Models & Labs
Google I/O 2026: Gemini 3.5 and Omni Models Unveiled
The Verge AIhigh confidence
Why it matters
- →Gemini 3.5 Flash enhances AI speed and safety, crucial for user trust.
- →Omni Flash's multi-input video generation expands creative possibilities.
- →Google's AI advancements signal a shift towards more interactive and versatile digital tools.
At Google I/O 2026, the tech giant unveiled its latest AI models, Gemini 3.5 and Omni, marking significant advancements in AI technology. Gemini 3.5 Flash, now integrated into the Gemini app and Search, offers enhanced speed and safety, while Omni Flash introduces multi-input video generation capabilities. These updates reflect Google's ongoing efforts to expand AI functionalities across its ecosystem, providing users with more dynamic and interactive tools. The announcements underscore Google's strategic focus on AI innovation and its potential to transform user experiences.
Read originalMore from The Verge AI
© The Verge AIMarket & Regulationagents
Hackers Exploit Chatbot 'Personalities'
Hackers are increasingly exploiting the 'personalities' of AI chatbots, using conversational tactics rather than technical skills to bypass safety protocols. This new wave of attacks involves manipulating chatbots through persuasive dialogue, revealing a vulnerability in AI systems that rely on human-like interactions. Companies have patched obvious loopholes, but the challenge remains in balancing useful conversation with security. As AI systems become more integrated into daily life, the need for psychological insight in cybersecurity is growing, highlighting a shift towards social engineering in AI exploitation.
The Verge AI
© The Verge AIVideo & Creative AIvideo
Google's Omni AI Model Enhances Video Generation
Google's new Omni AI model is pushing the boundaries of video generation, allowing users to transform any input into creative video content. The model, part of Google's AI video platform Flow, offers improved consistency and real-world knowledge integration compared to its predecessor, Veo. Users can now create videos with minimal effort, though the results can still be unpredictable, with occasional AI glitches. While not perfect, Omni represents a significant step forward in making realistic video generation more accessible, albeit at a cost in terms of credits and potential editing iterations.
The Verge AI
© The Verge AIMarket & Regulationbusiness
Elon Musk's Grok AI Struggles to Gain Traction
Elon Musk's AI chatbot, Grok, is facing significant challenges in establishing itself within the AI market. According to a Reuters report, Grok's presence in government projects is minimal, appearing only three times, while competitors like OpenAI and Google are used extensively. Despite Musk's ambitious vision, Grok is mainly deployed for basic tasks and is overshadowed by more sophisticated models. This situation casts doubt on its role as a key component of SpaceX's future business strategy, especially given its controversial outputs and reliance on rival models for training. Grok's current trajectory suggests it may struggle to meet the high expectations set by Musk, raising questions about its long-term viability.
The Verge AI
More in Models & Labs
Models & Labsmodels
llama.cpp b9297 release enhances tensor support
The b9297 release of llama.cpp brings a notable enhancement with the introduction of NVFP4 MTP scale tensors, boosting its tensor processing capabilities. This update also integrates Qwen3.5 MTP tensors, which improves performance across a spectrum of hardware configurations, including Apple Silicon, Vulkan, and ROCm on Ubuntu, as well as CUDA on Windows. The release supports a wide array of architectures, from macOS to Linux and Windows, ensuring compatibility with both CPU and GPU setups. While there are no new model architectures, the inclusion of KleidiAI on Apple Silicon and ROCm 7.2 on Ubuntu highlights llama.cpp's commitment to optimizing for diverse environments. This update reinforces llama.cpp's role as a flexible inference runtime, catering to a broad range of hardware setups.
llama.cpp Releases
Models & Labsmodels
llama.cpp b9309 release fixes integer overflows
The b9309 release of llama.cpp tackles significant integer overflow issues in its perplexity calculations, co-authored by Stanisław Szymczyk. This update is vital for enhancing the accuracy and reliability of the model's performance metrics, which are crucial for developers. By resolving these overflows, the release ensures that users can depend on precise data outputs. This fix is a testament to the ongoing efforts to improve the tool's robustness, allowing developers to trust the integrity of their AI computations. While it might seem like a minor adjustment, it plays a critical role in maintaining the tool's reliability.
llama.cpp Releases
© The AI Daily BriefModels & Labsmodels
OpenAI Achieves Math Breakthrough
OpenAI has made a significant advancement in mathematical capabilities within its AI models.
The AI Daily Brief