Delhi
--:--:--
London
--:--:--
Paris
--:--:--
Hong Kong
--:--:--
New York
--:--:--
THE PULSE WIRE
🔴 BREAKING AI: Autonomous Coding Agents cross 70% threshold on SWE-bench as test-time compute scales. 🛰️ ASTROPHYSICS: JWST confirms carbon-bearing molecules on habitable-zone exoplanet K2-18b. ⚡ FUSION: Commercial Tokamaks achieve 20-Tesla field with REBCO high-temperature superconductors. 🧬 BIOLOGY: In-vivo epigenetic reprogramming successfully reverses cellular aging biomarkers in mammalian models. 🔬 QUANTUM: Silicon CMOS spin qubits reach 99.9% gate fidelity for fault-tolerant error correction. 🚀 AEROSPACE: Rotating Detonation Rocket Engine completes supersonic flight test with 25% efficiency jump.

AI & Software Intelligence Hub

The Science Man AI & Software Intelligence Hub brings together our latest AI reporting, verified model records, software releases, company knowledge graph, original test lab and practical calculators.

Daily AI & Software Brief

The latest AI, Google software, apps and emerging-technology reporting from The Science Man.

  1. Amodei’s AI Slowdown Plan Faces the China Test
  2. What Is Physical AI? How Robots Are Learning to Work in Our World
  3. 2-Nanometer Gate-All-Around (GAA) Silicon Enters Mass Production in Leading Semiconductor Fabs
  4. The End of Copper Interconnects: How Co-Packaged Optics Is Solving the AI Bandwidth Bottleneck
  5. Solid-State Battery Field Tests Show 1,000-Kilometer Range with 12-Minute Ultra-Fast Charging
  6. Autonomous Coding Agents Achieve 70% on SWE-bench as Frontier AI Shifts to Test-Time Compute
  7. Liquid Neural Networks vs. Transformers: How Adaptive Time-Continuous AI Is Redefining Robotics
  8. Global Frontier AI Accord: International Regulators Establish Compute Threshold Protocols
  9. The Road to AGI: Tracking Compute Scaling, Reasoning Models, and the 2029 Horizon
  10. Recursive Self-Correction and Synthetic Thought: Can Frontier AI Formulate New Laws of Physics?

Open AI & Software Intelligence Hub · Google AI & Software

AI Model Tracker

Verified model records with official-source links and last-checked dates.

Software Release Radar

Officially sourced updates from AI, software and technology products.

No verified releases have been published yet.

AI Company Knowledge Graph

Entity pages that connect company background, models, releases and related reporting.

Original Test Lab

Transparent, repeatable tests with methods, inputs and limitations published alongside the result.

Test reports will appear here when original experiments are published.

AI Cost Calculator

Estimate usage cost from verified pricing you enter. The calculator does not assume vendor prices.

Estimated cost: 0.00

Enter the current official vendor price before using the result for a purchasing decision.

Weekly AI Intelligence

A seven-day view of AI, apps, software and emerging technology published by The Science Man.

1Amodei’s AI Slowdown Plan Faces the China TestSep 13, 20262What Is Physical AI? How Robots Are Learning to Work in Our WorldSep 13, 202632-Nanometer Gate-All-Around (GAA) Silicon Enters Mass Production in Leading Semiconductor FabsSep 10, 20264The End of Copper Interconnects: How Co-Packaged Optics Is Solving the AI Bandwidth BottleneckSep 10, 20265Solid-State Battery Field Tests Show 1,000-Kilometer Range with 12-Minute Ultra-Fast ChargingSep 10, 20266Autonomous Coding Agents Achieve 70% on SWE-bench as Frontier AI Shifts to Test-Time ComputeSep 10, 20267Liquid Neural Networks vs. Transformers: How Adaptive Time-Continuous AI Is Redefining RoboticsSep 10, 20268Global Frontier AI Accord: International Regulators Establish Compute Threshold ProtocolsSep 10, 20269The Road to AGI: Tracking Compute Scaling, Reasoning Models, and the 2029 HorizonSep 10, 202610Recursive Self-Correction and Synthetic Thought: Can Frontier AI Formulate New Laws of Physics?Sep 10, 2026

Open the editorial AI Tool Directory · Open Weekly AI Intelligence