The Telco That Bet Its Future on OpenAI
Let me start with the one that actually made me sit up straighter. On August 12, Malaysian mobile operator U Mobile announced a strategic collaboration with OpenAI—the company’s first telecommunications partnership in Malaysia.
Here’s why this matters. U Mobile isn’t just buying API access. They’re embedding OpenAI’s frontier models across internal process automation, software development, data analytics, network operations, cybersecurity, and brand and content work.
This is the first time a telco has committed to putting OpenAI at the center of its entire operations. Not as an experiment. As infrastructure.
Neil Tomkinson, U Mobile’s Chief Information Officer, tied it directly to their earlier 5G push: “Today, AI represents the next frontier of transformation”. They’re not just adopting AI. They’re betting the company on it.
And here’s what caught my attention: this fits a pattern GSMA identified in its Mobile Economy Asia Pacific 2026 report—operators across the region expanding into enterprise AI and sovereign cloud by partnering with hyperscalers and technology vendors. SoftBank with Oracle. Airtel with IBM. And now U Mobile with OpenAI.
The Network That Knows You Personally
While U Mobile was making headlines, Samsung and NTT Docomo were quietly doing something just as impressive.
On August 10, they announced they had successfully validated technology that uses AI to optimize mobile network quality for individual users. Not for everyone on a cell tower. For you. Specifically.
The system analyzes your service usage patterns and real-time wireless conditions. It predicts when your connection is about to degrade and automatically applies network settings tailored to you—like switching to the optimal frequency band before you even notice a problem.
Conventional systems apply the same configuration to everyone. This is different. This is AI that treats each user as an individual.
The results speak for themselves. In simulations using data from NTT Docomo’s commercial network, the frequency of service throughput degradation fell from 13.1% to 7.2%. That’s a 45% improvement.
The Chip That Cuts AI Costs to Nearly a Third
On August 11, Nvidia launched Nemotron 3.5 Lightning—a lightweight, customizable open AI model designed for autonomous agents.
Here’s the clever part: it uses a mixture-of-experts architecture with 30 billion total parameters, but activates only 3 billion for each task. That means it runs on a single GPU, including Nvidia’s DGX Spark or H100. And it delivers up to four times faster output than other open models in its class.
But the real story is what Nvidia released alongside it: NeMo Switchyard, an open-source routing library that automatically directs each AI task to the most capable and cost-effective model available. Internal tests showed it preserved frontier-level accuracy while reducing task-completion costs to nearly one-third of those incurred by using Anthropic’s Opus 4.8 alone.
And if that wasn’t enough, Nvidia also announced strategic partnerships with Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs and KKR to mobilize more than $500 billion in third-party capital for AI computing infrastructure.
The Model That’s Making Coding 75% Cheaper
On August 13, Google launched Gemini 3.7 Flash—its latest AI model designed for software coding and automated business tasks.
The company is pitching it as a lower-cost option for businesses building autonomous AI systems that can plan tasks, use software tools, and complete multi-step workflows. It has shown improved performance on coding tasks, including debugging, issue resolution, and production-ready code generation.
To drive adoption, Google is offering it at an introductory rate of 75 cents per million input tokens and $3.75 per million output tokens through the end of the year—half the original cost of Gemini 3.6 Flash.
The Nvidia Model That Could Reshape Open Source AI
And then there’s the one that genuinely surprised me. Nvidia is building Nemotron 4, a new AI model family with the largest version expected to have at least 1 trillion parameters—aimed at challenging top open-source models globally.
The chip giant is among the few major U.S. firms to release open-source models, which have drawn more attention this year as AI bills balloon and cheap Chinese models near the capabilities of top systems from leading American labs.
“Nvidia is investing in Nemotron because we believe every company and every country needs accessible frontier open models to strengthen safety and security, accelerate innovation, and provide a foundation they can rely on from one generation to the next,” said Kari Briski, vice president of generative AI.
The model could be ready as early as late fall.
These five developments—a telco betting its future on OpenAI, a network that optimizes for each user individually, a model that runs on a single GPU, a coding model that costs half as much, and a trillion-parameter open model that could reshape the industry—are not isolated events.
They are signals of a single trend: AI is moving from the lab into the core of how businesses operate.
U Mobile is embedding AI across its entire organization. Samsung and NTT Docomo are building networks that learn and adapt to individual users. Nvidia is mobilizing half a trillion dollars for AI infrastructure. Google is making coding AI affordable for every business. And Nvidia is building an open trillion-parameter model that could democratize access to frontier AI.
The infrastructure is being built. The economics are being rewritten. And the businesses that pay attention now will be the ones that define the next decade.
AI, #OpenAI, #Nvidia, #Samsung, #Google, #Telecom, #AIIinfrastructure, #AgenticAI, #Innovation, #TechNews, #BusinessTransformation, #MachineLearning, #Coding, #AIChips