传统的app设计是针对人的。会不会有新的针对ai agent。提供更底层的流量计费模式。比如http402协议。

  • 当前的矛盾是新的注意力范式转移和旧有的注意力攫取机制之间的矛盾。

传统上，对于用户侧是免费的，同时收割注意力税。对于商家又要抽佣。开发app的新范式可不可以是面向ai agent的。

新范式：

用户 -> 本地/云端agent <–> 服务 <–> 商家agent -> 商品、服务提供商

做一个自动撮合交易系统。自动做市商？挣的是流量的过路费。将渠道费成本降至最低，利润留给产品制造者、品牌方，达成一个正向的品牌效应和升级。

  • 如何保证中小微的提供商不被淘汰（规模成本压不下来，打差异化、个性化）
  • 这样的机制是否会造成劣币驱逐良币
  • 这样的机制是否会会造成一家独大
  • 这样的范式成功的前提的基础设施建设是否完善
  • 链条的参与方各自的忧虑会是什么

Translated from the Chinese original.

Traditional app design is aimed at people. Will there be a new one aimed at AI agents? Providing a more low-level traffic billing model. For example, the HTTP 402 protocol.

  • The current contradiction is the contradiction between the new attention paradigm shift and the old attention-harvesting mechanism.

Traditionally, it is free on the user side, while harvesting an attention tax. Merchants are also charged a commission. Can the new paradigm for developing apps be oriented toward AI agents?

New paradigm:

user -> local/cloud agent <–> service <–> merchant agent -> goods, service providers

Build an automated matching trading system. Automated market maker? What it earns is a toll on traffic. Bring channel-fee costs down to the minimum, leave the profit with product manufacturers and brands, and achieve a positive brand effect and upgrade.

  • How to ensure that small and micro providers are not eliminated (scale costs cannot be compressed; compete on differentiation and personalization)
  • Will such a mechanism cause bad money to drive out good money
  • Will such a mechanism cause one player to dominate
  • Whether the infrastructure, as a prerequisite for this paradigm to succeed, is in place
  • What would be the concerns of each participant in the chain