AI News
Reporting, analysis and features from AIM's AI News section.
Bengaluru-based Conversational AI Startup Gnani AI Raises $4 mn
Gnani AI has over 100 customers in banking and financial services, insurance, telecom, automotive and healthcare industries.
vandana.nair ·
Capgemini expands GenAI efforts with 350 new projects, 2,000+ deals in pipeline
Trained more than 120,000 employees on generative AI tools.
Vidyashree Srinivas ·
CP Gurnani's AIonOS Partners with Indosat to Accelerate AI Sovereignty in Indonesia
Pritam Bordoloi ·
NVIDIA- Acquired Startup Releases Brev v2.5 Custom Containers
This is the first launch announcement since NVIDIA acquired Brev.
vandana.nair ·
Indian GenAI Startup Devnagri Secures Funding Led by Inflection Point Ventures
The company was co-founded by Nakul Kundra and Himanshu Sharma, who collectively bring over 15 years of entrepreneurial experience.
Mohit Pandey ·
Google DeepMind's AlphaProof and AlphaGeometry Hit Silver Medal Mark at International Math Olympiad
The AlphaProof and AlphaGeometry teams continue to explore AI approaches for advancing mathematical reasoning and plan to release more technical details on AlphaProof soon.
Siddharth Jindal ·
NVIDIA Starts Supplying GH200 AI Chips to Tata Communications, Jio Platforms
Yotta being an elite partner received the first shipment of the 4000 H100s in March 2024.
Sagar Sharma ·
TWO Announces SUTRA Dual2L, a Multilingual AI Model Using Llama 3.1
The Dual2L model enhances Llama 3.1's capabilities by expanding support to more languages, including Korean, Gujarati, Tamil, Malayalam, and Japanese, with a 5-10% improvement in multilingual performance for the 70B model.
Siddharth Jindal ·
Mistral AI Unveils Mistral Large 2, Beats Llama 3.1 on Code and Math
It also supports 80+ coding languages, such as Python, Java, and C++.
Siddharth Jindal ·
Meta Llama 3.1 is Officially Out, Dethrones GPT-4o
“I think that open source AI is going to become the industry standard just like Linux did. "
Siddharth Jindal ·
NVIDIA’s New Model ChatQA-2 Rivals GPT-4 in Long Context and RAG Tasks
The Llama3-ChatQA-2-70B model can process contexts up to 128,000 tokens, matching GPT-4-Turbo's capacity.
Gopika Raj ·
Bhashini Signs MoU with New Media Wing of Ministry of I&B for Indian Languages Translation
The partnership will focus on language translation services, encompassing translation into 12 languages, text-to-text, text-to-speech, speech-to-text, audio localisation, live video/audio translation, and subtitling.
Mohit Pandey ·
Indian Government Boosts Space Tech Development with INR 1000 Crore Allocation
A 1000 crore venture capital fund will be set up by the government for space tech investments.
vandana.nair ·
Indian Govt Announces Internship for Youth with INR 5000 Monthly Allowance in Top Companies
Companies are expected to bear the training cost and 10% of the internship cost from their CSR funds.
Vidyashree Srinivas ·
Google Researchers Introduce Conditioned Language Policy Framework for Enhanced Multi-Objective Fine Tuning
The CLP framework enhances language models for summarization, conversational agents, and social norms encoding by balancing multiple objectives for real-world flexibility and usability.
Gopika Raj ·
Union Budget 2024-25: Indian Govt to Spend INR 1,16,342 Crore on IT and Telecom
In the 2023-24 Budget, the Telecommunication sector received Rs 59,740 crore.
Vidyashree Srinivas ·
Bhashini Provides Real-Time Translation of Union Budget 2024 Speech
Bhashini, which has recently reduced its translation lag from 6 seconds to 1-2 seconds, provided simultaneous captions during the Finance Minister’s budget speech.
Siddharth Jindal ·
Elon Musk’s xAI to Release Grok 2 Next Month, Grok 3 in December
“We haven't yet applied real world video from Tesla to Grok,” said Elon Musk.
vandana.nair ·
Cohere Reaches $5.5 Billion Valuation in Latest Funding Round
The latest funding round, led by Canadian pension investment manager PSP Investments, has raised $500 million for Cohere.
Siddharth Jindal ·
Apple's GenAI Takes a 'Lazy' Leap with LazyLLM
They’ve introduced LazyLLM, which selectively computes the key-value (KV) for important tokens during the pre-filling and decoding stages.
Tarunya ·