LLAMA 3 1
Binance sees outflows stabilising as exchange returns online after $6.3 billion heist
Binance customers pulled $3.64 billion, or 2.1% of its assets, as the world's biggest cryptocurrency exchange resumed service after a security breach linked to North Korea saw $6.3 billion stolen from the platform last week.
Binance sees net outflows of $3.64 billion after resumption of withdrawals
On November 1, Binance reported net outflows of around $3.64 billion as withdrawals resumed after a pause. During this period, withdrawal volumes reached over $4.64 billion while deposits amounted to $1.01 billion. The exchange stated that the percentage of outflows represented 2.1% of its total assets. Despite the large outflows, Binance highlighted that deposit volumes remained substantial and withdrawal volumes began to stabilize.
Luis Ceze, a professor, turned his lab's project into a startup. Nvidia bought it, made him Vice President, and he kept teaching
Nvidia acquired OctoAI, a software startup founded in 2019, reportedly for $165 million. Following the acquisition, OctoAI was dissolved as an independent entity, and its CEO joined Nvidia as VP. Luis Ceze focuses on AI systems software in his role at Nvidia. OctoAI specialized in generative AI platforms, having reportedly raised over $132 million since its inception.
Making people redundant is a no-win: Revolut India CEO on AI job cuts
As large private banks and financial institutions trim headcount to automate their workforce, Revolut’s Paroma Chatterjee believes it is critical to account for AI adoption first in processes before hiring.
Gnani AI launches Artha sovereign AI stack with 30-billion-parameter Evon 3.3
Gnani AI launched Artha, a sovereign AI stack for Indian companies and public institutions. This stack features the Evon 3.3 language model and the Plexus agentic platform. Evon 3.3 is an open-weights model trained on Indic languages and domain-specific data. The Artha stack allows organisations to run AI models within their own infrastructure.
ETtech Explainer: Meta's stop-start strategy for its open source AI models
Glimmer, a 30-billion-parameter model, is much smaller than leading AI models from rivals and is designed for agentic tasks. It can run on a Mac or PC with a single graphics card, aiming to tap into demand for artificial intelligence (AI) systems that run directly on people's devices.
- Go To Page 1
ETtech Explainer: How Meta's Muse Spark fares against Anthropic's Opus, OpenAI's GPT, Google's Gemini models
Muse Spark is the first model from the Meta Superintelligence Labs. It is a natively multimodal, reasoning-focussed AI model, designed to power the next generation of Meta AI products. While the model stands out in multimodal use cases and real-world integrations, it lags in core reasoning and agentic coding benchmarks.
AI breakthroughs spur race for superintelligence
ET traces the major LLM launches of 2025 — dubbed as the Year of AI model breakthroughs — and decodes the big hits and misses of the year so far.
Who is Shengjia Zhao, OpenAI co-creator and now Meta's new Superintelligence chief?
Zhao, a PhD from Stanford University, comes from OpenAI, where he played a big role in creating ChatGPT, GPT-4, and several models, including GPT-4.1 and o3. At Meta, he will set the strategy for the foundational models, especially the Llama series, and drive foundational AI research.
Meta names ChatGPT co-creator Shengjia Zhao as chief scientist of Superintelligence Lab
Zhao, a former research scientist at OpenAI, co-created ChatGPT, GPT-4 and several of OpenAI's mini models, including 4.1 and o3. He is among several researchers who have moved from OpenAI to Meta in recent weeks, part of a broader talent arms race as Zuckerberg aggressively hires from rivals to close the gap in advanced AI.
Startups rethink IPOs; US visa dilemma for students
Happy Tuesday! The global market rout may push Indian startups to reconsider IPO plans. This and more in today’s ETtech Morning Dispatch.
ETtech Explainer: How Meta's Llama 4 stacks up against Chinese AI models Qwen, DeepSeek, and Manus AI
Meta has launched Llama 4, its latest open-weight AI models, including Scout, Maverick, and Behemoth, offering advanced multimodal capabilities. These models excel in text, image, and video processing. Meta claims that they outpace rivals like OpenAI’s GPT-4 and China’s Qwen, with strong growth in open-source adoption globally.
Former Meta executives raise $15 million for AI assistant startup
The round was led by Rob Toews at Radical Ventures, with participation from other investors like Felicis, "AI godmother" Fei-Fei Li, and Google DeepMind chief scientist Jeff Dean. San Francisco-based Yutori is part of a slew of AI startups creating autonomous agents, or systems that use AI to perform actions on their own.
Qwen AI model by Alibaba offers low-cost alternative to DeepSeek
Researchers from Stanford University and the University of Washington, including Chinese-American “AI godmother” Li Feifei, have recently published their study regarding the training of the S1 reasoning model on the back of Alibaba's Qwen 2.5- 32b - Instruct model.
Alibaba releases AI model it claims surpasses DeepSeek-V3
The unusual timing of the Qwen 2.5-Max's release, on the first day of the Lunar New Year when most Chinese people are off work and with their families, points to the pressure Chinese AI startup DeepSeek's meteoric rise in the past three weeks has placed on not just overseas rivals, but also its domestic competition.
How DeepSeek is redefining AI, challenging Silicon Valley—will it be a turning point in the global AI landscape?
DeepSeek, a Chinese AI startup, is making waves in the tech world with its latest AI model, DeepSeek V3. This model has outperformed major competitors like Meta's Llama 3.1 and OpenAI's GPT-4o in tasks such as text understanding, coding, and problem-solving.
Has the Chinese startup DeepSeek outsmarted Meta and OpenAI? Here's all about its AI platform that was trained on a joke of a budget
DeepSeek, a Chinese AI start-up, has unveiled its latest large language model, DeepSeek V3, a game-changing development in artificial intelligence.
Meta's open source push: Llama's global impact and India's AI role
Llama is a foundational technology of intelligence. You want to make sure it’s the most efficient, responsible, safe and high-performance. You can only do so much as one company and must do it together with everyone. The best way to do it is to be an open ecosystem.
Meta unveils voice mode: You can now talk to and share photos with Meta AI
Over 400 million people are using Meta AI, out of which 185 million people are using it across Messenger, Facebook, WhatsApp and Instagram each week, the company said at the Meta Connect event at its headquarters in Menlo Park, California.
FirstCry Q1 financials; India Inc’s rush for AI talent
FirstCry reported a 17% jump in its April-June quarter revenue on year. This and more on today’s ETtech Top 5.
Meta’s Llama models hit 350 million downloads, leads open-source AI
Besides the community platform Hugging Face, Llama was also being accessed through Meta’s cloud partners such as Amazon Web Services (AWS), Microsoft’s Azure, Databricks, Dell, Google Cloud, Groq, NVIDIA, IBM watsonx, Scale AI, Snowflake, to name a few.
Meta just launched the largest 'open' AI model in history: here's why it matters
Meta, the parent company of Facebook, took up the fight for open-source AI in a big way by releasing a new collection of large AI models. These include a model named Llama 3.1 405B, which Meta's founder and chief executive, Mark Zuckerberg, says is "the first frontier-level open source AI model". In the case of ChatGPT, its parent company, OpenAI, releases neither the dataset nor code of its latest AI tools to the public.
ET Infographic | Running out of data
With foundation models running out of data for training, companies are placing their bets on the use of synthetic data for training the models. But based on a recent research paper, “indiscriminate use of model-generated content in training causes irreversible defects in the resulting models”.
Meta to let users to create custom AI characters
Meta Platforms launched AI Studio, which lets users design and share personalized AI chatbots. Built on Llama 3.1, this tool allows customized AI characters, assisting Instagram creators with common DM questions. Users can share their AI characters on Meta's platforms. Llama 3.1 competes with paid models from rivals like OpenAI.
Govt’s crackdown on financial fraud; Singapore new visa framework explained
Indian Cyber Crime Coordination Centre has shut illegal loan websites to prevent the increasing number of financial frauds. This and more in today’s ETtech Top 5.
Mark Zuckerberg, AI's 'open source' evangelist
Mark Zuckerberg, the founder of Facebook and CEO of Meta, has become an unexpected evangelist for open source technology when it comes to developing artificial intelligence, pitting him against OpenAI and Google. That concern has now spread to generative AI, but this time it is Microsoft-backed OpenAI and Google that are the closed-fence culprits that charge developers and keep a tight lid on their AI technology.
Meta’s Llama 3.1 to support Indian startups with synthetic data generation
Meta, on Tuesday, launched its largest-ever open-source AI model Llama 3.1 405B, and claims to have closed the performance gap with top closed-source models such as Open AI’s GPT-4o and Anthropic’s Claude 3.5 Sonnet. The latest iteration of Llama, which has 405 billion parameters, was trained on a massive 15 trillion tokens using 16,000 of Nvidia’s ultra-expensive H100 GPUs.
Meta unveils biggest Llama 3 AI model, touting language and math gains
Meta’s Llama 3, featuring 405 billion parameters, rivals OpenAI and Amazon. Enhanced coding and math, scoring 73.8 on MATH and 88.6 on MMLU, boost its appeal. Future multimodal versions will incorporate text, image, video, and speech. The Meta AI chatbot, gaining popularity, promises efficiency and attracts developers.
11.4 million wearables shipped globally in Q1 2015: IDC
Vendors shipped a total of 11.4 mn wearables in the first quarter of 2015, a more than two-fold jump from 3.8 mn wearables shipped in the year-ago period.
Load More
