The Macro Shift: From Model-Centric to Eval-Centric. The value is moving from the LLM itself to the proprietary evaluation loops that keep the LLM on the rails.
The Tactical Edge: Export production traces and build a "Golden Set" of 50 hard examples. Use these to run A/B tests on every prompt change before hitting production.
The Bottom Line: Reliability is the product. If you cannot measure how your agent fails, you haven't built a product; you've built a demo.
The transition from chatbots with tools to agents that build tools marks the end of the manual integration era.
Stop building custom model scaffolding and start building on top of opinionated agent layers like the Codex SDK.
In 12 months, the distinction between a coding agent and a general computer user will vanish as the terminal becomes the primary interface for all digital labor.
The Capability-Utility Gap is widening. We see a divergence where models get smarter but the friction of human-AI collaboration keeps productivity flat.
Deploy AI for mid-level engineers or low-context tasks. Avoid forcing AI workflows on your top seniors working in complex legacy systems.
The next year will focus on reliability over raw intelligence. The winners will have models that require the least amount of human babysitting.
The Macro Shift: Scaling laws are hitting a diminishing return on raw data but a massive acceleration in reasoning. The shift from statistical matching to reasoning agents happens when models can recursively check their own logic.
The Tactical Edge: Build for the agentic future by prioritizing high-context data pipelines. Models perform better when you provide massive context rather than relying on zero-shot inference.
The Bottom Line: We are 24 months away from AI that makes unassisted human thought look like navigating London without a map. Prepare for a world where the most valuable skill is directing machine agency rather than performing manual logic.
The transition from model-centric to loop-centric development. Performance is now a function of the feedback cycle rather than just the weights of the frontier model.
Implement an LLM-as-a-judge step that outputs a "Reason for Failure" field. Feed this string directly into a meta-prompt to update your agent's system instructions automatically.
Static prompts are technical debt. Teams that build automated systems to iterate on their agent's instructions will outpace those waiting for the next model training run.
The Macro Shift: The transition from writing to reviewing as the primary engineering activity. As agents generate more code, the human role moves from creator to editor.
The Tactical Edge: Build CLIs for every internal tool to give agents a native text interface. This increases accuracy and speed compared to visual automation.
The Bottom Line: Developer experience is the infrastructure for AI. Investing in clean code and fast feedback loops is the only way to ensure AI productivity gains do not decay over the next 12 months.
**Privacy Isn't a Feature; It's the Foundation.** For institutions, confidentiality is non-negotiable. Any network aiming to attract serious capital must offer privacy that allows for compliance without broadcasting every move to the world.
**Real Adoption Is a Long Game.** Chasing bull market hype is a losing strategy for enterprise adoption. Canton’s success with partners like Goldman Sachs, DTCC, and Citadel demonstrates the power of prioritizing utility and compliance over a premature token launch.
**The Next Wave Is Tokenizing Everything.** The goal is to move beyond crypto-native assets. The real prize is upgrading the rails for the world's existing financial system—equities, bonds, and treasuries—by making them digitally native, 24/7, and instantly settleable.
Focus or Fade. As the industry matures, companies must shed non-core business units to become world-class at one thing. For Blockworks, that's data, not news.
Buy the Theme. Public market investors will pay a massive premium for the only stock representing a major crypto trend (e.g., Securitize for tokenization), often making it a better trade than trying to pick winners among underlying assets.
Growth is Subsidized. Major L1/L2 foundations are actively paying for enterprise adoption (e.g., Solana and Western Union). This is a standard business practice to kickstart network effects, but the long-term ROI remains unproven.
Social Proof is the New Alpha. FOMO’s core bet is that transparently tracking successful wallets is a more powerful discovery mechanism than traditional research. By making on-chain activity legible and social, it unlocks a new paradigm for retail investing.
User Experience Wins the Next Cycle. The next 100 million crypto users will not be onboarded with seed phrases and gas fees. By abstracting away all technical friction and mirroring the seamlessness of Web2 apps, FOMO provides a blueprint for mass adoption.
Trading is Becoming a Spectator Sport. By turning trading performance into a form of content, FOMO is building a new financial creator economy. The best traders are the new influencers, and their alpha is the content that drives the entire ecosystem.
The Internet Gets Its Native Wallet. x402 uses crypto to finally fulfill the internet's original vision of direct, peer-to-server payments, unlocking an economy of micropayments for everything from accessing an article to running an AI model.
The Ad-Supported Web Is Obsolete. AI agents that retrieve information without viewing ads are killing the web's 20-year-old business model. x402 provides the new economic rails for a pay-per-use internet where value is exchanged for resources, not attention.
Build Composable Money Legos. The biggest opportunity lies in creating simple, single-purpose APIs that agents can easily discover and compose. Think of it as building for the App Store of the emerging AI agent economy.
Despite the brutal sentiment, both speakers remain bullish, predicting a sharp reversal and a new all-time high for Bitcoin by the end of the year once macro clarity emerges.
Macro is King. Bitcoin's fate is now tied to the broader economy. Forget four-year cycles; the key catalyst is a resolution to the government shutdown, which could unlock pent-up energy in the market.
The Treasury Trade is Toast. The era of companies boosting their stock by simply buying Bitcoin is over. Expect a painful shakeout as the market demands real utility and revenue, leading to more forced selling.
Distribution is the prize, not payments. The real money in gaming isn't in payment rails, which are being competed down to zero. The scarcest and most valuable asset is user attention, and Open Game Protocol is designed to monetize distribution.
Decouple to survive. Tying a single game to a single token creates a fatal dependency. A flexible, many-to-many protocol allows ecosystems to thrive by spreading risk and aligning incentives across multiple games and communities.
Content abundance requires new discovery. As AI makes game creation cheap and fast, the bottleneck shifts from production to discovery. Integrating games directly into high-intent surfaces like crypto wallets creates a powerful, non-advertising-based discovery channel.
