Episode summary

Why The Anthropic IPO Will Fail — With Gary Marcus

Big Technology Podcast · · Hosts: Ranjan Roy, Alex Stamos, Alex Kantrowitz · Guests: Gary Marcus

Why The Anthropic IPO Will Fail — With Gary Marcus – Big Technology Podcast

Summary

Gary Marcus argues that Anthropic may have strong products, but its IPO could still fail because the business economics look unsustainable. He says the company has grown fast, but its revenue growth is already slowing, its compute commitments are enormous, and its route to profitability is unclear given heavy competition and falling prices across the AI market.

Marcus’s core case is that Anthropic is selling an expensive, highly competitive frontier-model niche while carrying what he sees as nearly $900 billion in future liabilities. He says the company’s earlier growth was inflated by the “token maxing” era, and that the next phase will be defined by weaker growth, commodity pricing, and pressure from Chinese models and other large rivals. Even so, he acknowledges Anthropic has built useful hybrid systems that combine large language models with symbolic techniques, which he says vindicate his long-running neurosymbolic AI thesis.

The discussion also turns to whether Anthropic can expand beyond models into software products like Claude Code and office tools. Marcus concedes that this helps, but says the company still needs those products to work reliably, stay differentiated, and generate enough margin to cover rising infrastructure costs. He expects the most likely outcome is not a huge standalone public winner, but a sale or a much lower valuation if growth and profitability do not materialize.

Key takeaways

  • Gary Marcus says Anthropic can have a good product and still be a bad business if its costs and liabilities outrun revenue.
  • Marcus argues the company’s growth was boosted by the “token maxing” era and is already slowing as that fad fades.
  • He says Anthropic’s only durable niche is high-end models, but that niche is shrinking as competitors and Chinese open-weight models improve.
  • Marcus thinks Anthropic’s software push could help, but only if products like Claude Code and spreadsheet tools work reliably enough to create a real moat.
  • He sees the most likely long-term outcome as acquisition or distress, not a successful standalone trillion-dollar public company.

Notable quotes

“I still don't like the IPO. I'm super thrilled that they're using neurosyolic AI but I still don't think the economics make sense”

— Gary Marcus at 2:51

“I don't think that they're going to be profitable anytime soon. It's not clear that they will ever be profitable and I think there's a real chance they will go bankrupt”

— Gary Marcus at 5:27

“The revenue growth that really was sustaining that notion of valuation is already tailing off and so like they're probably not going to get the perfect storm”

— Gary Marcus at 19:10

“if Anthropic wants to say we did, you know, four points better on the benchmark, suddenly it's going to be 100 billion to them”

— Gary Marcus at 16:02

“I mean the problem again is that they need to be a rocket. like they can't just steadily sell at at the current levels because they're losing money at the current levels and they have these liabilities”

— Gary Marcus at 26:50

Topics

  • anthropic ipo
  • ai valuation
  • ai compute spending
  • competition from chinese models
  • turning ai into profit

Companies and tickers mentioned

  • Anthropic

Watch the episode

Watch “Why The Anthropic IPO Will Fail — With Gary Marcus” on YouTube

Summary, takeaways and quote selection by Octagon. Quotes are short excerpts; the full conversation is in the episode.

More from Big Technology Podcast

All Big Technology Podcast episodes →

Big Technology Podcast, this episode and its artwork belong to Alex Kantrowitz. Octagon is not affiliated with the show; this page is an independent summary.