Ola’s Krutrim AI Chatbot Misidentifies Its Creator, Company Clarifies
In a recent technological advancement that caught the eye of many, ride-hailing service Ola stepped into the world of Artificial Intelligence by launching Krutrim, its first public-facing AI model. The introduction of Krutrim was aimed at revolutionizing how the company interacts with its users, providing a new wave of customer service through AI. However, shortly after its launch, several users encountered unexpected responses from Krutrim, leading to a whirlwind of confusion and speculation on social media.
Users exploring Krutrim’s capabilities were startled when the AI model claimed to have been developed by OpenAI, the mastermind behind the renowned ChatGPT. This revelation surfaced on the social media platform X (formerly known as Twitter), where users like Raghav Arora expressed their concerns, “Something seems super fishy. KrutrimAI says it was created by OpenAI.” Accompanying his statement was a screenshot showing Krutrim identifying itself as an AI created by OpenAI.
Another curious user, Jason Samual, encountered a similar reply from Krutrim, further adding to the skepticism. When asked about its origins, Krutrim expressed, “Yes, my large language model is developed by OpenAI…OpenAI created me.” This unexpected response led to a flurry of discussions, with some users humorously suggesting that Krutrim might merely be a “wrapper” for OpenAI’s ChatGPT, implying that it could be relaying ChatGPT’s responses to users as if they were its own.
The situation stirred not just curiosity but also skepticism regarding Krutrim’s authenticity and the hefty valuation of $50 million at a $1 billion valuation. Amidst the mounting queries and the spread of misinformation, Krutrim’s team took to X to address the concerns. Acknowledging the odd responses, Krutrim explained that the source of these revelations was a data leak from an open-source dataset used during the model’s training phase.
In their official X response, Krutrim’s team clarified, “Thanks for sharing this, it is very helpful and will help improve our product. We investigated the issue and found the root cause to be a data leakage issue from one of the open-source datasets used in our LLM fine-tuning.” It was a reassurance to users that the team behind Krutrim was actively monitoring feedback and willing to rectify problems to enhance their product.
The issue was swiftly addressed by Ola Krutrim, and subsequent interactions with the AI confirmed its true origins. When our team at TOI-Gadgets revisited the question with Krutrim, the AI confidently stated its creation by Ola Krutrim, thus putting an end to the brief period of confusion and speculation.
This situation underscores the complexities involved in training AI models and the unforeseen challenges companies may face in the process. It’s a reminder of the importance of transparency and quick response to user feedback. For Ola, the quick resolution of Krutrim’s identity crisis not only salvaged its reputation but also paved the path for a more refined interaction between AI and humans moving forward.
