AnalysisBeyond the Transformer: How New AI Architectures Are Tackling the Memory Wall
State-space models, sparse attention, and hybrid memory systems are changing how engineers think about long-context artificial intelligence.
Technology, decoded
Clarity for a world built on code.
Independent reporting and practical analysis of the systems changing how we work, build, and live.
Intelligence
08Artificial Intelligence
Models, machine learning systems, research, and the infrastructure behind them.
Physical layer
08Hardware
Semiconductors, processors, computing systems, and the engineering inside devices.
Build systems
10Software
Programming languages, platforms, developer tools, and software architecture.
The newsroom
Latest intelligence
ExplainerENISA Launches EU Cyber Resilience Act Reporting Platform as Manufacturer Duties Begin
ENISA’s initial CRA Single Reporting Platform gives manufacturers a common route for qualifying actively exploited vulnerabilities and severe incidents; the launch does not make every Cyber Resilience Act obligation live.
NewsGitLab File-Read Flaw Draws Probes as CISA Adds It to Exploited-Vulnerability List
Probes have been reported against GitLab CVE-2026-85706, a critical self-managed-instance issue patched on 10 September; affected operators should check scope and fixed releases without presuming compromise.
ExplainerIndia Launches OneTag for FASTag Issuer-Bank Portability
OneTag is designed to let eligible NETC FASTag users change issuer bank while retaining the vehicle’s physical tag, Tag ID, and registration number; participation, verification, and availability details still matter.
ExplainerRetrieval-augmented generation: the production failure modes teams must test
A retrieval-augmented generation system can fail before a model writes a word. A practical test program should inspect the corpus, retrieval step, context assembly, response, and monitoring loop.
GuideAI agents need permission boundaries, audit logs, and reversible actions
An agent that can act on behalf of a person needs more than an instruction to be careful. Design boundaries around authorization, evidence, human review, and recovery before expanding its reach.
AnalysisHow to evaluate multimodal AI beyond benchmark scores
A multimodal system can perform well on a benchmark yet fail when cameras, microphones, documents, timing, or human workflows interact. Evaluation should follow the real task and its consequences.
TP / 01
Our point of view
Technology coverage should explain the system—not repeat the launch slide.
We connect architecture, constraints, evidence, and real-world use so readers can understand what changed and why it matters.
How we reportBrowse by field
Eight beats. One connected world.
01Artificial Intelligence
Models, machine learning systems, research, and the infrastructure behind them.
02HardwareSemiconductors, processors, computing systems, and the engineering inside devices.
03SoftwareProgramming languages, platforms, developer tools, and software architecture.
04CybersecuritySecurity research, defensive engineering, privacy, and digital resilience.
05Cloud & InfrastructureCloud platforms, data centres, networks, observability, and distributed systems.
06StartupsTechnology companies, product strategy, funding models, and operational lessons.
07GadgetsPersonal devices, wearables, displays, cameras, and practical buying context.
08ScienceComputing-adjacent discoveries in energy, space, climate, and biotechnology.