<!DOCTYPE html>

<head>

<title>Let's Learn Super Intelligence | letslearnsi.com</title>

<style>

body {

font-family: 'Inter', -apple-system, BlinkMacSystemFont, sans-serif;

background-color: #000000;

color: #F5F5F7;

overflow-x: hidden;

}

.glass {

background: rgba(255, 255, 255, 0.04);

backdrop-filter: blur(20px);

-webkit-backdrop-filter: blur(20px);

border: 1px solid rgba(255, 255, 255, 0.08);

}

.glass-hover:hover {

background: rgba(255, 255, 255, 0.07);

border-color: rgba(255, 255, 255, 0.16);

}

.text-gradient {

background: linear-gradient(135deg, #FFFFFF 0%, #A1A1AA 50%, #6366F1 100%);

-webkit-background-clip: text;

-webkit-text-fill-color: transparent;

}

</style>

</head>

const { useState, useMemo } = React;

const LABS = ["All Labs", "OpenAI", "Google", "Anthropic", "xAI", "Open Weights"];

const BREAKTHROUGHS = [

{

id: 1,

lab: "OpenAI",

tagColor: "bg-emerald-500/20 text-emerald-300 border-emerald-500/30",

title: "Autonomous Multi-Horizon Reasoning Agents",

date: "Live Milestone",

summary: "Test-time compute scaling enables deep research and multi-step verification across complex engineering, math, and financial workflows.",

technical: "Utilizes reinforcement learning on chain-of-thought token streams with dynamic verifier trees, trading inference latency for exponential accuracy gains on SWE-bench and FrontierMath."

{

id: 2,

lab: "Google",

tagColor: "bg-blue-500/20 text-blue-300 border-blue-500/30",

title: "Gemini Multimodality & Infinite Context Memory",

date: "Core Architecture",

summary: "Native audio, video, and code reasoning over million-token windows with near-perfect needle-in-a-haystack recall.",

technical: "Ring Attention combined with mixture-of-experts (MoE) routing allows sub-linear compute scaling across massive repository and video stream inputs."

{

id: 3,

lab: "Anthropic",

tagColor: "bg-amber-500/20 text-amber-300 border-amber-500/30",

title: "Constitutional AI & Computer-Use Orchestration",

date: "Agentic Frontier",

summary: "Direct OS-level GUI interaction paired with mechanistic interpretability mapping internal feature circuits.",

technical: "Sparse autoencoders (SAEs) isolate monosemantic concepts inside transformer residual streams to audit autonomous tool execution in real time."

{

id: 4,

lab: "xAI",

tagColor: "bg-purple-500/20 text-purple-300 border-purple-500/30",

title: "Colossus Coherent Supercluster Scaling",

date: "Infrastructure",

summary: "200,000+ GPU synchronous training fabric pushing real-time world knowledge synthesis and unthrottled reasoning.",

technical: "RDMA over converged Ethernet (RoCEv2) fabric eliminates all-reduce bottlenecks during trillion-parameter dense and sparse pre-training phases."

{

id: 5,

lab: "Open Weights",

tagColor: "bg-cyan-500/20 text-cyan-300 border-cyan-500/30",

title: "Distilled Frontier Reasoning at the Edge",

date: "Open Ecosystem",

summary: "Open-weight models rivaling proprietary flagships via RL distillation and quantized local execution.",

technical: "Group Relative Policy Optimization (GRPO) transfers long-chain reasoning behaviors into 32B–70B parameter footprints runnable on consumer workstations."

}

];

const MODELS = [

{ name: "Frontier Reasoning Class (o-Series / R-Class)", lab: "OpenAI / Open", reasoning: "98.4%", coding: "94.2%", agentic: "High Autonomy" },

{ name: "Gemini Ultra-Scale Multimodal", lab: "Google", reasoning: "97.8%", coding: "93.5%", agentic: "Native Tool + Video" },

{ name: "Claude Opus / Sonnet Agentic", lab: "Anthropic", reasoning: "97.5%", coding: "95.8%", agentic: "OS Computer Use" },

{ name: "Grok Frontier Reasoner", lab: "xAI", reasoning: "96.9%", coding: "92.9%", agentic: "Real-Time Search" }

];

const CURRICULUM = [

{

track: "01. Foundations",

title: "Transformers, Tokenization & Scaling Laws",

time: "8 min read",

desc: "How self-attention, Chinchilla scaling laws, and next-token prediction give rise to emergent world models."

{

track: "02. Reasoning & RL",

title: "Test-Time Compute & Inference Scaling",

time: "12 min read",

desc: "Why allowing models to 'think before speaking' via Reinforcement Learning unlocks PhD-level science and coding."

{

track: "03. Agentic Stack",

title: "Autonomous Agents, Tool Use & Agentic Payments",

time: "10 min read",

desc: "How AI systems plan, call APIs, write code, and execute cryptographic programmable settlements autonomously."

{

track: "04. Superintelligence",

title: "Recursive Self-Improvement & Alignment",

time: "15 min read",

desc: "Mechanistic interpretability, formal verification, and governing systems that surpass human cognitive ceilings."

}

];

function App() {

const [selectedLab, setSelectedLab] = useState("All Labs");

const [searchQuery, setSearchQuery] = useState("");

const [deepDiveMode, setDeepDiveMode] = useState(false);

const [activeModal, setActiveModal] = useState(null);

const filteredBreakthroughs = useMemo(() => {

return BREAKTHROUGHS.filter(item => {

const matchesLab = selectedLab === "All Labs" || item.lab === selectedLab;

const matchesSearch =

item.title.toLowerCase().includes(searchQuery.toLowerCase()) ||

item.summary.toLowerCase().includes(searchQuery.toLowerCase()) ||

item.lab.toLowerCase().includes(searchQuery.toLowerCase());

return matchesLab && matchesSearch;

});

}, [selectedLab, searchQuery]);

return (

{/* Sticky Cupertino Header */}

</div>

<div>

Let's Learn

</span>

Super Intelligence

</span>

</div>

<a href="#radar" className="px-3.5 py-1.5 rounded-full bg-white/10 text-white transition">Bento Radar</a>

<a href="#feed" className="px-3.5 py-1.5 rounded-full text-[#86868B] hover:text-white transition">Live Breakthroughs</a>

<a href="#models" className="px-3.5 py-1.5 rounded-full text-[#86868B] hover:text-white transition">Frontier Index</a>

<a href="#learn" className="px-3.5 py-1.5 rounded-full text-[#86868B] hover:text-white transition">Learn SI</a>

</nav>

Frontier Status: Live

</div>

</header>

{/* Hero Section */}

<span>✦</span> Tracking the Path to Superintelligence

</div>

The Frontier, Decoded.

</h1>

Signal over noise across OpenAI, Google DeepMind, Anthropic, xAI, and the open-weights ecosystem—engineered for builders, researchers, and curious minds.

</p>

{/* Search & Mode Toggle Bar */}

<input

type="text"

value={searchQuery}

onChange={(e) => setSearchQuery(e.target.value)}

placeholder="Search models, labs, or architectures (e.g., Anthropic, Reasoning, MoE)..."

className="flex-1 px-4 py-3 rounded-2xl glass text-sm text-white placeholder-[#86868B] focus:outline-none focus:border-indigo-500 transition"

<button

onClick={() => setDeepDiveMode(!deepDiveMode)}

className={`px-5 py-3 rounded-2xl text-xs font-semibold border transition flex items-center justify-center gap-2 ${

deepDiveMode

? "bg-indigo-600/30 border-indigo-400 text-indigo-200"

: "glass text-[#86868B] hover:text-white"

}`}

<span>{deepDiveMode ? "⚡ Technical Mode: ON" : "🔎 Explain Simply"}</span>

</button>

</div>

{/* Lab Filter Pills */}

{LABS.map(lab => (

<button

key={lab}

onClick={() => setSelectedLab(lab)}

className={`px-4 py-1.5 rounded-full text-xs font-medium transition ${

selectedLab === lab

? "bg-white text-black font-semibold shadow-sm"

: "glass text-[#86868B] hover:text-white"

}`}

{lab}

</button>

))}

</div>

</section>

{/* Apple-Style Bento Grid */}

{/* Large 2-Column Tile: Frontier Model Matrix */}

<div>

Intelligence Radar

</span>

<span className="text-xs text-[#86868B]">Updated Live</span>

</div>

Frontier Capability Matrix

</h2>

Benchmark synthesis across reasoning depth, autonomous software engineering, and multimodal tool execution.

</p>

<thead>

<th className="pb-3 font-medium">Architecture Class</th>

<th className="pb-3 font-medium">Math / Reasoning</th>

<th className="pb-3 font-medium">SWE Coding</th>

<th className="pb-3 font-medium">Primary Edge</th>

</tr>

</thead>

{MODELS.map((m, idx) => (

<td className="py-3.5 text-emerald-400 font-semibold pr-3">{m.reasoning}</td>

<td className="py-3.5 text-indigo-300 font-semibold pr-3">{m.coding}</td>

{m.agentic}

</span>

</td>

</tr>

))}

</tbody>

</table>

</div>

{/* Concept of the Day Tile */}

<div>

Concept of the Day

</span>

Test-Time Compute Scaling

</h3>

{deepDiveMode

? "Instead of scaling only pre-training FLOPs, models allocate dynamic inference compute via Monte Carlo tree search and RL-trained verification chains before emitting final tokens."

: "Earlier AI answered immediately on instinct. Modern reasoning models pause to break problems into steps, double-check their work, and backtrack when stuck—just like a human engineer."}

</p>

</div>

<span className="text-[#86868B]">Mode: {deepDiveMode ? "Technical" : "Intuitive"}</span>

<button

onClick={() => setDeepDiveMode(!deepDiveMode)}

className="text-indigo-400 hover:text-indigo-300 font-medium"

Switch View →

</button>

</div>

</section>

{/* Filtered Breakthroughs Feed */}

<div>

<h2 className="text-2xl font-bold tracking-tight">Live Breakthroughs & Milestones</h2>

Showing {filteredBreakthroughs.length} milestone{filteredBreakthroughs.length === 1 ? "" : "s"} ({selectedLab})

</p>

</div>

{filteredBreakthroughs.map(item => (

<div

key={item.id}

onClick={() => setActiveModal(item)}

className="glass glass-hover rounded-3xl p-6 cursor-pointer transition flex flex-col justify-between"

<div>

{item.lab}

</span>

</div>

{item.title}

</h3>

{deepDiveMode ? item.technical : item.summary}

</p>

</div>

<span>Inspect Architecture</span>

</div>

))}

</div>

</section>

{/* Learn SI Curriculum */}

Structured Curriculum

</span>

Master the Superintelligence Stack

</h2>

</div>

{CURRICULUM.map((c, i) => (

<div>

<span className="text-xs font-mono text-indigo-400 block mb-2">{c.track}</span>

<h4 className="text-base font-bold text-white mb-2">{c.title}</h4>

</div>

⏱ {c.time}

</div>

))}

</div>

</section>

{/* Detail Modal */}

{activeModal && (

<div

className="fixed inset-0 z-50 bg-black/80 backdrop-blur-md flex items-center justify-center p-6"

onClick={() => setActiveModal(null)}

<div

className="glass max-w-lg w-full rounded-3xl p-8 border border-white/20"

onClick={(e) => e.stopPropagation()}

{activeModal.lab}

</span>

<button

onClick={() => setActiveModal(null)}

className="text-sm text-[#86868B] hover:text-white"

✕ Close

</button>

</div>

<h3 className="text-2xl font-bold mb-3">{activeModal.title}</h3>

<strong>Executive Summary:</strong> {activeModal.summary}

</p>

<strong className="block text-white mb-1">Under the Hood (Architecture Note):</strong>

{activeModal.technical}

</div>

)}

{/* Footer */}

<p className="text-white/60">Built for the Frontier • Hosted on Cloudflare Edge</p>

</div>

</footer>

</div>

);

}

const root = ReactDOM.createRoot(document.getElementById("root"));

root.render(<App />);

</script>

</body>

</html>