Join our weekly Generative AI study group every Friday at 8 am PT! Dig into geospatial reasoning, share strategies for patch-text alignment, discuss KV cache routing, and more! For registration, head over to to sign up.

CNN REPORTS THAT 100–200 U.S. MILITARY ADVISERS ARE NOW IN SAUDI ARABIA PROVIDING REAL-TIME INTELLIGENCE, TARGETING AND GEOSPATIAL SUPPORT AGAINST THE HOUTHIS, MARKING A SIGNIFICANT ESCALATION IN U.S. INVOLVEMENT.

What if you could build an epidemic outbreak prediction model in minutes rather than weeks? Title: Planetary prediction engine: Automating global models via Earth AI URL: Google's Planetary Prediction Engine (PPE) automatically builds geospatial prediction models from natural language queries. For problems like public health, food security, and infectious disease outbreaks, it compresses weeks of manual data curation and feature engineering into minutes — no specialized engineering team required. Key Points 🌍 A three-stage automated pipeline — from query to model Stage 1 translates natural language into geographic constraints and retrieves covariates from Data Commons and Google Earth Engine. Stage 2 fuses structured covariates with embeddings from Population Dynamics Foundation Models and AlphaEarth. Stage 3 searches across regularized linear models, gradient-boosted trees, and MLPs. Each stage embeds a "Feature Gate" (leakage prevention) and an "Overfitting Guard Protocol" (self-correcting risk assessment). 🔬 Multimodal fusion outperforms either modality alone — across every experiment Combining structured statistical covariates with latent foundation model embeddings captures information neither source provides individually. This synergy was confirmed across all experiments, establishing a new architectural pattern for geospatial prediction. 📊 Validated across three high-stakes domains ・US CDC health indicators (21 types): R² 76.8% vs 60.0% baseline ・Nigeria food security downscaling: R² 66.1% vs 31.5% baseline (2× improvement) ・DRC Ebola outbreak prediction: Recall@10 83.3% (+10.3 percentage points vs Bayesian baseline) The real impact is democratization: humanitarian organizations and researchers without dedicated engineering teams can now get evidence-based predictions within minutes when it matters most. #EarthAI# #PublicHealth#

#DePIN# Leaders Just Dropped $43.74M in 30D Revenue ($12.35M) & #Helium# ($12.20M) dominating as real-world infrastructure turns into real cash on-chain. Decentralized compute, networks & geospatial the #tokenization# of physical assets is here. This is the backbone of #Web3#.

🚨 American military advisers and IRGC officers are now facing off through opposing sides of the escalating war in Yemen. CNN reports that more than 100 U.S. military advisers are operating in Saudi Arabia, with one U.S. official putting the number at roughly 200. At the same time, U.S. officials assess that hundreds of IRGC officers are inside Yemen working alongside the Houthis. This is not yet a direct United States-IRGC ground battle, but the two sides are increasingly embedded in opposing military efforts. American advisers are working inside a newly established Saudi joint forces command, providing intelligence sharing, geospatial targeting and real-time battlefield tracking technology reportedly comparable to the Pentagon’s Maven system. But there is a hard limit to that support. Axios reports that President Trump has ruled out the United States directly joining Saudi Arabia’s war against the Houthis. American forces are not flying strike missions, participating in Saudi strikes or refueling Saudi aircraft. So while American advisers are helping Saudi forces understand and track the battlefield, Saudi Arabia is still expected to fight the war itself. On the other side, the reported IRGC presence is far more extensive. U.S. officials believe hundreds of IRGC officers are assisting the Houthis, with one official assessing that they are ultimately working toward the ability to close Bab al-Mandab. That warning comes as the Houthis have seized Mocha, roughly 75 km north of the strait, and pushed government-aligned forces south toward Dhubab while expanding operations around the nearby islands. The geography explains why this confrontation matters far beyond Yemen. The Islamic Republic is already disrupting the Strait of Hormuz. Saudi Arabia has consequently become increasingly dependent on the Red Sea as an alternative route for its oil exports. Now the Houthis are advancing toward the other end of that route. Hormuz in the east. Bab al-Mandab in the west. The Islamic Republic does not need to physically close both waterways. Sustained missile, drone and anti-ship threats can disrupt shipping, increase insurance costs, force rerouting and put enormous pressure on global energy markets. Saudi Arabia therefore finds itself in an increasingly difficult position. It is absorbing Houthi ballistic missile and drone attacks while its Yemeni partners lose strategic territory, hundreds of IRGC officers are reportedly operating across the border, and American advisers are helping Saudi forces with intelligence and targeting. But the United States has drawn the line at joining the actual fight. What began as another chapter of the Yemen war is increasingly becoming another arena in the wider confrontation between the United States and the Islamic Republic, with American and IRGC personnel supporting opposing sides of the battlefield. For now, they are not directly fighting one another. But the distance between those two sides is getting increasingly smaller.

CNN reports that the U.S. has expanded intelligence and targeting support specifically for Saudi Arabia’s military campaign against the Iranian backed Houthis in Yemen. More than 100 U.S. military personnel are reportedly dedicated to that effort, with one U.S. official putting the number at roughly 200, operating as part of a newly established joint forces command. The U.S. is providing intelligence sharing and geospatial targeting support, including a battlefield tool that gives Saudi officers a near real time picture of the battlespace. However, U.S. personnel are not directly participating in Saudi strikes, refueling Saudi aircraft or providing other operational support. Sources characterized the effort as an extension of existing U.S.-Saudi cooperation and training. The expanded support comes as Houthi attacks against Saudi Arabia intensify, with U.S. officials saying Iran is providing substantial support to the group. CNN reports that U.S. officials believe hundreds of IRGC personnel are currently in Yemen and are helping the Houthis develop the capability to threaten the Bab el-Mandab Strait. Source:

An AI that can turn the right corner doesn't mean it can navigate an entire city. UrbanGround measures exactly that gap, at real scale in Hong Kong. UrbanGround: From Local Perception to Spatial Agency in a Real-Scale City 🏙️ Overview MLLMs have shown impressive local spatial skills — visual recognition, short-range movement, VQA — but whether those skills translate into sustained city-scale action was untested. UrbanGround is a physically-simulated replica of Hong Kong, built from geospatial data (OSM, satellite maps) in Unity, that evaluates MLLM spatial agency through 810 manually-verified task instances arranged across a five-level evaluation ladder. 🔍 The Problem Existing spatial reasoning benchmarks are mostly small-scale, synthetic, and short-range. None test the compounding of local decisions into kilometers-long routes, nor robustness to dynamic changes like weather, road closures, or pedestrian crowds. ⚙️ Methodology A three-layer framework (geospatial / simulation / agent) gives models first-person vision and an interactive map interface. Tasks escalate in five levels: ・Level 1: Visual recognition, orientation, active exploration ・Level 2: Short/long-range and instruction-constrained navigation ・Level 3: Implicit destination inference from description ・Level 4: Multi-task scheduling and route optimization ・Level 5: Dynamic adaptation to closures and pedestrians 📊 Results Visual recognition scores 77–93% — relatively strong. Orientation judgment drops to 23–58%. Short-range navigation success (~70%) collapses to near-zero for long-range tasks. Weather and lighting cut QA accuracy by 5–20 points. Pedestrian collision rates exceed 75% across all models. GPT-5.5 and Kimi-K3 lead overall, but all models share the same failure modes at scale. The core finding: "Local abilities do not compose into sustained exploration." Agents move locally compliant routes but cannot maintain spatial estimates beyond visible scenes or revise plans when routes become invalid. #MLLMs# #EmbodiedAI#