Newly tracked
Newly tracked — the first measured comparison begins once two weekly collections exist.
AcceleratingSee the evidence →
Weekly radar · 2026-W34
Follow technology. Not the noise.
There is more tech content than anyone can keep up with, but staying updated has never been more important. Tech Signal uses data and traction signals to identify which technologies are actually gaining momentum across articles, tutorials and videos.
- technologies monitored
- 55
- technologies monitored
- showing recent activity
- 49
- showing recent activity
- worth your attention this week
- 8
- worth your attention this week
- updated · Week 34 · 17–23 August 2026
- Weekly
- updated · Week 34 · 17–23 August 2026
What changed this week
Changes since 10 August 2026
Newly tracked
Newly tracked — the first measured comparison begins once two weekly collections exist.
GrowingSee the evidence →
Newly tracked
Newly tracked — the first measured comparison begins once two weekly collections exist.
GrowingSee the evidence →
Newly tracked
Newly tracked — the first measured comparison begins once two weekly collections exist.
StableSee the evidence →
This week’s signal
Measurement has caught up with the launch estimates, and the gap is the story. Coding agents are the only theme where adoption is visibly compounding — installs of the Claude Code CLI rose 51% in two weeks, while every other tracked theme moved less than 5%. Elsewhere the signal is research-heavy and adoption-light: vector search drew 190 arXiv papers against falling package downloads. Attention and adoption are not the same measurement.
Tech Signal editorial · 2026-08-22 · reviewed before publication
Trending technologies
Search all themes →Ranked by Traction Score — adoption data, growth, source breadth, enterprise relevance and learning value. Not press coverage.
AI Coding Agents
82
traction
AcceleratingMaturityAdoptDoAdopt
Software teams are delegating whole coding tasks to agents, not just autocompleting lines.
Managers + DevelopersOpen signal →
Agent Interoperability Protocols
75
traction
GrowingMaturityEmergingDoLearn
Open protocols like MCP are becoming the connection layer between AI systems, tools and companies.
Managers + DevelopersOpen signal →
Small Language Models & On-Device AI
68
traction
GrowingMaturityEmergingDoExperiment
Models small enough to run on laptops and phones now handle real production tasks.
Managers + DevelopersOpen signal →
AI Inference Infrastructure
66
traction
StableMaturityAdoptDoExperiment
Serving engines decide whether AI features are profitable — same GPUs, several times the throughput.
DevelopersOpen signal →
Vector Search & Agent Memory
66
traction
StableMaturityAdoptDoAdopt
Past the hype, retrieval quality — not model choice — is what limits AI answers on your own data.
DevelopersOpen signal →
Memory-Safe Systems Programming
63
traction
StableMaturityAdoptDoLearn
Governments and flagship codebases are making memory-safe languages the default for new systems code.
DevelopersOpen signal →
WebAssembly Beyond the Browser
62
traction
GrowingMaturityEmergingDoExperiment
A safe, fast way to run code anywhere — now spreading to servers, plugins and AI-agent sandboxes.
DevelopersOpen signal →
Post-Quantum Cryptography
58
traction
GrowingMaturityEmergingDoLearn
The migration to quantum-safe encryption has started — driven by regulators, not quantum computers.
ManagersOpen signal →
What should I pay attention to?
AI Coding Agents
Budget for licences and guardrails, update code-review policy, and ask your leads how much merged code is agent-written today.
Agent Interoperability Protocols
Ask which of your systems expose MCP endpoints, and put identity and audit for agent access on the security roadmap.
Small Language Models & On-Device AI
Evaluate the hybrid pattern — local models for routine tasks, cloud for hard ones — and ask what the inference bill looks like if most requests run locally.
Each theme page includes business impact, risks and the questions worth asking your engineering team.
Learn next
Four good uses of your next free hour — picked from this week's themes.
Quick overview
Ollama model library — what actually runs on a laptop
Small Language Models & On-Device AI · Ollama
Watch
Claude Code - Full Tutorial for Beginners
AI Coding Agents · YouTube
Deep dive
NIST Post-Quantum Cryptography project — why migration is happening
Post-Quantum Cryptography · NIST CSRC
Try it yourself
Connect your first MCP server to Claude Desktop (user quickstart)
Agent Interoperability Protocols · modelcontextprotocol.io
Editor’s note: First week scored from collected data rather than the seeded launch baseline. No collection ran in week 33, so growth is measured across a two-week window and normalised to a weekly rate. Where a score moved against the launch estimate, that mostly reflects measurement replacing editorial judgement — not the technology changing.