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.
Musk reposted Wait But Why''s 2015 AI intelligence curve. What the chart means, why he says we''re on the steep slope, and what builders should do.
Y Combinator''s Garry Tan says US open-weight labs should distill frontier models just like Chinese labs do — and that access to AI trained on public data should be a public good, not locked behind ToS. Here''s his argument and the pushback.
Musk posted Grok 4.8 at 2.5T parameters on a new C++ training stack; pretrain ends this week, then RL. What builders should watch on cadence vs eval time.
A skeptical read on AI's safety-pacing wave: data limits, serving costs, and IPO timing may explain the slowdown as much as caution does.
A new paper argues real recursive self-improvement means AI choosing its own improvement strategy, not just executing human-designed ones. Here''s the four-stage roadmap and why the framing matters more than the hype.
Meta wants Muse to run a small business end to end — and take a cut of every sale it processes. Here''s how the "business in a box" model works and what it means for sellers who don''t want to hand Meta the storefront.
Mustafa Suleyman's MAI Code of Conduct: 10 rules built on "AI must be subordinate to people," published Sept 14, 2026 for public comment.
Minecraft creator Notch says coding AI has basically arrived — his real regret is who owns it. His post reignited the local/open-weight AI debate. Here''s what he actually said and why replies pushed back.
NVIDIA Inception gives AI startups free cloud credits, SDKs, and investor exposure — no funding or equity required. Here''s who actually qualifies and what founders on X say the fine print leaves out.
OpenAI ran 10,000 parallel AI agents against a math problem open since the 1930s and produced a proposed solution. Here''s how the swarm approach worked, what''s been verified, and why Fields Medalists are skeptical of claims like this.
Paul Graham''s September 2026 essay reframes startup advice around one question: what makes this company more powerful, not just more profitable. Here''s every strategy in the essay, and where AI agents change the math.
Perplexity's Portable Computer shipped on Windows Sep 14, 2026 — but local inference still needs an NVIDIA RTX GPU with 24GB+ VRAM. Who qualifies.
