Expert profile
Founder & AI Product Leader
Yash Thakker is a Generative AI expert with over 12 years of experience in product leadership and technical strategy. As the founder of explainx.ai, he has taught over 300,000 learners and built AI platforms serving millions of users globally. He specializes in Agentic AI, Multimodal RAG, and the intersection of LLMs with consumer hardware.
Microsoft is creating Xbox Labs to explore AI game decompiling and recompiling. What Asha Sharma said, how recompilation works, and the unanswered IP questions.
Shing-Tung Yau launched a Beijing AI-for-math alliance of eight institutions and 1,000+ researchers while Western mathematicians debate a boycott.
Harvard study of 700+ firms: AI coding agents add 30% more code but no gain in resolved issues, and PR merge time rises 49%. What teams should do.
Bloomberg says AI is supercharging a $442B scam economy bigger than cocaine. What the GASA, CRS and UN data support, and how to protect yourself.
AlphaProof Nexus is now in Science: 9 of 353 Erdős problems, 44 OEIS conjectures, Lean-verified proofs, and why peer review matters amid the OpenAI boycott.
Anthropic says small AI models may be months from self-replicating across cheap IoT devices. Here is what its tests show, what they skip, and how to defend.
Apple EMNLP paper: LLM vs human translation quality annotation across 70 language pairs; agreement mostly low for both.
Apple Minions pools idle Macs into a GPU cloud for inference: pull-based design, watchdog fault tolerance, and no benchmarks yet.
SelfCompact lets LLM agents choose when to compact context: up to +18.1 math points and 30-70% lower cost, no fine-tuning needed.
Army AI drone test: soldiers reported overheating, short range and drones not on the NGC2 network. What the reporting shows and what is unknown.
Viral claim: China spends $3.9T on its grid for AI. Verified: State Grid plans 4 trillion yuan, about $574B. What the numbers do and do not show.
LightOn released ColGrep-Agent-2B, an Apache-2.0 local code-search agent. Delegating search cut Claude Code API tokens 34% in its tests, at a time cost.
