GENIUSPreço de Genius

Preferred Shares Are Not the Catalyst for Corporate Bankruptcy; MicroStrategy's Dollar Reserves Can Cover Dividend and Interest Payments Until February 2027. This article analyzes the nature of preferred shares used by MicroStrategy (MSTR). Legally equity but economically similar to debt, these shares, including its Bitcoin-linked STR convertible preferred notes (STRC), offer fixed or floating dividends. Crucially, MicroStrategy's preferred shares lack rigid redemption clauses, meaning they are not classified as traditional debt. This eliminates principal repayment pressure and means missed dividends do not constitute default or trigger bankruptcy, creating a "self-contradictory virtuous cycle." The article clarifies that if funds are short, MicroStrategy can defer or suspend preferred share dividends (except for non-cumulative types like STRD) without immediate risk. The real potential crisis point lies with its convertible bonds. If a prolonged bear market prevents conversion, MicroStrategy might need to sell Bitcoin to repay these bonds starting from the earliest maturity in September 2027, potentially creating a downward spiral. Preferred dividend suspensions would only exacerbate market panic in such a scenario. Recent financial activity shows MicroStrategy strengthened its position through four weeks of common stock (MSTR) issuances, raising over $851 million without issuing new preferred shares. It increased its dollar reserves to approximately $1.4 billion, which is sufficient to cover all preferred share dividends and interest until around March 2027. While Bitcoin purchases slowed recently, this prioritization of cash reserves enhances the company's near-term financial safety. The analysis concludes that if the Bitcoin bear market ends by early 2025 as anticipated, MicroStrategy can resume issuing MSTR stock in a rising market to replenish reserves and manage future dividend obligations, thereby reducing the long-term pressure from its preferred share structure.

Andrej Karpathy has proposed a revolutionary concept for managing personal knowledge: treating notes as immutable "source code" and using LLMs as "compilers" to build a structured, interlinked wiki. This approach fundamentally shifts the cognitive workflow away from the limitations of RAG (Retrieval-Augmented Generation), which merely retrieves and pieces together fragments, leading to contradictions and "digital mummies"—unused, decaying notes. The LLM-Wiki framework introduces a three-layer architecture: the **Raw Layer** for original, immutable notes; the **Schema Layer** defining rules for structuring knowledge; and the **Wiki Layer**, where the LLM continuously compiles and maintains a coherent, cross-referenced knowledge base. Key operations are **Ingest** (adding new material, which triggers updates across related pages), **Query** (asking the compiled wiki, with answers that can become new pages), and **Lint** (periodic AI audits to find contradictions, outdated claims, or gaps). This system automates the tedious maintenance—updating links, resolving conflicts, keeping summaries fresh—that has historically made large-scale personal knowledge management unsustainable. It realizes Vannevar Bush's 1945 "Memex" vision by finally solving the maintenance problem. Karpathy's proposal represents a third piece in human-AI collaboration, following "Vibe Coding" and "Agentic Engineering." It liberates human attention from organizational drudgery, refocusing it on what matters: deciding what to read and deriving meaning.

Former Huawei "Genius Youth" Li Bojie recently drew public attention by criticizing his interview experience with DeepSeek. The controversy escalated when Du Jun, co-founder of Web3 investment firm ABCDE Capital, publicly accused Li of being "the founder with the least sense of contractual spirit" he had ever cooperated with, sparking a dispute over Li's startup project, Metagent. Li detailed a frustrating DeepSeek interview where he was accused of potential plagiarism, leading him to end the session. The spotlight then shifted to his venture, Metagent, a Web3+AI project aiming to tokenize AI agents. ABCDE invested $1.5 million, with an initial $500k disbursed. Du Jun claimed the project's progress was severely lacking, with a poor-quality demo and minimal social media activity. He alleged Li stopped communicating, deleted his Telegram, and failed to provide proper financial reporting. In response, Li argued the remaining $1 million was never received, crippling operations and forcing salary cuts. He stated he left Metagent in October 2024 due to family reasons and Web3 compliance concerns, with board approval. He claimed to have fulfilled disclosure duties and that his subsequent projects avoided conflicting fields. Other investors, including ArkStream Capital, shared negative due diligence experiences, citing unprofessional contracts and evasive answers on tokenomics. Metagent's social media went silent in June 2024, effectively stalling. Li has since moved to a new consumer AI agent platform, Pine AI (formerly Logenic AI), which has raised $25 million in Series A funding. He served as its Chief Scientist but recently left, clarifying he was not the founder and departed due to a shift in research interests.

Six U.S. federal agencies are progressing with rulemaking for the GENIUS Act payment stablecoin framework, aiming for a July 18 deadline. This development provides crucial regulatory clarity on stablecoin reserves, issuer capital requirements, and new licensing pathways, particularly for commercial banks. The primary impact is on market structure and long-term operational certainty, not immediate token prices. For traders, builders, and compliance teams, the update offers a concrete data point in a rapidly evolving sector. Stablecoins are a foundational component of crypto, linking exchanges, payments, and settlements, meaning changes to their regulation have broad implications. The story's significance lies in its contribution to understanding the regulatory trajectory for stablecoins in the current cycle. It underscores a move towards clearer rules rather than representing final legal certainty. The value is in a precise, source-based assessment of what has changed, who is affected, and what steps come next, avoiding overblown conclusions.

GPT-5.6 has reportedly achieved an IQ score of 136 on Tracking AI's proprietary offline test, surpassing the human "genius" threshold of 130 for the first time. This places it above an estimated 99% of humans in this specific metric. The test is designed to prevent memorization by using a private question bank. Multiple GPT-5.6 variants, including the vision model, consistently scored 136, leading competitors like Claude-5 Fable (130). User anecdotes suggest practical superiority over rivals in real-world coding and problem-solving tasks, such as building a physics simulation or a customer service app from a single prompt. While some speculate this approaches AGI for most users, the article notes IQ tests only measure a narrow slice of cognitive ability like pattern recognition. The significance lies in GPT-5.6's apparent ability to translate high test scores into effective task performance on novel, real-world problems.