The week that was in AI, more opinions on the Anthropic IPO, and a Must Read on “The End of the World as Business Model” (updated)

History before our eyes. Or how the mighty can break a leg, and it isn’t over. By now you’ve absorbed another departure from flight, the resignation (read, ouster by the board) of Nvidia co-founder and CEO Jensen Huang. The worst of the revelations was his extra-marital relationship with an alleged Chinese intelligence asset, a certain Alice Huang (no relation; likely this is an anglicized version of a Chinese name). Piling on are allegations of an affair with Nvidia’s VP of healthcare, Kimberly Powell, as well as a third Nvidia employee. Nvidia is probably the most valuable company right now in the world, the 9,000 lb. bull elephant in producing graphics processing units (GPUs), central processing units (CPUs), and artificial intelligence infrastructure software. The company that was born in a Denny’s (a chain diner) that nearly went broke a few times, that was linchpinned by Huang’s marriage to an early microchip designer he met at Oregon State University, now his wife, who now co-controls millions of Nvidia shares in family trusts.

Alice Huang and Megaspeed, a Singapore company, are being investigated by the US Department of Justice (DOJ) on circumventing American export restrictions on advanced AI chips into China, plus reportedly a separate effort to smuggle Super Micro server products with Nvidia chips into China. Huang, a Hong Kong native, was formerly CEO of Megaspeed. Today she was barred from reentering the country over inaccurate information provided to Singapore police. Singapore is reportedly frustrated with the slowness of the US law enforcement response as well. South China Morning Post

Today Nvidia stock bounced around a bit, but it is a company too dominant, too much part of everyday infrastructure and AI, to take much of a hit. For investors, it may be a buying opportunity. The Nvidia board swiftly named an interim CEO, the well-regarded CFO Colette Kress. Yahoo Finance

Updated 18 Nov: Despite calls for Jensen Huang’s reinstatement by Elon Musk and President Donald Trump, the Nvidia board has refused to bring him back, citing its own investigation into his behavior and whether it compromised national security or trade secrets. They have hired Wachtel Lipton to conduct said investigation. CNBC  And the rumor mill is swirling on cui bono, with Anthropic leading the list of Suspects. Huang was an advocate of open-sourced AI, which isn’t in ANT’s or OpenAI’s interest. At all. Short sellers? Congress? Howard Lutnick at Commerce versus Scott Bessent at Treasury over not blacklising Megaspeed? In any case, Senator Tim Scott (R-SC) has requested that both Huang and Lutnick show up before the Senate Committee on Banking, Housing, and Urban Affairs before Christmas leave on 3 and 10 December respectively. Yahoo Finance, Bassell Haidar on LinkedIn

Of course, Mr. Huang retains a considerable number of shares: 800 million to 812 million shares (about 3.4% to 3.5% of the company), most of which are in joint trusts held with Lori Huang. Not inconsiderable, but not the kind of control that Larry Ellison has at Oracle (40%).

Rather harder to believe are the all too expansive blatherings of Anthropic’s CEO Dario Amodei in his IPO roadshow video. This Editor, who comes not only from marketing but also from Startup/Early Stage Health Tech Communications World, is rather inured to “save the world” claims made by CEOs. Fine for the press, but not for the investors; “say it safely”, man! Example: we all made book on (well, some of us hoped) that telehealth and associated healthcare tech would make healthcare more affordable, accessible, and available to all from NYC to Uttar Pradesh. And in large part it’s done so, but it took two decades, a pandemic, a trail of tears, and high profile failures to do it.

So here we have a video with CEO Amodei making this absurd claim: “If today we look at things like cancer, Alzheimer’s, heart disease and mental illness, these will be relics of the past in I think less than a decade.” Can Amodei and Anthropic deliver? Of course not. Anyone in healthcare, not even biotech, would be able to rebut him. Since drug development not only is tied to a long-term clinical timeline and a clinical trial for safety and efficacy framework that could be shortened (good) but also that every one of these diseases is multi-factorial–especially cancer, but also diseases of the brain such as dementia and mental illness (not good)–any claim that any or all of these diseases will be ‘relics of the past’ in 10 years is pure hype. Improving healthcare? Probably. Even the eternal optimist Eric Topol, MD can’t get on board with Amodei. Nor can the chair and CEO of Eli Lilly, Dave Ricks. On X, Amodei was savaged or, less dramatically, ROTFL’d. 

You don’t want to be either before a zillion-dollar IPO.

To say these things in an IPO video is clownish, but also verges on investor fraud. SEC, where are you? 

It definitely takes away from Anthropic’s business case for the public investor, to put it mildly. Constellation Research has a sober take on the IPO, from ‘absurd’ to ‘interesting’; if you like numbers, this is for you. Over on Substack, Marcus on AI lands on the side on the absurd and compares Amodei with Elizabeth Holmes, along with the hindquarter leavings of horses and bulls. In this Editor’s view (in this article’s comments), she expressed the deep desire of Anthropic’s investors to exit as they are so far down the rabbit hole they are looking to tunnel out, thus the urgency to IPO. She also commented on another comment likening tech CEOs to cult leaders, which is a really chilling thought.

Which leads us to this week’s long Must Read, The End of the World as Business Model. It is published on an unlikely platform,  LinkedIn, by an author I haven’t come across before, Joe Lang. It is a finely tuned roundup of the panic around AI and P(doom) ending humanity, which were the hot topics that closed out the summer [TTA 17 Sept]. But it moves on to outline and sledgehammer the contradicting narratives ranging between P(doom) and AI as Mankind’s Savior. It touches on the philosophy, from the neo-paganist (my term) worship of computers, to the ‘singularity’ where man loses control over technology, to ‘effective altruism’ which is a radical form mixing utilitarianism and alleged  rationality, a philosophy that would give Ayn Rand, the ultimate rationalist, indigestion on its premises.

Mr. Lang and I agree that no one is going to stop this freight train, but that we the public should recognize that within the proponents of AI there are different groups pulling our strings and attempting to confuse us: Doomers increasing our fatalism to cover their writing the rules, Accelerationists who point to national interest while they pull up the ladder on small AI, Centrists who want regulation but don’t want Big AI writing the rules, and Believers in the triumph of the singularity. There are too many people within AI ‘high on their own supply’. Real issues get obfuscated: data centers with their huge footprint, energy (acknowledged by Jensen Huang), and water needs, the destruction of jobs ironically in the information sector and management overseeing systems, and very real cognitive decline researched quantitatively among heavy and young AI users. What is inevitable? (F)doom–financial doom–will catch up to many AI companies well before (P)doom. (Whether a failure of Oracle, OpenAI, or even Anthropic will crater the economy scares me more than (P)doom.)

Towards the end, Mr. Lang discusses the difference in using AI as scaffolding or substitution for the thought processes that go into doing the work. Warning: it is a long way to the end of this article, he touches on nearly everything, but ultimately it will be worth your effort. 

Categories: Latest News and Opinion.

Leave a Reply

Your email address will not be published. Required fields are marked *