SRMED
Englishعربي
Englishعربي
Google used its annual I/O conference to show off a broad new wave of artificial intelligence products, but one of the company’s most closely watched moves was what it did not do: it held back the launch of Gemini 3.5 Pro, its next major model, and said it would arrive next month instead. The delay disappointed some attendees, according to Business Insider, but it also underscored how carefully Google is managing the rollout of its biggest AI systems as competition intensifies across the industry.
At the same event, Google unveiled Gemini Spark, a personal AI agent designed to work in the background, draft emails, assemble documents, monitor inboxes and eventually make purchases even when a laptop is closed, as reported by VentureBeat and Business Insider. Google also said it was tightening the connection between Search and its AI tools, introducing a major overhaul of the search box and new “information agents” that can watch topics over time and alert users when something changes, according to TechCrunch and Business Insider. Taken together, the announcements showed Google pushing AI deeper into everyday consumer products rather than relying on a single headline model release.
That is why the absence of Gemini 3.5 Pro drew so much attention. Google executives did not publicly give a detailed explanation for the delay, but the decision appears to fit a broader strategy of staggering major launches while the company builds out the surrounding ecosystem. Recent reporting from Fast Company described Google’s AI direction as increasingly focused on products for search, coding, design and media generation, while TechCrunch noted that the company is also making it easier for users to build Android apps in minutes with its AI Studio tools. In other words, Google seems to be emphasizing the platform around the model, not just the model itself.
The timing also reflects a crowded and fast-moving market. Rival companies and investors are racing to define how AI will be used in the workplace and in consumer products. Bloomberg reporting this week highlighted debates over how U.S. and Chinese AI models are diverging, while Alibaba announced a new AI chip for training and inference, showing how major players are investing across the full stack of AI hardware and software. OpenAI, meanwhile, announced a new AI lab in Singapore, adding to the sense that the global race is shifting from model launches alone to infrastructure, partnerships and deployment.
Google DeepMind chief executive Demis Hassabis used the conference to press a separate point: that companies should not use AI as a blanket justification for layoffs. According to WIRED and India Today, Hassabis argued that firms should use AI-driven productivity gains to expand what they do, not simply cut jobs. His comments landed at a time when AI’s labor impact is under intense scrutiny, with many workers and analysts worried that new systems will automate parts of entry-level and routine work before the broader gains show up elsewhere.
The conference also highlighted Google’s push into more specialized AI tools. Fast Company reported that the company is building out products for creatives, including image and video generation, while TechCrunch said Google is positioning new design tools as accessible to teachers, small business owners and nontechnical users. Business Insider also noted that Google described rising usage of its AI products, including a joke from CEO Sundar Pichai about “tokenmaxxing,” a reference to heavy AI consumption. All of this suggests Google is trying to prove that its AI business is more than a single model release; it is a layered effort spanning Search, agents, coding, design and media creation.
For users and developers, the immediate takeaway is that Google’s AI push is widening even as its next flagship model waits in the wings. The company is asking people to adopt new agents and tools now, while leaving room for a more powerful Gemini release next month. That approach may frustrate some of the conference crowd, but it also reflects the reality of today’s AI race: success is increasingly measured not just by who has the biggest model, but by who can turn it into products people actually use.