{
      "@type": "WebPage",
      "@id": "https://afcp.ai/our-thesis/#webpage",
      "url": "https://afcp.ai/our-thesis/",
      "name": "AFCP AI | Our Thesis — Efficient AI, Experiential AI & the Agentic Internet",
      "description": "AFCP AI's founding thesis on building efficient, experiential AI for agentic commerce. Covers the Agentic Internet, Efficient AI architecture, and Experiential AI as the successor to Generative AI.",
      "isPartOf": { "@id": "https://afcp.ai/#organization" },
      "breadcrumb": {
        "@type": "BreadcrumbList",
        "itemListElement": [
          { "@type": "ListItem", "position": 1, "name": "Home", "item": "https://afcp.ai" },
          { "@type": "ListItem", "position": 2, "name": "Our Thesis", "item": "https://afcp.ai/our-thesis/" }
        ]
      }
    },

    {
      "@type": "FAQPage",
      "@id": "https://afcp.ai/our-thesis/#faq",
      "mainEntity": [

        {
          "@type": "Question",
          "name": "What is Efficient AI?",
          "acceptedAnswer": {
            "@type": "Answer",
            "text": "Efficient AI is AI that is like electricity — scalable, affordable and reliable. It must consume fewer chips, compute and energy than today's LLM-first approaches. AFCP AI believes a new hybrid AI architecture is needed for efficiently carrying out thinking, planning, reasoning and acting, rather than relying solely on large language models whose compute and energy costs make them unsuitable for efficient scaling in the long run."
          }
        },

        {
          "@type": "Question",
          "name": "What is Experiential AI?",
          "acceptedAnswer": {
            "@type": "Answer",
            "text": "Experiential AI is a new approach to building AI that is human-first, collaborative and self-learning. Unlike Generative AI, which relies on large pre-trained models, Experiential AI builds agents that learn continuously from real-world experiences and human intervention, making them self-adaptive over time. AFCP AI uses Experiential AI to build shopper agents that are not only active helpers for consumers but also interactive and fun to be with."
          }
        },

        {
          "@type": "Question",
          "name": "Why is an LLM-first approach unsuitable for agentic commerce?",
          "acceptedAnswer": {
            "@type": "Answer",
            "text": "Compute and energy costs are rising, making LLMs unsuitable for efficient scaling in the long run. Models cannot scale with the real world — they lack continual learning and memory. Generative AI uses a lot of training and inference compute that is not necessary. AFCP AI aims to consume 10x less infrastructure, compute and energy compared with model-first approaches, by implementing better strategies to build AI agents that can scale efficiently."
          }
        },

        {
          "@type": "Question",
          "name": "What is the DC vs AC analogy for AI?",
          "acceptedAnswer": {
            "@type": "Answer",
            "text": "AFCP AI draws an analogy between the DC distribution crisis of electricity and today's LLM-first AI paradigm. AI built on an LLM-first approach is analogous to DC electricity — it works but cannot scale to meet the growing needs of businesses, the environment and consumers. The solution is AC: an efficient agent-first approach, just as AC electricity replaced DC for large-scale distribution. Efficient AI can help deploy AC-like agents for the world."
          }
        },

        {
          "@type": "Question",
          "name": "What is the Agentic Internet?",
          "acceptedAnswer": {
            "@type": "Answer",
            "text": "The Agentic Internet is a reimagined version of the current internet where consumers and businesses interact directly with agents through voice, text or video. It uses the existing Web1, Web2 and Web3 infrastructure as the backbone but adds an agentic layer for the experiencing part. Building the Agentic Internet requires solving five foundational challenges: accessibility of agents, identity of agents, security and privacy of agents, control back to consumers, and trustworthiness of agents."
          }
        },

        {
          "@type": "Question",
          "name": "What is an artificial shopper's mind?",
          "acceptedAnswer": {
            "@type": "Answer",
            "text": "An artificial shopper's mind is AFCP AI's approach to building AI that understands and anticipates shopper behaviour with precision. Rather than relying on large models alone, it is an artificial mind specific to cracking shopping intelligence — one that plans, remembers, and pursues goals for a shopper based on real-world experiences. It enables instant hyper-personalised searches and suggestions, agentic buying and guiding for instant or repeat purchases, and automation of shopping tasks like research."
          }
        },

        {
          "@type": "Question",
          "name": "How does Experiential AI differ from Generative AI?",
          "acceptedAnswer": {
            "@type": "Answer",
            "text": "Generative AI was the last revolution. Experiential AI is the new approach needed for the Agentic future. While Generative AI relies on large pre-trained models and high inference compute, Experiential AI builds agents that are human-first, collaborative and self-learning. The architecture can be hybrid and use components of Generative AI, but without being restricted to it. Experiential AI agents learn by experience continually, consume less energy and compute, can be governed, and are safe, reliable, and self-adaptive."
          }
        }

      ]
    }

  ]
}