In this interview, Legendary Entertainment founder Thomas Tull explains his investment philosophy: using long-term capital and embedding AI teams inside traditional industries like insurance and manufacturing, and that 'avoiding stupidity' matters more than 'being smart.' He is cautiously optimistic about AI adoption—slower than expected but with huge long-term potential. Key holdings: Acrisure (insurance broker, AI embedded for transformation), Figs (medical scrubs, early-stage tech support), and Colossal (gene-editing startup aiming to revive the woolly mammoth, led a $15M seed round).
Thomas Tull, founder of Legendary Entertainment, produced box-office blockbusters including The Dark Knight trilogy, the Hangover series, and Inception. His investment firm Tulco focuses on leveraging artificial intelligence (AI) to transform large-scale industries. He is also a part-owner of the Pi
Guest: Thomas Tull, founder of Legendary Entertainment, producer of box office hits such as the Dark Knight trilogy, the Hangover series, and Inception. His investment firm Tulco focuses on leveraging artificial intelligence (AI) to revolutionize large industries. He is also a part-owner of the Pittsburgh Steelers and plays guitar in the band Ghost Hounds, touring with the Rolling Stones.
Core Judgment: Tull views his cross-domain success (film, AI investment, sports, music) as a unique advantage in his investment research, emphasizing the disruptive application potential of AI in traditional industries, but his core investment philosophy is "avoid doing stupid things" rather than "pursue doing smart things."
Tull argues that introducing long-term institutional capital into traditional industries (e.g., the film industry) is a prerequisite for disruptive innovation, because short-term market pressure (such as quarterly earnings reports from listed companies) distorts management's decision-making, preventing it from bearing the risk of "breaking the rules."
Historical Context: When Tull founded Legendary in 2004, the film industry was a $30 billion business, but with no institutional capital involved—all funding came from individual investors or bank loans. He successfully persuaded institutions such as Fidelity and T.Rowe Price, becoming the first player to bring "permanent institutional capital" into the film industry. He explained: "If I raise money from my successful neighbor who is a dentist, next year that dentist might say 'I need the money,' whereas institutional capital can take a 5- to 10-year view on an investment."
Mechanism Breakdown: Tull's holding company, Tulco, uses "permanent capital" (a holding company structure) rather than the "vintage year" model of traditional funds (which must return capital within an agreed timeframe). This structure allows him to "buy entire companies or majority stakes and embed the AI lab team inside them," without worrying about short-term exit pressure.
Comparative Data: Tull notes that management at publicly traded companies (such as traditional film studios) is constrained by a system that "evaluates them quarterly and ties bonuses to stock prices," making it impossible to say to management: "Don't count on our financial forecasts for the next two years, because we are reshaping the business." He adds: "Capitalism does not work that way right now."
Extrapolation: Tull believes that this "unruly capital structure" is the key to his successful application of AI in traditional industries such as insurance and manufacturing. Falsification Condition: If a majority of the companies in Tulco's portfolio fail to achieve long-term value, the perceived advantage of its capital structure would be overestimated.
Tull's core approach: Embedding Tulco's AI lab team directly inside the traditional companies it acquires, rather than outsourcing to a third party, because "traditional industries lack technical talent and cannot independently recruit top-tier AI teams."
Case Mechanism: Tull cites the insurance company Acrisure (headquartered in Grand Rapids, Michigan) as an example — its CEO Greg Williams "truly understands the power of AI in actuarial science, predictive analytics, and automating back-office processes." Tulco's AI team was "placed inside the company" to help it transform using its existing data systems. He explains: "We did a deal that was entirely AI-driven, selling off some insurance assets."
Another Case: Medical scrubs company Figs — led by two female founders. Tulco helped early on to "build the platform, recruit staff, and ensure the company embraced technology," while the founders were "extremely obsessed" with their customers (healthcare workers) and the product. Tull believes success comes when "obsession with the customer" is combined with "embracing technology."
Competitive Landscape: Tull emphasizes that most large traditional companies "cannot easily pivot to technology" because "changing the course of such a large ship is very difficult, especially if technology is not part of the company's DNA." He adds: "I hear people say, 'Why can't they transform?' — because it's hard."
Extrapolation: Tull believes the pace of AI adoption in traditional industries is "slower than most people realize," because everyday tasks like maintaining balance while standing remain extremely difficult for AI. Time Horizon: Over the next 5–10 years, AI penetration in industrial sectors (e.g., insurance, manufacturing) will accelerate, but AI applications in the physical world (e.g., robotics) will take longer. Uncertainty: Tull acknowledges that "AI is happening, but it is happening more slowly than people imagine."
Tull sees “judging people” as the core of investing, and the standard is not the business plan (“plans change”), but rather “whether, when bad news comes, this person is willing to be the first to call and tell you.”
Judgment framework: Tull cites the approach of Buffett and Munger: “We don’t think about ‘what smart things to do,’ but rather ‘what stupid things could we do here’ – this stays with me whenever I analyze a problem.” He adds that in investing, “bad news should travel faster than good news.”
Specific indicators:
Connection to personal experience: Tull recalls growing up “poor in upstate New York, with a single mother working two jobs,” which taught him “no excuses, you do what has to be done.” He has translated this into an investment philosophy: “If you know someone is going to ‘screw things up,’ you’d rather find out in advance than be forced to react after the fact.”
Deduction: Tull believes that in investing, “people” are 100 times more important than the “business plan,” because “plans change, but a person’s character and ability to adapt do not.” Falsification condition: If, within Tulco’s investment portfolio, the failure rate attributable to founder character issues is higher than the industry average, then his judgment framework is invalid.
Tull believes that his successful experience across multiple fields—film, music, sports, AI investing—enables him to understand the essence of "value creation" more comprehensively than traditional financial investors. The key, however, is to maintain the calm of a "third-party analyst."
Mechanism analogy: Tull compares filmmaking to "every project being a startup"—which requires "a great director (CEO) + a great script (business plan) + chemistry among actors (team collaboration)." He emphasizes: "In the film industry, if the director and script are wrong, nothing else can make up for it." This aligns with his approach in AI investing: "first look at the founding team, then at the business plan."
On the "comfort zone": Tull notes that his company Tulco's "outsider" perspective (stemming from non-traditional financial backgrounds such as film and music) helps him avoid "making mistakes due to industry conventions" when evaluating traditional sectors. He quotes Buffett: "The most important thing in investing is not 'being smart,' but 'not being stupid.'"
Difference from market consensus: Tull explicitly states that when Netflix decided in 2015 to "release all episodes at once," "everyone I knew in Hollywood said, 'They are idiots,' but it turned out they were right." He believes that such "counter-consensus" judgments require a cross-disciplinary perspective—only those who simultaneously understand technology, capital, and the psychology of creators can identify true disruption.
Extrapolation: Tull believes that over the next 10 years, virtual reality (VR) will become the next "great storytelling medium," but this will require "new creators—people we don't know yet—to understand how to create an emotional experience in this environment." Uncertainty: He acknowledges that "the commercialization timeline for VR remains unclear, but it is certain in the long run."
| 标的 | 嘉宾态度 | 关键数据 |
|---|---|---|
| Figs | 看好（投资支持） | 医护工作服公司，由两位女性创始人领导，Tulco早期帮助建设平台和AI能力 |
| Acrisure | 看好（投资支持） | 全球最大保险经纪之一，AI驱动交易，Tulco团队嵌入其内部 |
| Colossal | 投资（种子轮领投） | 1500万美元种子轮，目标通过基因编辑恢复灭绝物种（如猛犸象） |
| Re:Build Manufacturing | 投资（个人大幅投资） | 美国高科技制造业投资公司，由MIT校友和亚马逊前高管Jeff Wilkie共同创立 |
| Legendary Entertainment | 未明示（已退出） | 2004年创立，制作Dune、Jurassic World、Dark Knight Trilogy等，2018年出售给Wanda Group |
1. "Avoiding stupidity is more important than seeking brilliance": Tull cites the Buffett/Munger approach — in investing, "doing smart things" is less effective than "not doing stupid things," because "smartness" often leads to overconfidence, while "avoiding stupidity" is the cornerstone of long-term value. (Support: Tull reveals that every time he analyzes a problem, he asks: "What is the stupid thing we might do here?")
2. "Long-term institutional capital is a necessary but not sufficient condition for innovation success": Tull believes that the reason traditional industries lack institutional capital is "structural barriers" — short-term pressure prevents management from bearing the cost of trial and error, while Tulco's permanent capital structure is its core advantage. (Support: In 2004, he introduced institutional capital to the film industry, which had $30 billion in scale with zero institutional participation before that.)
3. "AI deployment in traditional industries requires embedding teams inside the company, not outsourcing": Tull uses the Acrisure case to illustrate that AI teams must "eat and live" with business teams to understand industry-specific data structures and decision-making logic. (Support: The Acrisure transaction was "entirely AI-driven," but only because "we already had a data collection system in place.")
4. "In investing, a person's character is 100 times more important than the business plan — because plans change, but character does not": Tull emphasizes that when evaluating founders, the key is to see "whether he calls you first when bad news arrives" and "whether he is willing to say 'I don't know.'" (Support: He cites the "Theranos" case, arguing that most fraud comes from "a small step toward unethical behavior," not overnight villainy.)
5. "Constraints are catalysts for innovation, not obstacles": Tull believes that too much capital actually weakens resilience, because "without the risk of failure, there is no real motivation for innovation." (Support: He observed in the film industry that "if the budget is unlimited, directors lose creativity, and appropriate constraints such as deadlines and budget limits force teams to generate genius ideas.")
6. "Cross-domain experience is a 'dimension-enhancing weapon' for investment research": Tull combines the "every project is a startup" mindset from filmmaking with AI investing, believing he can simultaneously understand the three languages of "technical feasibility, capital structure, and creator psychology." (Support: He successfully predicted Netflix's disruption, even though Hollywood at the time unanimously thought "Netflix is an idiot.")
7. "The key to reshoring U.S. manufacturing is not tariffs, but 'human-machine collaboration efficiency'": Through the Re:Build Manufacturing project, Tull advocates using AI and robotics to enhance productivity at the "human-machine interface," rather than simply moving factories back. (Support: He cites the lesson of supply chain ruptures during the pandemic, arguing that "the cost of self-sufficiency should be factored into national calculations.")
8. "Investment philosophy: From 'doing the right thing' to 'not making mistakes' — it is a form of contrarian thinking": Tull believes that most investors fall into the trap of "seeking brilliance," while true wisdom is "doing the basics well and maintaining a long-term perspective." (Support: He quotes Buffett's view that "Don't try to do smart things, but avoid doing stupid things," and calls it the "only thing I am sure of.")