Tuesday, April 28, 2026
AI Radar — Live Dashboard
Filter
Significant Developments
The most important signals — click any card for deeper context
China's NDRC blocks Meta's $2B acquisition of Manus AI — formal security review prohibits foreign acquisition
SignificantChina's National Development and Reform Commission issued a formal security review decision on 2026-04-27 prohibiting Meta's acquisition of the Manus AI project. The decision langu...
Dirac open-source agent scores 65.2% on TerminalBench — beats Google's official 47.8% and closed-source Junie CLI's 64.3%
SignificantGodelNumbering published results on 2026-04-27 from a fully leaderboard-compliant run of their open-source Dirac agent on the TerminalBench coding-agent benchmark (TerminalBench ev...
Also Worth Knowing
YourMemory: AI memory with biological decay — 52% Recall@5 on LoCoMo, 84% token reduction vs static RAG
NotableShow HN from SachitRafa (2026-04-26): implements the Ebbinghaus forgetting curve (the empirically measured pattern of how human memory decays exponentially with time unless reinfor...
HauhauCS publishes plagiarized abliteration package — AGPL-3.0 violation of Heretic confirmed by original author
NotableHauhauCS (a HuggingFace account with 5M+ combined monthly downloads across 22 uncensored models) published `reaper-abliteration` on HuggingFace and PyPI. Philipp Emanuel Weidmann, ...
Luce DFlash: 2× speculative decoding speedup for Qwen3.6-27B on a single RTX 3090
Notablesandropuppo released Luce DFlash (MIT license, C++/CUDA, standalone, no Python runtime): a speculative decoding implementation (speculative decoding — a small "draft" model propose...
Developing Trends
When multiple independent signals point at the same outcome, something is likely about to happen. These are the trends we're tracking.
Defense AI Becomes a Standalone Public Market Category by Mid-2027
At least one defense-focused AI company (Shield AI, Anduril, or similar) will IPO at a $15B+ valuation by mid-2027, establishing defense AI as a distinct investable category separate from general AI o...
Inference Cost Collapse Enables Mass Agent Deployment by Late 2026
The combined effect of memory compression, alternative CPU architectures, and post-training optimization will reduce inference costs by 5-10x within 6 months, making it economically viable to deploy p...
AI IPO Wave Creates Public Market Investment Window in H2 2026
At least 3 major AI companies (from: Anthropic, OpenAI, Shield AI, Anduril, Databricks, Scale AI) will file for IPO in H2 2026, creating a concentrated public market entry window for AI-focused invest...
AI Safety Evidence Cascade Triggers Regulatory Action Wave by Q4 2026
At least 2 major regulatory actions (new laws, binding enforcement actions, or significant new mandatory compliance frameworks) specifically targeting AI behavioral safety (sycophancy, manipulation, s...
Agent Orchestration Standardization
Multi-agent orchestration emerges as a distinct infrastructure category with de facto standards, as 5+ independent projects converge on solving the same coordination problem simultaneously.
Roadblocks Holding AI Back
Constraints slowing progress across multiple areas. When one clears, it unlocks a cascade.
AI agents aren't reliable enough yet
BlockingOnly 37% success rate in enterprise tasks. 5 major companies are working on this. 6 recent signs of progress.
The tool-connection protocol has design gaps
BlockingMCP (how AI agents connect to tools) can't represent all tool types correctly — agents silently fail on complex tasks. 3 teams working on fixes.
AI Cybersecurity Offense/Defense Asymmetry
Getting worseHow It's Blocked
Not enough datacenters are getting built
ImprovingHalf of planned US datacenters delayed due to equipment shortages from China. Only 4GW of 12GW is actually under construction.
Talent Concentration in Frontier Labs
ImprovingHow It's Blocked
How Fast Things Are Moving
Not just where things stand — whether they're speeding up or slowing down.
VRAM cost per GB (consumer, 32GB workstation)
The cost of GPU memory is dropping fast — running AI locally gets cheaper.
KV-cache compression ratio (zero accuracy loss)
AI models can compress their memory use dramatically — more users per GPU.
Frontier model FrontierMath score
How well frontier models solve hard math — a key capability benchmark.
Largest single AI investment
The largest single bet anyone has placed on AI keeps getting bigger.
OpenAI total private funding (single round)
+$122B in valuation in 6 days — Amazon $50B anchor
Anthropic ARR growth rate vs OpenAI
Anthropic is growing revenue 3x faster than OpenAI.
Defense AI company valuation (Shield AI)
+140% YoY
Legal AI company valuation (Harvey)
+$1B/month