openai launches google while nvidia expands its platform

Key Takeaways

• AI agents now research hotels for corporate buyers before any human contact, scanning reviews and rates to create negotiation briefs.
• Booking.com's 2026 report shows 42% of travelers go off-peak due to climate anxiety, with searches for cooler destinations like Norway up 33%.
• Accor reported Q1 2026 revenue of 1.313 billion euros, up 2.3%, and RevPAR growth of 5.1%, but the UAE saw a 9% drop after the Middle East conflict began.
• Grok-4.20 Expert Mode and OpenAI GPT 5.4 Pro Vision tied for the smartest AI model in TrackingAI's April 2026 benchmark, both scoring 145 on the Mensa Norway test.
• Google's Gemini 3.1 Pro Preview scored 141 on the Mensa Norway test, closely behind the top models.
• OpenAI launched ChatGPT Images 2.0, creating highly realistic pictures at up to 2K resolution, raising concerns about misinformation and deepfakes.
• NVIDIA released a new version of FLARE that lets developers convert existing training scripts into federated clients with only 5 to 6 lines of code.
• Brivo expanded to the UK with a new London office and AI security products; CEO Dean Drako said the UK is Brivo's fastest-growing European market.
• Lafayette College's counseling center warned against using AI chatbots for mental health, citing lack of human connection and privacy concerns.
• The Trump administration and Congress are pursuing AI legislation that would override state AI laws, a move opposed by technology, labor, and civil rights groups.

AI agents research hotels before buyers call, climate anxiety shifts travel, Accor reports growth

AI agents are now researching hotels for corporate buyers before any human contact, scanning reviews and rates to create negotiation briefs. A hotel sales director shared how a buyer used AI to benchmark rates and attach a TripAdvisor summary, giving a 24-hour deadline. Meanwhile, Booking.com's 2026 report shows 42% of travelers now go off-peak due to climate anxiety, with searches for cooler destinations like Norway up 33%. Accor reported Q1 2026 revenue of 1.313 billion euros, up 2.3%, and RevPAR growth of 5.1%, but the UAE saw a 9% drop after Middle East conflict began in late February.

Hotel sales teams now compete with AI agents before buyers call

A corporate travel buyer who booked over 400 room nights sent a rate request email benchmarked against three competitors with a TripAdvisor summary, all compiled by her AI. The author argues that AI agents now evaluate hotels before any human conversation, scanning reviews, rates, and sales director profiles. PwC data shows 76% of millennials are likely to use AI agents for travel recommendations. The piece urges sales leaders to audit AI-facing data weekly, write content for algorithms, and train teams for pre-qualified buyers.

Tattoo artists debate whether AI is friend or foe

Tattoo artists in Charleston are discussing how AI affects their trade. Some use AI to create reference photos but insist on drawing the final design themselves. Clients sometimes bring AI-generated images, but artists refuse to copy them directly, saying AI art all looks the same. They emphasize that tattooing relies on human connection and collaboration, which AI cannot replace. A communications professor said AI-generated tattoos would cheapen the meaning of the art.

College counseling center warns against using AI chatbots for mental health

Lafayette College's counseling center released an advisory warning students against using AI chatbots for mental health support. The center's prevention coordinator said AI lacks human connection and may not challenge unhealthy thought patterns. A licensed psychologist noted AI cannot replicate the special interaction between patient and therapist. Psychology professors raised concerns about privacy, data use, and AI validating unhealthy behaviors. The center recommends using free anonymous support groups instead.

Groups oppose federal preemption of state AI laws

The Trump administration and Congress are pursuing AI legislation that would override state AI laws, a move opposed by technology, labor, and civil rights groups. The administration made preemption a key part of its National Policy Framework for Artificial Intelligence. Representatives from the Center for American Progress, AFL-CIO Tech Institute, and other organizations will discuss why preemption is not the right path. They argue that meaningful federal AI legislation should focus on other priorities.

Five reasons to be careful using ChatGPT for financial advice

Many people use ChatGPT for money questions, but experts warn it has serious limitations. The AI can confidently give incorrect answers and may affirm your existing beliefs instead of challenging them. To get useful results, you must share sensitive financial data, which raises privacy concerns. Unlike human financial advisors, chatbots have no legal accountability or regulations to protect users. A human expert should always review any financial plan before taking action.

Brivo expands to UK with new London office and AI security products

Brivo announced its expansion into the UK market with a new regional office in London. The company will showcase its AI-first unified security platform at The Security Event. CEO Dean Drako said the investment includes local sales, marketing, technical support, and training spaces. Brivo Managing Director for EMEA Rishi Lodhia noted the UK is Brivo's fastest-growing European market due to high demand for AI-driven security upgrades.

NVIDIA FLARE makes federated learning easier with less code changes

NVIDIA released a new version of FLARE that makes federated learning easier for developers. Federated learning allows AI training across multiple sites without moving sensitive data. The new API lets developers convert existing training scripts into federated clients with only 5 to 6 lines of code. It supports no data copy, compliance controls, and privacy techniques like encryption. The goal is to help teams move from pilot projects to production without major rewrites.

Grok and OpenAI tie for smartest AI model in 2026 ranking

TrackingAI's April 2026 benchmark ranks Grok-4.20 Expert Mode and OpenAI GPT 5.4 Pro Vision as the smartest AI models, both scoring 145 on the Mensa Norway test. Gemini 3.1 Pro Preview follows closely at 141. The top tier is crowded with several models separated by only a few points. Scores have risen sharply from 2025, showing rapid improvement in AI reasoning. The test uses visual pattern puzzles, so results vary depending on how questions are presented.

Google Gemini offers eight tips for spring cleaning and organizing

Google's Gemini AI can help with spring cleaning and home organization tasks. Users can ask Gemini to generate personalized cleaning checklists, audit clutter by uploading photos, and clear the fridge by identifying leftovers. It can also troubleshoot home repairs, streamline shopping lists, and suggest room makeovers. For plant care, Gemini can diagnose issues and advise on watering and light. The tool can also help declutter your inbox by archiving and creating tasks.

OpenAI releases ChatGPT Images 2.0 with highly realistic pictures

OpenAI launched ChatGPT Images 2.0, an upgraded image generator that creates far more realistic pictures than before. The new system produces images at up to 2K resolution with sharper text and supports multiple aspect ratios and foreign languages. It can search the internet in real time and use a slower thinking model for more refined results. Researchers worry the realistic images could make it harder to spot fakes and may fuel misinformation and deepfakes. OpenAI says the tool is for creativity and is working on safeguards.

Sources

• HN Brief: AI Agents Are Researching Your Hotel Before Buyers Call, Climate Anxiety Shifts 42% of Travelers Off-Peak, Accor Posts 5.1% RevPAR Growth Despite Middle East Headwinds
• The Agentic AI Guest: Why Your Hotel Sales Team Is Now Competing With Chatbots
• Is artificial intelligence a friend or foe to local tattoo artists?
• Counseling center responds to rise of AI chatbots
• Preemption Of State AI Laws Is Not The Path Forward
• 5 Reasons to Think Twice Before Using ChatGPT—or Any Chatbot—for Financial Advice
• Brivo Announces UK Expansion, Showcases AI-First Unified Security at The Security Event
• Federated Learning Without the Refactoring Overhead Using NVIDIA FLARE
• Ranked: The Smartest AI Models of 2026
• 8 Gemini tips for organizing your space (and life)
• AI Images Just Got Way Too Real