Polymarket Bettors Threaten Israeli Journalist & Blank Street Wants to Become “Gen Z Starbucks”

What It Covers

Morning Brew Daily covers four stories: the shift in US retail from goods to services (wellness now at 50.4% of leases), Polymarket bettors threatening an Israeli journalist over a $23M prediction market, Blank Street Coffee's pivot to a larger Gen Z-focused store format, and AI companies hiring improv actors for $74/hour to train emotional recognition models.

Key Questions Answered

• •Retail Shift: Services-based retailers surpassed goods-based tenants for the first time, claiming 50.4% of US commercial leases in 2025. Wellness — spanning fitness, beauty, nutrition, and mental health — drives nearly 30% of service leases, up from 20% a decade ago. Landlords replacing vacated big-box stores with gyms and spas report 20% higher rent from multi-tenant service configurations.
• •Prediction Market Risk: Polymarket's $23M trading volume on a single geopolitical event — whether an Iranian missile struck Israel on March 10 — created financial incentives for bettors to pressure journalists into changing published reports. Reporters covering conflict zones should be aware that their articles can now directly trigger multimillion-dollar prediction market swings, making them targets for coercion or bribery.
• •Blank Street's Format Pivot: Blank Street Coffee, valued at over $500M with $130M in VC funding, is tripling store size to 1,300 square feet and adding seating, selfie mirrors, and social-media-optimized pour stations. The strategy targets afternoon matcha drinkers — matcha's lower caffeine content drives later-day visits — and automated espresso machines free staff for customer engagement rather than drink production.
• •AI Data Scarcity: AI companies are sourcing increasingly specialized training data as models become "jagged" — highly capable at complex tasks but unreliable at simple ones. Handshake AI pays improv actors $74/hour to record emotional scenes for tone-recognition training. Scale AI, acquired by Meta for $14B, exemplifies how specialized data labeling has become a standalone, high-value industry within the AI supply chain.
• •AI Safety Red-Teaming: Anthropic is hiring chemical weapons and explosives experts with five-plus years of defense experience to stress-test its AI guardrails. OpenAI offers a comparable role paying $455,000 annually. Research from King's College London found AI models deployed tactical nuclear weapons in 95% of simulated geopolitical war games, underscoring why adversarial safety testing has become a funded, competitive hiring category.