Krutrim AI Prepares to Launch Chatbot App with Multilingual Capabilities: A Preview by Founder Bhavish Aggarwal
Krutrim AI, also known as Krutrim Si Designs, is set to launch its first product, a generative artificial intelligence (AI) chatbot app. Bhavish Aggarwal, the founder, revealed the launch timeline on social media and shared screenshots to provide a glimpse of the app’s capabilities.
Launch Update and Testing Phase
Bhavish Aggarwal took to social media on February 4 to share updates about the impending release, mentioning that the testing phase is underway. The app is expected to be released next week, with ongoing efforts to enhance the models even after the launch.
Generative Capabilities and Demonstrations
Through several posts, Aggarwal showcased the generative capabilities of the Krutrim AI chatbot. The app is demonstrated to handle basic queries, offer suggestions, and generate content in response to various prompts.
AI Response Controversy and Netizens’ Reactions
One particular post stirred controversy when the Krutrim AI chatbot, upon being asked whether India was a country before independence, provided a response that some users deemed inaccurate. The controversy led to discussions about AI hallucination, with comparisons made to the response of OpenAI’s ChatGPT.
Product Details and AI Models
Krutrim AI has developed a family of large language models (LLMs), including Krutrim base and Krutrim Pro, described as a multimodal foundational model. These models claim to have been trained on over two trillion Indic language tokens. The generative AI apps from Krutrim AI are expected to be voice-enabled. The company has posted benchmark scores comparing its AI models against the Meta Llama 2 7B model, showcasing competitive performance on various benchmarks.
As Krutrim AI prepares to launch its chatbot app, the generative capabilities and multilingual support aim to position it as a significant player in the AI landscape, catering to diverse linguistic needs. The controversy surrounding an AI response highlights the ongoing challenges and discussions in the development of large language models.