Homechevron_rightArchiveschevron_rightTechnology

Archives

Explore Stories by Date

23 Dec 2025

Splat app uses AI to transform your photos into coloring pages, signaling shift in kid-friendly tech

The team behind photo-sharing app Retro has launched Splat, a new app that uses generative AI to convert any photo into customizable coloring book pages for kids. Parents can choose from artistic styles like anime, cartoon, or manga, then print or color on-screen. The app offers an ad-free alternative to cluttered websites, charging $4.99 per week for 25 pages or $49.99 annually for 500 pages.

2 Sources

Salesforce AI gains 6,000 customers while quietly pulling back on large language models

Salesforce's Agentforce added 6,000 enterprise customers in one quarter, reaching 18,500 total users and $540 million in annual recurring revenue. Yet executives now admit confidence in AI has dropped, with the company scaling back its use of large language models after facing reliability issues and laying off 4,000 support staff.

2 Sources

Yann LeCun vs Demis Hassabis: Public Clash Over General Intelligence Reopens AGI Debate

Meta's Chief AI Scientist Yann LeCun and Google DeepMind CEO Demis Hassabis engaged in a heated public disagreement over whether general intelligence exists. LeCun argues human intelligence is highly specialized, while Hassabis defends it as genuinely general. The debate carries significant implications for artificial general intelligence development across the AI industry.

3 Sources

Anthropic releases Bloom, an open-source tool to test AI models for bias and dangerous behaviors

Anthropic unveiled Bloom, an open-source agentic framework that automates the testing of AI behavior in frontier models. The tool evaluates problematic traits like sycophancy, self-preservation, and bias by generating custom scenarios and analyzing responses. Benchmark results across 16 models from Anthropic, OpenAI, Google, and DeepSeek reveal alignment challenges that could shape AI safety standards.

2 Sources

OpenAI admits prompt injection attacks on AI agents may never be fully solved

OpenAI acknowledges that prompt injection attacks targeting AI agents like ChatGPT Atlas represent a long-term security challenge unlikely to ever be completely resolved. The company is deploying an LLM-based automated attacker using reinforcement learning to identify vulnerabilities, but warns that AI-specific attack vectors continue to outpace traditional cybersecurity frameworks.

14 Sources

22 Dec 2025

Google upgrades Gemini for Home cameras with smarter face and pet recognition

Google is rolling out significant AI model improvements for Gemini for Home-enabled cameras from Nest and other compatible devices. The updates enhance Familiar Face detection by filtering out low-quality images, improve animal identification to better distinguish between cats and dogs, and provide more detailed descriptions in Home Brief when faces aren't fully visible.

3 Sources

OpenAI launches Your Year with ChatGPT, a Spotify Wrapped-style recap for AI chatbot users

OpenAI has rolled out its first year-end recap feature called Your Year with ChatGPT, offering users a personalized review of their interactions with the AI chatbot throughout 2024. The Spotify Wrapped-style experience includes conversation themes, usage statistics, user archetypes, AI-generated poetry, and custom images, available now to eligible users in select countries.

11 Sources

StubHub launches ChatGPT app that transforms how fans find tickets for concerts and sports

StubHub has integrated with ChatGPT to create a conversational ticket discovery experience. Fans can now search for concerts, sporting events, and theater tickets using natural language prompts instead of traditional filters. The AI integration remembers preferences throughout the conversation and displays real-time inventory and pricing, fundamentally changing the ticket buying experience.

3 Sources

Anna's Archive scrapes 300TB of Spotify data, sparking fears of AI training misuse

A piracy activist group called Anna's Archive claims it scraped 86 million music files and 256 million rows of metadata from Spotify, totaling nearly 300 terabytes. The shadow library says it's building a music preservation archive, but experts warn the dataset could fuel AI training on pirated content. Spotify confirmed unauthorized access and disabled accounts involved.

5 Sources

Uber and Lyft partner with Baidu to bring robotaxis to London in 2026 trials

Uber and Lyft announced separate partnerships with Chinese tech giant Baidu to launch robotaxi services in London next year. Testing begins in the first half of 2026 as part of a British government pilot program, positioning the UK as a frontline for self-driving taxi services. The move sets up direct competition with Waymo and other autonomous vehicle companies.

8 Sources

Google Gemini debuts in Samsung's AI Refrigerator, marking first home appliance integration

Samsung is bringing Google Gemini to its Bespoke AI Refrigerator Family Hub and AI Wine Cellar at CES 2026. The integration marks the first time Google's AI assistant appears in home appliances, promising enhanced food recognition that goes beyond the current 37 fresh and 50 processed food types. The AI Vision feature will automatically track inventory, suggest recipes, and manage shopping lists without manual registration.

4 Sources

21 Dec 2025

AI Hiring Creates a 'Doom Loop' as Job Seekers and Companies Struggle With Automated Screenings

Companies are rapidly adopting AI for talent acquisition, with 78% now using it in hiring processes. But the shift is creating unexpected problems: nearly nine in 10 HR professionals report being ghosted by candidates, and research shows AI use actually reduces hiring rates. Both job seekers and employers find themselves trapped in what one CEO calls a 'doom loop' of impersonal interactions.

3 Sources

Mac Cluster Computing Gets Major AI Boost with RDMA Support on Thunderbolt 5

Apple's macOS Tahoe 26.2 introduces RDMA over Thunderbolt 5, allowing Mac clusters to pool up to 1.5TB of unified memory for running massive AI models. Real-world tests show performance tripling with four Mac Studios, handling trillion-parameter models at a fraction of traditional supercomputer costs. But expiring memory supply agreements threaten Apple's cost advantage.

4 Sources

20 Dec 2025

OpenAI lets users adjust ChatGPT's warmth and enthusiasm in new personalization update

OpenAI rolled out new ChatGPT personalization settings that allow users to fine-tune the AI chatbot's warmth, enthusiasm, and emoji usage. The update addresses ongoing user feedback about conversational style after GPT-5.2's professional tone sparked complaints. Users can now choose between more, less, or default levels for these characteristics alongside existing personality options.

10 Sources

NotebookLM's Data Tables feature converts scattered notes into structured spreadsheets

Google has upgraded NotebookLM with Data Tables, its most requested feature that transforms messy notes into organized spreadsheets exportable to Google Sheets. Powered by Gemini 3, the AI-powered feature lets users synthesize information from various sources using natural language commands, eliminating manual data compilation for students, researchers, and business owners.

2 Sources

Get all updates into your inbox 💌, Stay Up-to-date for Free 🥳

Your Daily Dose of Curated AI News

Don’t drown in AI news. We cut through the noise - filtering, ranking and summarizing the most important AI news, breakthroughs and research daily. Spend less time searching for the latest in AI and get straight to action.

The Outpost

News

About