Google's Gemini chatbot is enhancing user experience by integrating with Google Photos, allowing subscribers to generate personalized AI images. This feature, available to U.S. Google AI Plus, Pro, or Ultra subscribers, uses your photos for context to create more accurate images of yourself or loved ones without lengthy descriptions. Google assures users that their private photos are not used for training AI models. Meanwhile, Databricks and Google Cloud are deepening their partnership, integrating Google's Gemini models into the Databricks Lakehouse Platform. This collaboration, announced at Google Cloud Next 2026, aims to help businesses securely build and deploy generative AI applications using proprietary data, with Databricks now a first-party provider for Gemini models.
Further expanding its utility, Databricks has introduced a new connector that brings real-time, governed data from its Lakehouse platform directly into Google Sheets. This tool empowers business users, like those at Nubank, to analyze enterprise data within a familiar spreadsheet interface, ensuring data accuracy through Unity Catalog. In a different application of AI, Paperclip CEO Doda showcased an open-source platform where AI agents can orchestrate "zero-human" companies. This platform supports models like Claude and plans to integrate Gemini, demonstrating an AI CEO managing tasks such as hiring and development plans, with performance and costs tracked.
Moody's CEO Rob Fauber emphasized that building trust, rather than just advanced models, is the primary challenge for businesses adopting AI. He argues that curated, connected data is more crucial than the AI model itself, comparing data to a navigation system for an AI "car." Organizations must ensure their AI data is organized, normalized, and tested to make defensible decisions. Reflecting a broad approach to AI, Nvidia CEO Jensen Huang explained his company's strategy of investing widely across many tech companies, including major AI firms like Anthropic and OpenAI, as well as various startups, rather than attempting to pick specific winners in the evolving AI landscape.
AI is also proving useful in specialized searches, with assistants like Claude capable of finding financial advisors by analyzing web content and trust signals such as third-party reviews. However, the rise of AI also brings new threats. Aave Labs launched Aave Checkpoint, an AI-powered system combining automated analysis with human verification to enhance governance security for the Aave DAO. Conversely, a cybersecurity firm demonstrated how hackers are leveraging AI to breach networks, a concern echoed by U.S. officials warning of Iranian cyberattacks. The FBI reported nearly $21 billion in cybercrime losses last year, highlighting the growing threat. Disturbingly, Iran is reportedly using AI-generated Lego videos with rap music for propaganda, spreading antisemitism and conspiracy theories to millions online.
Key Takeaways
- Google's Gemini chatbot now uses Google Photos (opt-in) to create personalized AI images for paid subscribers (AI Plus, Pro, Ultra) in the U.S., without using photos for model training.
- Databricks and Google Cloud are enhancing their partnership, integrating Google's Gemini models into the Databricks Lakehouse Platform to help businesses build and deploy generative AI applications.
- Databricks launched a new connector to bring real-time, governed Lakehouse data directly into Google Sheets, enabling business users to analyze enterprise data easily.
- Paperclip CEO Doda demonstrated an open-source platform that orchestrates AI agents to run "zero-human" companies, supporting models like Claude and planning Gemini integration.
- Moody's CEO Rob Fauber states that trust, built on curated and tested data, is more critical for AI success than advanced AI models themselves, especially in high-stakes sectors.
- Nvidia CEO Jensen Huang advocates for a broad investment strategy across many AI tech companies, including Anthropic and OpenAI, rather than picking specific winners.
- AI assistants, such as Claude, can effectively find financial advisors by analyzing web content, profiles, and trust signals like third-party reviews.
- Aave Labs introduced Aave Checkpoint, an AI-powered security system combining automated analysis with human verification to enhance governance security for the Aave DAO.
- Hackers are utilizing AI to breach networks, contributing to nearly $21 billion in cybercrime losses last year, while U.S. officials warn of Iranian AI-powered cyberattacks.
- Iran is reportedly using AI-generated animated Lego videos with rap music for propaganda, spreading antisemitism and conspiracy theories to millions online.
Gemini uses Google Photos for personalized AI images
Google's Gemini chatbot can now create personalized AI images by accessing your Google Photos. If you opt in, Gemini uses your photos and labels to understand prompts better and generate more accurate images. This feature simplifies creating images of yourself or your family without needing long descriptions. Google states that your photos are not used for training AI models, only for prompt context. Personal intelligence is off by default and currently available for paid Google AI subscribers.
Gemini app creates custom images with your photos
The Gemini app now uses your Google Photos and personal preferences to create custom AI images. You can ask for pictures featuring yourself or loved ones, and Gemini will use your photos for context without needing long prompts. This feature is rolling out to U.S. subscribers of Google AI Plus, Pro, or Ultra. Gemini does not train its models on your private photo library, ensuring your privacy. You can refine results or swap reference photos to maintain creative control.
Databricks and Google Cloud boost AI partnership
Databricks and Google Cloud are strengthening their partnership by integrating Google's Gemini models into the Databricks Lakehouse Platform. This collaboration aims to help businesses build and deploy generative AI applications more easily and securely. The partnership, announced at Google Cloud Next 2026, focuses on using proprietary data with AI. Databricks is now a first-party provider for Gemini models, allowing users to manage the entire AI lifecycle. This integration has seen significant growth, powering various AI use cases for enterprise clients.
Databricks connects Lakehouse data to Google Sheets
Databricks has launched a new connector that brings real-time, governed data from its Lakehouse platform directly into Google Sheets. This tool allows business users to access and analyze enterprise data within a familiar spreadsheet interface without manual exports. The connector ensures data accuracy and governance through Unity Catalog, with options for automatic refreshes or manual updates. Nubank is already using this integration to empower non-technical users with data exploration capabilities. The Databricks connector for Google Sheets is now available on the Google Marketplace.
Paperclip CEO shows AI agents running zero-human companies
Paperclip CEO Doda demonstrated an open-source platform that orchestrates AI agents to run 'zero-human' companies. The platform allows users to create AI agents, assign them roles, and manage tasks within a company structure. It supports various AI models like Claude and plans to integrate Gemini. The demonstration showed an AI CEO managing hiring and development plans, with performance and costs being tracked. Paperclip aims to enable advanced automation and efficiency through AI collaboration, with plans for enhanced features and scalability.
Moody's CEO: Trust, not models, is key for AI success
Moody's CEO Rob Fauber argues that trust, not just advanced AI models, is the main challenge for businesses adopting AI. He believes that curated, connected data from multiple sources is more important than the AI model itself. Fauber compares AI models to cars and data to navigation systems, emphasizing the need for reliable intelligence to guide decisions. He states that organizations must ensure their AI data is organized, normalized, and tested against real outcomes to build trust and make defensible decisions, especially in high-stakes sectors like finance.
AI can find financial advisors using web searches
An AI assistant like Claude can help find financial advisors by performing web searches and analyzing information from sources like LinkedIn, advisor websites, and directories. The AI looks for specific keywords in profiles and content to match advisors with user requests. Clear descriptions of services, target clients, and published content are crucial for AI to identify suitable advisors. Third-party reviews and testimonials also act as strong trust signals for AI in recommending advisors. The AI prioritizes plain language and specific details over general statements.
Nvidia CEO Jensen Huang explains why they invest broadly in AI
Nvidia CEO Jensen Huang believes in investing broadly across many tech companies rather than picking specific winners in the AI space. He cites Nvidia's own history, where it was initially not expected to succeed, as a reason for this strategy. Huang stated that it's not Nvidia's role to pick winners and that they invest in AI labs, biotech, robotics, and self-driving companies. Nvidia has made significant investments in major AI firms like Anthropic and OpenAI, as well as various startups.
Iran uses AI Lego videos for propaganda
Iran is reportedly using AI-generated animated videos featuring Legos and rap music for propaganda aimed at denigrating the United States. These videos have gone viral on social media, reaching millions of views by exploiting cultural elements and legitimate concerns about U.S. politics. Experts note that new technologies allow Iran to create more culturally appealing propaganda than before. The videos contain antisemitism and conspiracy theories, and their spread is occurring as U.S. programs to counter foreign influence have been cut. U.S. officials have condemned the videos as disgusting and detached from reality.
Aave Labs launches AI security system Checkpoint
Aave Labs has launched Aave Checkpoint, an AI-powered system designed to enhance governance security for the Aave DAO. This system adds a multi-layered review process for all proposals before they are executed on the blockchain. Checkpoint combines automated AI analysis with mandatory human verification, working alongside existing manual reviews. The goal is to strengthen security and ensure thorough review of every proposal. This initiative is part of Aave's structured approach to managing governance risks.
Hackers use AI to breach networks, cybersecurity firm warns
A cybersecurity firm has demonstrated how hackers are using artificial intelligence to breach business and consumer networks. U.S. officials are also warning about Iranian cyberattacks targeting these groups. A recent FBI report indicated that cybercrime losses reached nearly $21 billion last year. The report highlights the growing threat of AI-powered cyberattacks and offers advice on how individuals and businesses can protect themselves from these breaches.
Sources
- Gemini can now create personalized AI images by digging around in Google Photos
- New ways to create personalized images in the Gemini app
- Databricks & Google Cloud Forge AI Partnership
- Databricks Connects to Google Sheets
- Paperclip CEO on Building Zero-Human Companies
- Moody’s CEO: AI has a trust problem – better models won’t fix it
- How AI Is Changing How People Find Financial Advisors
- Jensen Huang explains why Nvidia does not try to pick AI winners
- Banned AI-generated Iran propaganda videos using Legos have gone viral
- Aave Labs Launches Checkpoint, AI-Powered Governance Security System: Aave Labs
- Cybersecurity firm shows how hackers are using AI to breach networks
Comments
Please log in to post a comment.