Ross Douthat
Key Views & Dialogues
Why Tech C.E.O.s Are Blaming A.I. for Mass Layoffs
AI now anchors major layoffs at Atlassian, Block, and reportedly Meta, while Block’s 17% share jump shows how productivity narratives can reframe overhiring or mismanagement. Meta’s planned $135 billion capital expenditure could shift costs from payroll to compute, but abandoned Behemoth and delayed Avocado expose execution risk. Token consumption adds a potentially enormous labor cost, making output and revenue—not usage leaderboards—the unresolved test.
View summary (9)
Tech’s latest cuts are an early warning, but not yet clean proof that AI has directly replaced workers. Atlassian is eliminating 10%, roughly 1,600 jobs; Block about 40%, or 4,000; and Meta was reportedly preparing cuts of 20% or more, potentially 16,000. Casey Newton’s calibration is the useful one: AI differs in causal importance at each company, yet “sooner or later, I do think we’re going to have to believe them.”
The hosts disclose relevant conflicts: Kevin Roose works for The New York Times, which is suing OpenAI, Microsoft, and Perplexity; Casey Newton’s fiancé works at Anthropic.
For public companies under pressure, AI offers both a productivity thesis and a marketable explanation for old-fashioned restructuring. Block had expanded from roughly 3,800 employees in 2019 to more than 10,000, then spent $68 million flying 8,000 people to an event with Jay-Z five months before its cuts; its shares rose 17% the next day. That makes “AI-washing” difficult to separate from correcting pandemic-era overhiring and management failures.
Kevin argues that Meta’s proposed labor cuts could shift costs from payroll to AI infrastructure rather than reduce aggregate spending. The company plans $135 billion of capital expenditure this year while Zuckerberg argues that projects once requiring large teams can now be completed by “a single very talented person.” Yet the productivity case remains speculative: Meta abandoned Behemoth, reportedly delayed Avocado for missing targets, and apparently achieved only a slight improvement over Gemini 2.5.
Employees are being placed in a no-win adoption test: use AI heavily to demonstrate alignment, or avoid proving that their work can be automated. The hosts argue that repeated layoffs also quiet internal dissent, whether or not workforce discipline is an explicit objective. Kevin Roose consequently revives a prediction he previously got wrong: fear could drive tech workers toward unionization and bargaining over retraining or reassignment.
Chatbots’ mediocre literary voice may be a consequence of product optimization, not an underlying inability to generate surprising prose. Jasmine Sun preferred aspects of GPT-2 and GPT-3 because they were variable, strange, and capable of style matching, whereas post-training and RLHF pushed later systems toward the “helpful assistant”: chirpy, sycophantic, safe, and repetitive. The commercial demand is for excellent corporate emails, while artistic quality lacks the verifiable rewards that accelerated coding.
The defensible near-term writing product is a personalized collaborator, not an autonomous author. Sun says text generation occupies only about 25% of her work; reporting, idea selection, reading, judgment, and lived experience remain harder to reproduce. Her productive Claude workflow instead learns her archive, retrospective notes, audience, and aspirations, then asks questions that push her toward “the best version of myself as a writer.”
Token consumption is becoming a new and potentially enormous component of technical labor costs. OpenAI’s top employee reportedly used 210 billion tokens in seven days—about “33 Wikipedias” of text—while Anthropic’s highest individual Claude Code user spent more than $150,000 in one month. One Swedish engineer said he probably spends more on Claude than his salary, turning unlimited access into both a job perk and a retention mechanism.
Token leaderboards confuse adoption with output and invite classic Goodhart’s-law gaming. Companies are incorporating usage into performance reviews—an employee might be challenged for using “only” 70 million tokens—despite no clear relationship between consumption and value. The investor-relevant question is therefore not who is token maxing, but whether escalating inference spend produces products, revenue, or merely “tasks of uncertain value.”
Interesting Times: A Mind-Bending Conversation with Peter Thiel
Thiel expects AI to be internet-sized rather than a guaranteed societal rupture, with its economic significance depending on whether it breaks stagnation beyond software and bits. Institutional conformity, regulation, and geopolitical risk may limit the payoff, while efforts to control existential threats could create centralized global surveillance.
View summary (7)
- Thiel’s base case is that AI will be internet-sized, not a total transformation of society. He estimates the internet added a few percentage points to GDP and perhaps 1% annual growth for 10–15 years; AI may do something comparable, creating great companies without ending stagnation. The investable case rests on AI being “the only thing we have,” not on a guaranteed superintelligence cascade.
- The bottleneck to progress may be institutions and heterodoxy rather than raw intelligence. Society already has many smart people yet remains stuck, leading Thiel to doubt that higher AI “IQ Elo scores” automatically produce dementia cures, factories, or Mars colonies. A conformist AI could instead generate infinite acceptable content, displace workers, and deepen stagnation—but he still prefers that risk to “nothing at all.”
- The wider economy remains stagnant because progress after 1970 concentrated in bits while atoms, medicine, transport, and energy slowed. Thiel contrasts 1750–1970’s accelerating ships, trains, cars, aircraft, Concorde, and Apollo with 1985–2025’s superficially similar built environment. His practical tests are whether 30-year-olds live better than their parents did and whether research yields justify rising inputs.
- Growth is not merely upside in Thiel’s model; it is what keeps middle-class expectations, fiscal bargains, and political institutions coherent. He defines the middle class as people who expect their children to do better, while describing Reagan’s “consumer capitalism” and Obama’s “low-tax socialism” as growth-dependent contradictions. Without renewed dynamism, taxes rise or socialism ends; a static “Greta future” would be a radically different and oppressive settlement.
- Trump remains “by far the best option” chiefly because disruption can clear the field for construction—not because Thiel expects effortless technological greatness. Silicon Valley’s shift followed years of increasing the liberal “dose” until companies spent hundreds of millions, “go completely woke, and everybody hates you.” Nuclear deregulation offers a possible path from deconstruction to building, but Thiel now calls political involvement both “incredibly important” and “incredibly toxic.”
- AI undermines the fantasy that entrepreneurs can escape politics by moving offshore or to Mars. Elon Musk told Thiel, “There’s nowhere to go,” while Demis Hassabis’s warning that superhuman AI could follow Musk to Mars underscored that Mars might not function as an alternative political society. The same technology could deter China from Taiwan by widening its disadvantage—or accelerate an invasion under a “now or never” calculation.
- Thiel’s ultimate tail risk is centralized rule sold as protection from technological catastrophe. His formulation, “Antichrist or Armageddon,” links nuclear, climate, bioweapon, and AI fears to global compute control capable of logging every keystroke; the FDA’s and Nuclear Regulatory Commission’s de facto worldwide influence are his milder precedents. Douthat’s strongest challenge is that such totalitarianism may first require the very AI, surveillance, and military systems he says Thiel is investing in.