Japan’s AI play and Taiwan’s space ambitions
The Financial Times reports on Japan's AI strategy and Taiwan's aerospace initiatives, focusing on broader Asian tech trends.
Financial Times Technology·23 Jul 2026
Use this view to inspect the underlying evidence corpus. For ranked developments, decision posture and interpretation, use Signals.
Raw feed or Signals?
Raw feed is chronological evidence. Signals ranks and interprets material change.
The Financial Times reports on Japan's AI strategy and Taiwan's aerospace initiatives, focusing on broader Asian tech trends.
Chinese AI models narrowed the performance gap with US counterparts to 6% in June from 9% in May, according to Bloomberg Intelligence.
ServiceNow invested $40 million in BusinessNext, an Indian banking software specialist, to expand its AI-powered financial services offering.
Thailand received $40.6 billion in foreign direct investment applications in H1 2026, up 80% year-on-year, driven by AI and data center projects.
Laguna S 2.1 model released, claimed to be cheaper than Deepseek v4 Flash and to outperform Deepseek v4 Pro in benchmarks.
STMicroelectronics forecast Q3 sales below analyst expectations, leading to a stock drop and dimming hopes for AI data center-driven recovery.
Poolside AI claims to have developed a 'model factory' capable of training custom 118B MOE models that outperform larger open-source alternatives.
Anthropic released Claude Opus 4.7, showing performance improvements over version 4.6 in advanced software engineering tasks.
Anthropic announced the release of Claude Opus 4.5, updating its frontier model family with improved capabilities.
Francisco Partners co-founder DJ Deb states AI will not kill the software sector as the firm raises $21bn for new investments.
Research explores using "model-selection Bayesian wind tunnels" to enable transformers to perform Bayesian model selection, identifying correct hypothesis classes.
New benchmark, CruiseBench, for aero-engine remaining useful life (RUL) prediction uses real-flight profiles, extending N-CMAPSS with more realistic data.
New dataset and benchmark, Air Quality Arena, evaluates time-series foundation models for large-scale, multi-region air quality forecasting.
Research investigates explainability in continual learning for time series forecasting, particularly with Experience Replay strategies in non-stationary environments.
Research proposes Unified Geometric Surgery (SUM) to address spatial and temporal interference in Federated Class Incremental Learning (FCIL) for distributed, evolving AI systems.
STN-TGAT proposes a Transformer-Graph Attention Network for stock ranking and portfolio construction, combining temporal dynamics and cross-sectional dependencies.
Research proposes LAARA, a framework for adaptive rank allocation in parameter-efficient fine-tuning (PEFT), addressing the suboptimality of uniform rank.
Research explores using Graph Neural Networks to predict groundwater arsenic concentrations, aiming to improve spatial risk assessment for public health.
Prefix-GRPO is a reinforcement learning framework that reuses teacher trajectories for training small language models in long-horizon tasks.
New research proposes Memory Merge DQN, an alternative target network update for Deep Q-networks to improve stability and preserve value function structure.
Research explores leveraging offline supervision to improve reinforcement learning for vision-language-action (VLA) models, addressing out-of-distribution (OOD) performance.
Research investigates statistical label fusion for consensus segmentation, finding STAPLE underperforms voting under common conditions.
CoTFormer proposes a recurrent transformer architecture that formalizes Chain-of-Thought as preserving intermediate states for explicit reasoning.
Research demonstrates an agentic AI system, backed by a frozen LLM, can perform NMR elucidation comparable to graduate-level chemistry students.
Research demonstrates that the discrepancy in optimal population sizes for Evolutionary Strategies (ES) in LLM fine-tuning is primarily due to reward design and normalization, not population size itself.
Research presents an Adaptive Multi-Expert Graph Transformer for interpretable EEG-based diagnostics, modeling dynamic functional connectivity.
New research challenges current machine unlearning evaluation methods, finding they may favor models that retain forgotten data, proposing a restoration-based audit.
Open-access dataset for marine engine predictive maintenance released, featuring data from controlled fault experiments on a testbed.
Research introduces a framework to audit silent safety failures in tool-augmented LLM agents, focusing on malformed tool responses.
Research audits machine learning models for predicting crypto extrema on Binance Spot, finding them unprofitable after costs.
© 2026 OneBench: AI Insights. All rights reserved.
Evidence before opinion