Meet the Majorana 2, a next-generation topological quantum chip developed with the help of Microsoft Discovery’s agentic AI.

☢️You remember? Google optimized Shor's algorithm. The algorithm that breaks asymmetric cryptography (RSA, elliptic curves) once you have a quantum computer with enough Qubits. The US government blocked the paper. So Google published a Zero Knowledge proof instead: a mathematical proof that they have the result, without revealing how. Cryptographic sorcery 🧙 But the Internet is sneaky. Someone launched a contest to re-discover the result with AI. The LLM searches a huge space of circuits (each one a candidate optimization of Shor's), and tests whether it beats the previous best. The clever part: they use the ZKP verifier as the reward function. No false positives, and it turns out to be a very efficient signal. In less than 2 days, the community re-discovered Google's result !!! 🔔15 days later, the LLMs are still improving it. They're already 44% ahead of Google. Hard to say where this stops, ie. what the true minimum quantum complexity for Shor's is. But we will not close the full gap. You still need a Quantum Computer with a relatively large number of qubits. The only thing that changed is that this number drops a little every day, and it has been dropping for 15 days straight.

Claude Code fully dissected! Researchers from UCL reverse-engineered the leaked Claude source. What they found changes how you should think about agent design. Only 1.6% of the codebase is AI decision logic. The other 98.4% is operational infrastructure. Permission gates, tool routing, context compaction, recovery logic, session persistence. The model reasons. The harness does everything else. This is the opposite of what most agent frameworks do today. LangGraph routes model outputs through explicit state machines. Devin bolts heavy planners onto operational scaffolding. Claude Code gives the model maximum decision latitude inside a rich deterministic harness, and invests all its engineering effort in that harness. The core loop is a simple while-true. Call model, run tools, repeat. But the systems around that loop are where the real design lives: A permission system with 7 modes and an ML classifier. Users approve 93% of prompts anyway, so the architecture compensates with automated layers instead of adding more warnings. A 5-layer context compaction pipeline. Each layer runs only when cheaper ones fail. Budget reduction, snip, microcompact, context collapse, auto-compact. Four extension mechanisms ordered by context cost. Hooks (zero), skills (low), plugins (medium), MCP (high). Each answers a different integration problem. Subagents return only summary text to the parent. Their full transcripts live in sidechain files. Agent teams still cost roughly 7x the tokens of a standard session. Resume does not restore session-scoped permissions. Trust is re-established every session. That friction is the point. The bet behind all of this is simple. As frontier models converge on raw coding ability, the quality of the harness becomes the differentiator, not the model. Paper: Dive into Claude Code (arXiv:2604.14228) We've shared an article on Agent Harness and what every big company is building. Read it below.