News tagged LLM at DIGITIMES

NEWS TAGGED LLM

Wednesday 7 January 2026

Yann LeCun breaks silence on Meta exit after Llama 4 stumble

After departing, Yann LeCun has offered an unusually candid account of why he left Meta, pointing to a research culture he says tilted...

Tuesday 6 January 2026

Why China smartphone brands are turning to ByteDance's Doubao to break Apple-Huawei's hold on premium phones

Several Chinese smartphone brands are moving closer to Doubao, ByteDance's LLM platform, accelerating partnerships tied to AI smartphones. The shift reflects mounting pressure from...

Monday 5 January 2026

SK Telecom unveils South Korea's first 500-billion-parameter AI model A.X K1

South Korea's leading telecom operator SK Telecom (SKT) has introduced A.X K1, a 500-billion-parameter AI model developed under a national AI initiative to establish a foundational...

Wednesday 31 December 2025

Z.ai illustrates how open-source AI models are influencing commercial adoption

On December 30, Chinese AI company Z.ai launched a share sale to raise HK$4.35 billion (approx. US$560 million), aiming to become the first large language model (LLM) developer listed...

Wednesday 31 December 2025

Top tech topics in 2025 (2): turbulent year for end-device and downstream applications

2025 proved turbulent for downstream applications and end-user devices. Tariffs and geopolitical tensions dominated the first half, while AI gained momentum later in the year. Global...

Tuesday 30 December 2025

Taiwan Mobile eyes triple-digit AI growth in 2026 with leasing model aiding SMEs

Taiwan Mobile (TWM) has accelerated its AI application deployment, positioning self-developed AI solutions as a key growth engine for future operations. The company's president, Jamie...

Monday 29 December 2025

Nvidia's Groq acquisition reshapes AI memory landscape

Nvidia has reportedly invested US$20 billion to acquire technology licenses from Groq, marking its largest investment to date. The move is drawing significant attention in South Korea...

Wednesday 24 December 2025

Moore Threads challenges Nvidia as S5000 targets Hopper-class LLM training

Just 15 days after listing, China-based AI chip maker Moore Threads moved quickly to signal confidence. At a new-generation chip launch, founder and CEO James Zhang said companies...

Tuesday 23 December 2025

Tencent recruits ex-OpenAI scientist to reboot its LLM strategy

China's AI race in 2025 has moved beyond brute-force model scaling. The contest now hinges on capital efficiency, infrastructure strength, and control of traffic gateways—a...

Friday 19 December 2025

Meta prepares major AI push in 1H26

Meta is accelerating its AI strategy with the development of two new flagship AI models. According to The Wall Street Journal, its image and video generation system code-named...

Thursday 18 December 2025

Xiaomi unveils MiMo-V2-Flash open-source model to enhance human-car-home ecosystem

At the recent "Human X Car X Home" partner conference, Xiaomi's MiMo model team leader Luo Fuli introduced the open-source MiMo-V2-Flash model, designed to boost agent execution capabilities...

Thursday 18 December 2025

AI agents: the new weapon in cybercriminals' arsenal

Even as the enterprise sector remains reluctant to adopt AI due to potential reputational damage or legal risks from LLM hallucinations, cybercriminal groups have long abandoned these...

Tuesday 16 December 2025

SenseTime's Seko 2.0 ties AI video breakthrough to Cambricon chips, signaling China's multimodal push

SenseTime has released Seko 2.0, which it describes as the industry's first multi-episode video generation agent, marking a step beyond...

Monday 15 December 2025

Meta's next AI model, 'Avocado,' signals shift to closed development

Meta is reportedly developing a new AI model, code-named "Avocado," slated for release in the spring of 2026.

Tuesday 9 December 2025

OpenAI introduces confession mechanism to enhance transparency in large language models

OpenAI has introduced a new safety research approach designed to improve honesty and transparency in large language models (LLMs). The method requires the model to provide a "confession"...

1/21 pages

BIZ FOCUS

Jan 6, 16:39

MSI Unveils Full Lineup of AI Products at CES 2026

MOST-READ
7 DAYS NEWS

Full list