Building an AI-Centric Business

In this episode, host Mark talks with James Thornton, CEO of Daz 3D / Tafi, about what it really means to build an AI-centric business. James shares how his company evolved from a 3D content and avatar business into a key AI data provider for some of the world's largest tech and gaming brands. They discuss: The shift from scraping web data to bespoke, rights-clean AI training data Using AI across product, marketing, and customer service workflows Real-world applications in VR, robotics, gaming, and product visualization How AI is transforming email marketing and localization What the next 3 years of AI adoption will look like inside typical organizations Key Topics & Timestamps [0:00:00] – Introduction & Guest Background Mark opens the show and frames the topic: AI-centric businesses with a marketing slant. James introduces himself and his company Daz 3D / Tafi. His background: CEO at Daz for over a decade Previous role growing a creative product brand from near zero to a $1B retail brand in 3.5–4 years How Daz has continually pivoted into AI trends [0:01:18] – What It Means to Be an AI-Centric Business James defines AI-centric as "AI-first" thinking across the business, especially in marketing. Emphasis on balancing AI with people, expertise, and intuition, rather than AI replacing human judgment. [0:01:37] – Daz / Tafi's Evolution into AI Company roots: 25-year-old business with a strong direct-to-consumer freemium model Deep community of 3D digital content creators (avatars, products for gaming, etc.) Past pivots: work in Web3 and NFTs, then parlaying that into today's AI data and services. They now both: Support other companies' AI initiatives Use AI internally to streamline content creation and business workflows [0:03:37] – How They Help Other Companies with AI Core specialty: 3D digital content, avatar systems, and digital assets. Use AI and procedural tools to: Scale their content library Produce rights-clean, structured, annotated AI training data Customers include top tech, gaming, and enterprise brands training foundation models for: Text-to-character Product creation & visualization (including physical product development) Robotics and other AI applications [0:04:51] – Concrete Benefits of AI for Clients and Daz Central thesis: Good AI requires high-quality training data. For clients: Bespoke, properly licensed data for training a wide range of models For Daz's own business: AI embedded in product workflows and content production AI in product search, customer service, QA, and other functions to drive efficiency and effectiveness [0:06:07] – AI Agents and Localization Use of AI agents in customer service to handle technical tickets and customer questions. Heavy focus on localization: AI to adapt website and e-commerce for global markets Integrating translation tools directly into their software so customers can use products in their native language James calls this a "huge lift" for the business. [0:07:45] – Tools: Open "Claw" and Enterprise AI Options They use Open "Claw" (contextually, likely a reference to a general-purpose LLM) through virtual machines across business units, including marketing. Acknowledge recent changes in that ecosystem and the need to adapt. Considering more enterprise-focused alternatives, such as the "Nemo Claw" and other corporate-friendly solutions, with an eye on governance and parameters. [0:08:54] – Their Process for Helping Enterprise Clients First step: deeply understand voice of the customer and specific AI data needs. Industry shift: Moving away from scraping the internet or generic licensing for training data Toward custom, bespoke training sets Unique position of Daz / Tafi: Own the full rights to their 3D content library (avatars, scenes, props) Avoid ethical and legal issues around unauthorized training data Process: li::marker]:font-[sans-serif]" start="1"> Scope the client's needs. Use AI and internal workflows to generate customized datasets. Example: For a major tech company building VR headset experiences, they created tens of thousands of unique characters, avatars, clothing, and assets to train their models. James notes they might be the only company able to deliver this level of tailored solution at that scale. [0:11:34] – The Next 3 Years: How AI Changes Organizations Based on discussions with CEOs and senior leaders: Debate continues on whether AI creates or eliminates jobs; James personally believes it will create jobs. Near-unanimous view: every company must integrate AI into workflows. Employees and companies with a clear AI strategy will win versus those without one. Expectation: Continued job creation plus a strong push for AI-centric efforts in every part of the business Focus on efficiency, output, and more bespoke customer solutions [0:13:25] – Advice to Marketers & Final Takeaways James emphasizes "just get started" and experiment with AI. For marketers: AI enables hyper-specific adaptation to customer needs. Allows new ways of marketing that aren't feasible manually today. Example from their own marketing stack: Their largest marketing channel is email They manage a seven-figure, vetted email list mailed daily AI is used to: Optimize send times Customize products and offers by demographic and other signals Overall message: Embrace AI to become more effective and deliver better, more tailored customer experiences. [0:14:58] – How to Contact James Website: Daz3D – daz3d.com Direct email: [email protected] James invites listeners to contact him directly with questions.