Let's talk about a few directions that look promising and those that do not for the second half of the year～

1. Promising prediction markets, AI oracles + permissionless predictive products = a new form of AI-driven casino, matching the needs of gamblers and money launderers.

2. Promising the Moand ecosystem, where the top founders taking the wildest paths can actually break the deadlock.

3. Promising the new version of the economic model, which is more complex and random, e.g., more airdrop allocations paired with task unlocks.

4. Promising permissionless, stable DeFi/RWA yield layers.

1. Not optimistic about meme coins backed by market makers, they are all crashing one after another.

2. Not optimistic about reinventing the wheel in ways that do not align with the fundamental conditions of the chain, such as doing RWA on BSC or Dex on Sui.

3. Not optimistic about AI workflows and Multi AI in Crypto, the situation changes too quickly to keep up, taking one step ahead does not guarantee success, and it demands too much from team capabilities.

Finally, the quality requirements for projects have decreased; as long as it looks passable, the corresponding shell premium is also decreasing, and the demands on project creators' pockets are higher. With the elimination of market makers creating opening prices, retail investors flocking in, perhaps everyone is truly competing on whose product is used, establishing a reasonable connection between the product and the token～