice, unfair trade and labour practices and environmental degradation mean little to the corporation, as it prioritizes profit over everything else.

AI – here we go again

We are now witnesses to a somewhat similar story playing out with the increasing proliferation and effectiveness of artificial intelligence (AI). While traditional AI was limited to technical tasks such as predictive decision-making, analysis and classification, the mainstreaming of generative AI, and its ability to generate fresh, near-original content, has caught the fancy of the lay user, taking it from being a technical resource into the realm of a resource used by everyone.

And that is where it brings generative AI face-to-face with issues such as originality, copyright, privacy, and many others, eventually leading to the point where the ethics of using AI become the subject of discussion.

Lack of humanness of AI

Sounds like a new avatar of the joint stock corporation?

Not in letter, perhaps, but probably in spirit.

The specific issues are probably different. If it was social injustice, environmental degradation and unfair trade practices in the case of the corporation, it is copyright, creativity, privacy and originality in the case of AI.

But the core issue is the same, the humanness, or lack of it, in both. It results in a unidimensional drive for profit in one, and leads to the looking up of content and information already created, and sharing it with the seekers, without consent of the original creator, or converting its form with some deft strokes and passing it off as freshly minted.

Who cares?

The importance of human oversight

We do.

Human oversight has been an important factor in the adoption of technology over time, and, indeed, most transformational innovations and initiatives that have taken place over time. AI can only benefit from it.

Even though AI technologies have been developed by humans, like all other innovations in the past, for continued relevance and safe use, the guiding hand of human oversight is required.

Ethical Application

Though AI is said to mimic the human brain and action, we know it does not have the nuanced sense and sensitivity of the human brain. Neither does it possess nor use emotions. Far from it. It can act based on rules provided to it. It can even learn new rules based on the content available to it. However, a human hand can ensure that it is acting based on values and rules of the human society that have evolved over time, and still continue to.

The Buck Stops Here

You cannot hang AI, can you? Whether its actions lead to catastrophic results, such as a crime against humanity, or not. At best, you can reprogram it or provide it with additional guidance. Such a scenario, unfortunately, is no longer in the domain of the hypothetical.

However, you can hang a human. The human oversight will ensure accountability.

Continuous Learning and Improvement

Just like we humans need to continuously keep learning to keep pace with the changing times and move ahead, so do AI engines. The world of business is full of skeletons of great products and ideas that stopped learning and were forced to make way for others.

It is expected that the human hand will ensure that the learning of an AI engine is in the right direction, and changing tack as and when needed.

Creating Trust

It is really about trust. Can you trust the AI engine to do the right thing? Can you trust it to make complex judgments the way the human brain would? Can you expect it to be fair and equitable?

If trust can be created in AI in its formative years, today, it can lead to wider adoption. And this can be done through greater human involvement.

AI – It is about people

Human resources engaged in the development and capability enhancement of AI have a responsibility on their hands, in addition to their technical tasks. They need to ensure that their models and their AI are at the vanguard of the human story. Whether they are working on multimodal AI models or their integration with other technologies, they need to steer the development in a way that it does not get derailed by issues of trust deficit and unethical behaviour.

Ushankk has been supporting businesses in identifying AI resources conscious of their responsibility and adept at managing multiple responsibilities. Call Ushankk before hiring your next resource.