The bottleneck in most AI-heavy businesses is approving much more so than generating. Just delivered an assessment to a founder who already runs Claude Code, a second brain on a Mac Mini, and a one dedicated wiki per client. His constraint is that every call summary and framework has to be manually reviewed before it enters the knowledge base. Generation is instant but ratification piles up. The fix, in two sizes: Quick win (this week) -Add his assistant to TailScale so she can reach the Mac Mini -She runs the existing intake skill on each Fathom transcript -Output goes to him for a yes or no, then into the second brain Major project (later) -Fathom fires a webhook when a transcript is done -An agent listens and runs the intake skill automatically -A routing skill sends the result to the right wiki (main brain first, then the client's) -Assistant approves or rejects in one shared channel Net savings we estimated: 1.5 to 2 hours a week after 30 to 60 minutes of review time.

The SEC just drew a clearer legal line through the tokenized-stock market. On one side are tokens that represent the stock itself, or preserve the same rights as the underlying security. On the other are products that simply track the economics of a stock through notes, certificates, derivatives or synthetic exposure. That distinction now matters much more because @SECGov has opened a five-year pathway for qualifying tokenized NMS stocks to trade through permissioned AMMs on public blockchains. The important detail is that a large portion of today’s tokenized-equity market does not automatically qualify. — ● The SEC did not broadly approve tokenized stocks On September 17, 2026, the SEC introduced a temporary Innovation Exemption for qualifying tokenized NMS securities. The framework allows Tokenized Securities Venues to operate permissioned AMMs without registering under the same structure as traditional exchanges, provided they satisfy the exemption’s conditions. Some liquidity providers can also receive conditional dealer relief. The underlying smart contracts can still operate on public blockchains, but participation in the regulated market remains permissioned. The structure therefore looks more like: Public blockchain -> permissioned participants -> AMM liquidity -> regulated tokenized equities This is a controlled market-structure experiment rather than a blanket approval of every tokenized stock already trading onchain. — ● The biggest distinction is what the token legally represents The SEC framework recognizes two broad models. • The first is issuer-native tokenization. A company, together with its transfer agent or another authorized party, can issue the actual share directly onchain. The token represents the security itself. • The second is a third-party custodial model. In this case, the underlying stock remains in traditional custody while an onchain token represents an ownership interest in that security. For the token to fit the framework, it has to preserve the same rights and privileges as the traditional share, including things such as dividends and voting rights. That is where the market begins to split. A token that represents actual equity ownership is fundamentally different from a token that only follows the price of the equity. — ● Two tokenized-equity markets are now emerging • The first market consists of rights-preserving tokenized securities. These products try to bring actual equity ownership, or a legally equivalent ownership interest, onto blockchain rails. Platforms such as @DinariGlobal and newer U.S. structures from @Ondo are moving closer to this model, while firms such as @Securitize are already building infrastructure around regulated tokenized securities. Dinari’s dShares, for example, are backed by securities held through regulated custody and are designed to preserve dividends, voting rights, corporate actions and redemption rights. • The second market is built around economic exposure. This includes structures where the token tracks the performance of a stock without giving the holder direct shareholder rights. @xStocksFi uses tracker certificates that provide exposure to underlying equities but do not confer voting rights. @RobinhoodApp’s existing international Stock Tokens are structured as tokenized debt securities that provide economic exposure without giving holders legal or beneficial ownership of the underlying company. Stock perpetuals on venues such as @HyperliquidX sit even further from direct equity ownership because they are derivatives. The price exposure can look similar across all of these products, but the legal claim behind them is very different. — ● Ondo shows why the distinction can get complicated @Ondo now operates across different tokenization structures. Its older global model uses products issued through Ondo Global Markets in the BVI. Those instruments provide economic exposure to the performance of underlying stocks through note-like structures rather than direct equity ownership. Its newer U.S. model is different. The U.S. structure uses traditional custody and market infrastructure, with Broadridge supporting proxy voting and shareholder communications. That moves the model much closer to the rights-preserving structure the SEC is now accommodating. The important takeaway is that saying “Ondo tokenized stocks” is no longer specific enough. The legal wrapper matters just as much as the underlying ticker. — ● The SEC is also keeping the experiment deliberately small The exemption does not suddenly move U.S. equity liquidity onto AMMs at scale. The SEC imposed strict symbol and volume limits. • Tier 1 covers S&P 500 stocks, Russell 1000 names and qualifying ETPs. Each Tokenized Securities Venue can support up to 75 symbols, with trading capped at 0.25% of the underlying stock’s prior-month average daily volume. • Tier 2 covers other qualifying NMS stocks. Each venue can support up to 250 symbols, with trading capped at 2.5% of prior-month average daily volume. These limits allow the SEC to test permissioned AMM market structure without immediately shifting meaningful portions of traditional equity liquidity onchain. The current framework is therefore better understood as a sandbox for regulated secondary-market infrastructure. — ● The interesting mismatch is where liquidity already sits The tokenized-equity market is already meaningful in size. @RWA_xyz tracks roughly $3.01B in distributed value and more than $12B in monthly transfer volume. Major platforms include: • @Ondo at roughly $854M • @bstocksfinance at roughly $754M • @xStocksFi at roughly $557M • @Securitize at roughly $358M But these numbers combine products with very different legal structures. Some represent regulated securities or ownership interests. Others are debt instruments, tracker certificates, derivatives or synthetic exposure. That creates an important mismatch. A large share of current liquidity already sits in products that are not automatically aligned with the SEC’s new framework, while the more legally aligned structures remain comparatively smaller. — The first phase of tokenized equities was mainly about bringing stock exposure onchain. The next phase is becoming much more focused on the legal quality of that exposure. The market is starting to separate simple price exposure from actual shareholder rights. The SEC’s Innovation Exemption begins solving the secondary-market problem for a narrow class of tokenized equities that preserve those rights. It does not turn every existing stock token into an onchain share. That means the most important signal from here is where liquidity begins to migrate. If capital starts moving toward structures that combine real shareholder rights, regulated custody and onchain AMMs, the tokenized-equity market will be entering a very different phase. It will be moving from simply tokenizing stock prices toward actually bringing the stock market onchain.

The US has reportedly proposed extending the US–China trade truce by six months. Beijing wants a longer extension, reportedly covering the remainder of Trump’s presidency, according to the Financial Times. No extension has been agreed. My reading: the duration itself is a strategic bargaining instrument for the second summit as a continuation from the first one. Washington wants regular opportunities to assess compliance and retain leverage. Beijing seeks greater predictability and fewer recurring deadlines for renewed pressure. A longer truce could reduce immediate trade uncertainty while leaving the underlying systemic competition intact. Both powers have reasons to buy time and play with tradeoffs, and apply their own plans for how to use it. But the shared interest remains the same from May: a managed confrontation and avoiding global system collapse.

The most expensive thing in systematic trading is often not a losing strategy. It's the two years you spent not being in the market. Everyone underestimates this because time doesn't show up anywhere in your P&L. A blown account is visible and painful and you learn from it fast. Three years spent building a custom machine learning backtest engine, a feature pipeline and an ML framework before you've placed a single live trade costs you *far* more, and the damage is completely invisible. And the cruel part is that the building genuinely feels like progress. You're learning, you're solving real problems, you're getting better at something. But you're getting feedback about your code, not about the market. You only get it by being on the field. Meanwhile the thing you were going to build usually already exists in a simpler form for almost nothing. If you want trend, start with a basic 12 month lookback and a monthly rebalance. That's your first sleeve done in an afternoon, live, compounding, teaching you things. Then every afternoon after that goes into the second sleeve, and the third. Five boring strategies stacked together will beat the one brilliant one you're still building, because their drawdowns land in different places. So before you build anything, ask what you can do *right now*, with the tools you already have, that is likely to make money. Do that this week. Complexity comes later, if the simple version genuinely can't harness the edge.

The first scaling axis was data & model size. The second one was RL environments & compute. I think model companies are now scaling on third axis: agent swarms. If you notice all the recent eye-popping behaviours are an outcome of 10,000+ agents working together intelligently.