Veo video tools, Gemini models and custom chips redefine racing – Azat TV

Google’s multifaceted AI offensive: Setting a new industry standard

In late 2025, the artificial intelligence landscape has changed dramatically. Google, once seen as a laggard behind OpenAI and Nvidia, has now emerged as a formidable leader in the AI ​​race. Recent moves such as the announcement of the Veo AI video generator, the launch of the Gemini 3 language model, and the release of a new generation of custom AI chips have not only impressed investors, but also raised deep questions about the future of technology, creativity, and trust.

Veo 3.1: A new era of video generation

Veo 3.1, Google’s latest AI-powered video tool, is making waves for its ability to create stunningly realistic videos from simple text prompts. The tool is available via Google Cloud and includes features such as richer native audio, more nuanced narrative control, and advanced image-to-video conversion capabilities (MM+M).

When testing Veo 3.1 against OpenAI’s Sora, reviewers found Google’s results to be more realistic and convincing. For example, when asked to create a TV ad for a headache medication, Veo produced an 8-second spot that reflected the style and sophistication of real-world pharmaceutical ads. The AI’s portrayal of humans was so realistic that it became really difficult to tell the difference between a computer-generated actor and a real human. It’s a step up from Sora, where visual cues often revealed the artificial nature of the video.

But such power comes with new challenges. Google’s Veo employs a set of safety filters designed to block the creation of harmful or misleading content. The company refuses to produce videos featuring public figures, living or dead, a safety measure not mirrored by its competitors. However, this system is not perfect. When asked to create ads based on disputed medical claims, such as the alleged link between acetaminophen and autism, Veo created content that was realistic and compelling, regardless of scientific validity. The tool’s ability to generate such material, even with initial errors, highlights the double-edged sword of AI-driven creativity.

Gemini 3 and Ironwood TPU: Google’s technological leap forward

Veo isn’t the only thing pushing Google to the forefront. The company’s Gemini 3 model, released in November 2025, outperformed the latest version of OpenAI’s ChatGPT on several industry benchmarks (CNBC). The rapid adoption of Gemini 3, which has increased from 450 million users in July to 650 million users by October, has garnered attention. The stock market reacted similarly. Shares of Google’s parent company Alphabet have risen more than 66% this year and more than 30% this quarter alone, outpacing traditional tech giants like Nvidia and Microsoft.

Fueling this momentum is Google’s seventh-generation custom chip, the Ironwood Tensor Processing Unit (TPU). Built specifically for AI workloads, these chips will be made available to customers outside of Google Cloud for the first time, potentially threatening Nvidia’s dominance in the AI ​​hardware space. Broadcom, Google’s TPU manufacturing partner, also benefited, with its stock up 65% since the beginning of the year. The market is showing signs of change. For the first time since 2016, stocks tied to Google’s AI stack are trading at a premium over stocks tied to the Nvidia and OpenAI ecosystems.

Competition is intensifying. Nvidia’s GPUs have long been the backbone of AI research and deployment, but as Google’s TPUs gain momentum, they are facing existential questions about their role. Meanwhile, OpenAI has reportedly declared a “code red” and delayed other projects focused on improving the quality and relevance of ChatGPT.

Industry upheaval: From creative teams to AI ethics

These technological leaps are not happening in isolation. The rise of tools like Veo and Gemini has sparked a heated debate about the future of the creative industries, advertising and public trust. Adam Daly, vice president of social at CG Life, says there is growing concern among marketers and patient advocates. “Many of these tools can be extremely dangerous,” Daly warned, noting that AI-driven campaigns risk eroding the trust and credibility that has been hard-won within the patient community.MM+M).

Daly, who works closely with rare disease patients, emphasizes that AI-generated stories cannot fully capture the nuances and individuality of a real patient’s journey. “People want real voices from real actors,” he says. Backlash against AI influencer campaigns, such as a leukemia awareness drive featuring virtual persona Lil Miquela, has sparked both mass engagement and criticism, indicating that the public remains wary of synthetic authenticity.

Meanwhile, Apple is closely monitoring these developments. In response to industry pressure and the company’s slow AI innovation, Apple recently named Amar Subramanya, a veteran of Microsoft and Google’s AI divisions, to lead its revamped AI efforts. Subramanya will oversee Apple’s fundamental model, AI safety, and a long-awaited overhaul of Siri, with the aim of closing the gap with more aggressive competitors.tech radar).

Competition and the way forward

The AI ​​sector is entering a new phase of specialization and fragmentation. For example, Amazon has launched its own Trainium3 AI chip, challenging both Nvidia and Google in the hardware space (bloomberg). As the market matures, the days of “AI lifted all boats” are coming to an end. Investors and industry watchers are now picking winners and losers based on technological differentiation, real-world adoption, and strategic vision.

For Google, the combination of Veo’s creative power, Gemini’s sophisticated language capabilities, and Ironwood TPU’s raw processing power made for an unparalleled trio. But the very strengths that drive this progress—realism, scale, and speed—also increase risk. As these systems become more pervasive and persuasive, misinformation, dehumanization, and ethical dilemmas will only increase.

Google’s rise in the AI ​​race is as much about technological excellence as it is about navigating the evolving ethical and social responsibilities that come with it. The industry is at a pivotal moment as AI-generated content becomes indistinguishable from reality. Do these advances increase creativity and efficiency, or undermine trust and credibility? As Google’s experience shows, the answer depends as much on human judgment as it does on machine intelligence.