Gaurav DhooperAssistant Vice President| GenpactNoida, U.P., India
Generative AI presents both opportunities and challenges for companies undergoing major transformations. Striking the right balance between automation and human interactions is crucial to safeguard brand reputation, preserve customer trust and drive sustained business success.
What are your thoughts around striking this right balance?
Senior Projects Manager | Field & Marten AssociatesNew Westminster, British Columbia, Canada
Gaurav, I totally agree with you. Without a certain degree of human interaction, at least to verify outcomes, things can go out of hand so this balance is very much needed! Saving Changes...
As with most balances, the tricky part is to figure out the ratio of human to machine interaction which provides sufficient safety while still benefiting from the efficiencies of the machine.
Kiron Saving Changes...
Sergio Luis ConteHelping to create solutions for everyone| Worldwide based OrganizationsBuenos Aires, Argentina
Generative AI, used at organizational level, needs and leverage the ancient AI concept called human in the loop. So, this answer your question. The problem is the marketing that generative AI is gaining then it becomes a new buzzword and lot of people talk, create posts, write papers without knowledge about it. Saving Changes...
Surupa ChakravartyBusiness Development Manager| InfosysToronto, Ontario, Canada
I agree to your point. There is a need for human interaction. Without which, it will be difficult to measure and verify qualitative outcomes. Balance is must. Saving Changes...
From the technology perspective, there is no alternative to the human interaction in any of the application development cycles: - Before training, a popular reinforcement learning technique needs human feedback - During usage, one can improve performance by giving a few examples, called few-shot learning
Also, the model, after the deployment, should be continuously monitored and improved upon by latest technology trends, which involves human interaction via prompt engineering and the corresponding evaluation. Saving Changes...
The repetitive task is any way getting automated as human don't love to do task repetitive in nature. How does AI different from the repetitive task? Can AI do anything new; NO.. Human can only add values. Human intervention is more desirable by companies. Problem arise when human start doing repetitive task without adding any value. there comes AI.. Saving Changes...
Raman ChadhaManager| DeloitteMillbrae, United States
Evaluating different tasks and classifying them into categories before trying to automate everything is important. A simple (perhaps too simplistic) example is to classify each task into a risk-reward matrix of automation. E.g., I would be wary of using AI (or any other tech) for tasks / use cases that are high risk, and especially if the reward is low, without having (the right) human-in-the-loop. Saving Changes...
Community Champion
