Martin Shkreli: Big Pharma's Blind Spot, Betting Against Quantum & What Prison Taught Him
嘉宾:Martin Shkreli
精读 · 按主题梳理核心观点、论证与案例。
§01A provocative introduction and a two-part question
The segment opens with a montage that frames the guest through his notoriety: he is described as having been called "the most hated man in America" after raising the price of a life-saving drug by 5,000%, as "the poster child for capitalistic greed," and as someone who spent seven years in prison. The montage includes lines defending profit maximization, such as "You can be prosecuted for not maximizing profit," and a reference to buying the only copy of a legendary Wu-Tang Clan album (rendered "Woutain clan" in the transcript) with no plans to play it.
Ego is the fountainhead of all progress.
The hosts then welcome Martin Shkreli (spelled "Schrelli" in the transcript) and set up the conversation with a two-part question. First, they ask for his thesis on why quantum computing might be a fraud, noting he had said people might go to jail over it. Second, half-jokingly, they ask what advice he would give CEOs who do end up in prison, given his own experience and reports that he became influential inside.
§02The case against quantum computing as a business
The answer begins with personal context. The speaker credits his wife, Chelsea, who is described as one of the first people at OpenAI, as having studied at MIT and taught quantum computing with Scott Aaronson ("Aronson" in the transcript). Over many hours, she explained to him what quantum computing actually is. The speaker also jokes that her father was one of the first people at Nvidia, while in his own family he "was one of the first people in prison."
The central claim is that people wrongly equate quantum computing with faster computing in general. According to the speaker, the field grew out of a thought experiment by Peter Shor ("Peter Schorer" in the transcript): the idea that some algorithms run faster on different physical media. He calls that a revolutionary idea in computing but not in business, because essentially one class of algorithm benefits.
To make the point concrete, the speaker uses sorting, the subject of the opening pages of the standard algorithms textbook CLRS. When he asks quantum enthusiasts which machine sorts a deck of cards faster, he says they answer the quantum computer, and he says that is wrong. His reasons: modern quantum computers run at clock speeds of hundreds or thousands of hertz rather than gigahertz, and they are single-core and single-threaded for a task like merge sort.