prakhar@affmantra.com

ChatGPT can now pay for you

Yes, you read that right – soon you’ll be able to chat, shop, and pay right inside ChatGPT. No app-switching, no typing UPI IDs – just tell the AI what you want, and it’ll handle the rest (with your permission, of course). India’s NPCI, Razorpay, and OpenAI have teamed up to test this new feature…

Read More

Noise Master Buds Max headphones launched in India at ₹9,999 featuring Sound by Bose Technology

Noise has unveiled the Master Buds Max, its first over-ear headphones under the flagship Master Series, following the Master Buds TWS launch earlier this year. The new model debuts with ‘Sound by Bose’ technology – marking a major collaboration that brings premium acoustic tuning to an accessible price segment. Unveiled earlier at IFA Berlin 2025,…

Read More

OpenAI partners with Broadcom to build 10 Gigawatts of custom AI accelerators

OpenAI has announced a major multi-year partnership with Broadcom to develop and deploy 10 gigawatts of custom AI accelerators, marking a significant step toward enhancing its computing infrastructure for large-scale AI model training and deployment. Under this collaboration, OpenAI will design the accelerators and system architecture, leveraging its deep expertise in developing frontier AI models….

Read More

Airtel, Google team up for $15 Billion AI data hub in Visakhapatnam

Bharti Airtel and Google have announced a landmark partnership to establish India’s first large-scale Artificial Intelligence (AI) hub in Visakhapatnam, Andhra Pradesh. The initiative marks one of the largest technology infrastructure investments in India, aimed at boosting AI adoption and strengthening the nation’s digital ecosystem. The collaboration involves a $15 billion investment from Google, spread…

Read More

Portronics Tune Plus wireless car adapter launched with built-in YouTube and Netflix

Portronics has introduced the Portronics Tune Plus, a wireless car adapter that transforms traditional wired Apple CarPlay and Android Auto setups into a seamless wireless experience. The device also comes with pre-installed streaming apps, including YouTube and Netflix, bringing entertainment to your car like never before. The Portronics Tune Plus eliminates the need for a…

Read More

NVIDIA Researchers Propose Reinforcement Learning Pretraining (RLP): Reinforcement as a Pretraining Objective for Building Reasoning During Pretraining

NVIDIA AI has introduced Reinforcement Learning Pretraining (RLP), a training objective that injects reinforcement learning into the pretraining stage rather than deferring it to post-training. The core idea is simple and testable: treat a short chain-of-thought (CoT) as an action sampled before next-token prediction and reward it by the information gain it provides on the…

Read More

A Coding Implementation of Advanced PyTest to Build Customized and Automated Testing with Plugins, Fixtures, and JSON Reporting

In this tutorial, we explore the advanced capabilities of PyTest, one of the most powerful testing frameworks in Python. We build a complete mini-project from scratch that demonstrates fixtures, markers, plugins, parameterization, and custom configuration. We focus on showing how PyTest can evolve from a simple test runner into a robust, extensible system for real-world…

Read More

Andrej Karpathy Releases ‘nanochat’: A Minimal, End-to-End ChatGPT-Style Pipeline You Can Train in ~4 Hours for ~$100

Andrej Karpathy has open-sourced nanochat, a compact, dependency-light codebase that implements a full ChatGPT-style stack—from tokenizer training to web UI inference—aimed at reproducible, hackable LLM training on a single multi-GPU node. The repo provides a single-script “speedrun” that executes the full loop: tokenization, base pretraining, mid-training on chat/multiple-choice/tool-use data, Supervised Finetuning (SFT), optional RL on…

Read More

Alibaba’s Qwen AI Releases Compact Dense Qwen3-VL 4B/8B (Instruct & Thinking) With FP8 Checkpoints

Do you actually need a giant VLM when dense Qwen3-VL 4B/8B (Instruct/Thinking) with FP8 runs in low VRAM yet retains 256K→1M context and the full capability surface? Alibaba’s Qwen team has expanded its multimodal lineup with dense Qwen3-VL models at 4B and 8B scales, each shipping in two task profiles—Instruct and Thinking—plus FP8-quantized checkpoints for…

Read More