Overview of Act of Claude: Anthropic’s IPO bid
This episode of The Intelligence from The Economist centers on three stories: Anthropic’s long-awaited IPO filing and what it reveals about the economics of frontier AI; a close Alaska Senate race where fishing, not national culture-war politics, is the defining issue; and a technical but important update from metrology, as experts move toward redefining the second using more advanced atomic clocks.
Anthropic’s IPO: a profitable AI giant with big spending ahead
Miles Kruppa explains that Anthropic’s prospectus gives the clearest look yet at how one of the leading AI firms is doing financially — and what its future may cost.
What the filing shows
- Anthropic’s growth has been extraordinary, with revenue rising sharply after the launch of Claude Code.
- The company reported:
- $11.6 billion in revenue in Q2
- Over $900 million in operating profit
- That makes it unusually strong for a private AI company of its size and age.
Why investors are still cautious
- The filing makes clear that current profits may not last at the same level.
- Anthropic expects major increases in:
- capital expenditures
- data center rental commitments
- chip and compute spending
- The company is signaling that the path to staying at the AI frontier will likely be expensive.
Strategic shift in infrastructure
- Anthropic has relied heavily on cloud providers like Google, Microsoft, and Amazon.
- It is now moving toward:
- buying chips directly from vendors such as Broadcom
- installing them in its own leased data centers
- This could improve long-term control and profitability, but it is capital-intensive.
Safety, governance, and the IPO message
- Anthropic presents itself as an AI safety-first company and a public benefit corporation, allowing it to weigh more than shareholder profit.
- The filing includes extensive discussion of risks, including:
- models behaving unpredictably
- catastrophic misuse
- the difficulty of insuring against AI harms
- The episode frames Anthropic as trying to prove that a frontier AI company can be both profitable and responsible.
Alaska’s Senate race: fishing as the dominant political issue
The episode then shifts to Kodiak Island, Alaska, where correspondent Aaron Braun reports from a Senate debate that shows how local issues can outweigh national polarization.
Why Alaska is unusual
- Alaska is geographically huge but sparsely populated.
- Party identity is less rigid than in most states.
- The state has a tripartisan coalition in its legislature.
- Even with Trump popular there, Democrat Mary Peltola remains competitive.
The candidates
- Mary Peltola:
- former Democratic congresswoman
- native Alaskan
- emphasizes her fishing background and local roots
- Dan Sullivan:
- Republican incumbent senator
- focuses on defending fishermen and criticizing Peltola’s record
The central issue: fisheries
- Fishing is not just an industry in Alaska; it is a way of life and a major political issue.
- Peltola uses local identity and industry expertise to appeal beyond party lines.
- Sullivan argues she has not done enough for fishermen.
- Peltola counters that Sullivan has not gone far enough, especially on protecting wild Alaskan fish stocks rather than farmed fish.
Federal cuts and NOAA concerns
- Peltola also uses national issues strategically, especially:
- federal staffing cuts
- weakened government services
- NOAA data reliability
- Alaskans depend on NOAA for:
- weather information
- fish stock monitoring
- quota decisions
- The episode notes that Alaska has lost a notable share of its federal workforce, which could affect remote communities and fisheries management.
Broader takeaway
- The race is close and may take time to count because of Alaska’s geography.
- The debate illustrates how a candidate can remain viable in a deeply red state by focusing on practical local concerns rather than national ideology.
Redefining the second: atomic clocks push timekeeping forward
The final segment looks at a surprisingly consequential standards issue: redefining the SI second.
Current definition
- The second is currently based on cesium-133.
- It is defined as 9,192,631,770 cycles of radiation associated with cesium atoms.
- This system underpins:
- international atomic time
- satellite navigation
- financial systems
- telecommunications and other precision technologies
Why change it now
- New atomic clocks using elements such as:
- strontium
- ytterbium
- aluminum
- lutetium
- are now significantly more accurate than cesium clocks.
- The issue is that modern clocks are becoming so much better that the old definition of the second is now the limiting factor.
The challenge
- Redefining the second would improve precision and better match scientific reality.
- But the process is not purely technical:
- different national labs have competing clock designs
- there is prestige in having your standard adopted internationally
- politics among metrology institutions matters
Bottom line
- The change won’t affect daily life in obvious ways, but it matters for the precision infrastructure behind modern society.
Key takeaways
- Anthropic’s IPO filing suggests real revenue and profit, but also major future compute costs.
- The company is trying to convince investors that scaling frontier AI requires heavy upfront spending, not just software margins.
- Alaska’s Senate race is being shaped by local industry concerns, especially fishing and federal resource management.
- Scientists are preparing to redefine the second because atomic clocks have outgrown the cesium standard.
Notable themes
Profit vs. ambition
Anthropic is trying to show that frontier AI can be commercially viable without abandoning safety.
Local politics over national identity
In Alaska, voters appear more responsive to concrete issues like fisheries than to national partisan messaging.
Precision as progress
The timekeeping segment underscores how scientific standards evolve when technology becomes more accurate than the definitions built to measure it.