Google Cloud is introducing AI tools for retailers, including Google Agentspace for building intelligent agents that enhance customer experiences and streamline operations. Vertex AI Search has been improved for better product discovery, while Connected Stores enhances the omnichannel experience. Retailers like Wayfair are leveraging these tools for efficiency. Partnerships with NVIDIA and Everseen focus on optimizing marketing and in-store experiences.
Nvidia launched Nemotron, a family of AI models for agentic tasks, including language and vision applications. The models come in three sizes: Nano, Super, and Ultra, optimized for different computing resources. Nvidia also introduced orchestration blueprints to manage multiple agents, partnering with various companies. New blueprints include converting PDFs to podcasts and summarizing videos, aiming to streamline the deployment of AI agents.
Meta is introducing AI-generated accounts on Instagram and Facebook to attract younger users. AI influencers like Aitana Lopez and Kimochii are emerging, while companies like 1337 allow user interaction with AI entities. However, there are growing concerns about misinformation and harmful content from these AI accounts. The rise of AI in social media could significantly alter user engagement and trust.
Get the most out of your Google Analytics data with Marketing Auditor – a powerful tool that helps you uncover and fix errors in minutes. Generate white-label reports packed with actionable insights and a customized action plan to optimize your analytics setup effortlessly. Customize reports with professional themes or your branding, and export them in editable formats like PowerPoint or Google Slides.
Google CEO Sundar Pichai stated that the AI model Gemini will be the main focus in 2025, which he called a critical year. He stressed the need for Google to accelerate its AI efforts and acknowledged the company must catch up. Pichai noted that while Gemini is gaining momentum, there is work to be done to close the gap and establish leadership, with scaling Gemini on the consumer side as the top priority for the year ahead.
xAI, Elon Musk’s AI company, raised $6 billion in a Series C financing round, boosting its valuation to $45 billion. Major investors included Andreessen Horowitz, Blackrock, and Kingdom Holdings, which invested $400 million. The company launched the AI supercomputer Colossus and developed models like Grok 2 and Grok 3. The funding will enhance infrastructure and product development focused on beneficial AI. xAI is also hiring for mission-driven roles.
2024 was significant for Google AI, featuring updates like Circle to Search and NotebookLM’s Audio Overviews. Key highlights included the launch of Gemini 1.5, advancements in health AI, and new generative tools. Major events included Google I/O, where AlphaFold 3 was introduced, and the Made by Google event showcasing new hardware. December marked the introduction of Gemini 2.0 and a quantum chip. Google aims to maintain this momentum into 2025.
OpenAI announced the o3 model family, successor to o1, during its “shipmas” event. The o3 and o3-mini models are claimed to approach AGI under certain conditions. OpenAI skipped o2 to avoid trademark issues. A preview for o3-mini starts soon, but o3 is not widely available yet. The model uses “deliberative alignment” and can adjust reasoning time for better performance. Internal benchmarks show o3 outperforms o1, but external validation is still needed.
Google has introduced an experimental AI model named Gemini 2.0 Flash Thinking Experimental in AI Studio, designed for multimodal understanding, reasoning, and coding in complex fields. It self-fact-checks but typically takes longer to provide answers. Initial tests showed issues, such as miscounting letters. While many companies are developing reasoning models, there are concerns about their high costs and whether they can sustain their performance over time.
Google's Gemini has updated guidelines for contractors, requiring them to evaluate AI responses even outside their expertise. Previously, they could skip prompts they weren't qualified to assess, but now they must rate what they understand and note their lack of expertise. This raises concerns about the accuracy of AI outputs on sensitive topics like healthcare, as contractors may misjudge complex issues. Skipping is now allowed only for missing information or harmful content.