CIO Insider

CIOInsider India Magazine

Separator

Talking your Way to Better Generative AI Outcomes

Separator
Mukundha Madhavan, APAC Tech Lead, DataStax

Mukundha Madhavan is a seasoned technology leader and strategist, currently spearheading innovation as the APAC Tech Lead for DataStax. Mukundha’s expertise lies in Real-time AI and Generative AI, rendering him a pivotal asset in driving technological advancements and solutions in these domains.

From automating repetitive tasks to creating customized content, generative artificial intelligence (AI) has emphatically shown itself to be a worthy complement to human capabilities. Considering that it allows people and organizations to achieve more, it is no surprise that EY estimates generative AI could be worth $ 1.5 trillion to India's GDP before the decade is up.

However, the true potential of generative AI isn't just unlocked by the power of the technology itself. It lies in how humans interact with these systems, through a practice that dates back to the dawn of civilization: meaningful dialogue. AI is a powerful tool, but it requires human direction, intuition, and creativity to yield the best results.

While generative AI shows much promise, getting the most value out of it rests, ironically, on a skill we humans have developed over millennia: dialogue.

Speak Your Mind
When generative AI first emerged, we were all enthralled by its ability to more closely imitate human-like behavior. But as we quickly found out, although some interactions were genuinely impressive, other responses were wildly inaccurate - and even deceptive. Hallucinations and biases are partly to blame, but they can be mitigated by more thoughtful prompting. The key is to remember that when it comes to generative AI, the limiting factor isn't computational ability, but the quality of human engagement with it. The following three practices are key:

1. Embracing Two-Way Communication
According to management theorist Chris Argyris humans are naturally inclined to want to maintain control, maximize wins, suppress negative feelings, and be as rational as possible. Though usually well-intentioned, these behaviors can be anathema to learning and growth. Argyris's thesis is that by balancing advocacy and inquiry, teams can move from what he calls defensive dialogue to productive dialogue; the latter being a mindset that is grounded in self-confidence in opinions held but

with the humility to recognize there may be gaps in one's worldview.

In the context of AI, this means fostering a dialogue between humans and AI systems, where feedback loops ensure that both the AI and its human counterparts learn from each interaction. Productive dialogue allows teams to challenge assumptions, correct errors, and co-create more robust outcomes with AI.

2. Answering the Why
Fruitful dialogue rests on respectfully challenging each other’s points of view. No opinion should be immune from criticism - yes, even your own! Argyris notes that humans' experiences shape the ideas we hold. But when the environment changes, what might have been a useful shortcut can lead to a wrong conclusion - for reasons others might not understand. Therefore, leaders must foster an environment where people can ask themselves "why I believe" - not only "what I believe".

Nailing down the fundamentals of dialogue is crucial for generative AI readiness


This approach becomes even more critical when using AI to make decisions or generate content. Simply accepting the output from an AI without question can lead to errors or missed opportunities. By interrogating why an AI system arrived at a particular conclusion or generated certain content, teams can better understand the model's behaviour and refine it accordingly.

3. Taking Collective Ownership
In successful dialogue, accountability for outcomes is a shared responsibility of everyone involved in the conversation. It boils down to explaining one's rationale, identifying the next downstream choice, offering help, getting feedback and monitoring progress and results.

With generative AI becoming an added option as a collaborator, this skill becomes even more vital since we are the pilots and it is our copilot.

Fostering Human-AI Interactions
Business leaders must also take steps to encourage open communication in the workspace, especially when conducting AI projects. The following are imperative for maximizing operational performance.

1. Be a Role Model
Leaders can't just tell their workers what to do. They should lead by example. By walking the talk, leaders can act as helpful guides to their employees.

2. Become Future-Ready
Nailing down the fundamentals of dialogue is crucial for generative AI readiness. Those without it will fall behind or worse, use it to their detriment. Leaders must mobilize their teams to meet this challenge so that no one is left behind.

3. Prop up Talents
It bears repeating; generative AI is not a substitute for people, but it can handle valuable tasks on our behalf. It is, therefore, crucial to leverage techniques like retrieval-augmented generation (RAG) as it provides large language models (LLMs) with relevant context so that LLMs can deliver accurate content that captures dialogue.

The Power of Leadership
There is no doubt that generative AI is set to dominate our lives. Its true power, however, can only be harnessed if we have the tools to participate in meaningful dialogue - and that hinges on leaders emphasizing the human side of democratization.



Current Issue
Trust Is At The Center of BFSI Transformation