OpenAI scales revenue fast, but compute costs test its path to a durable moat

Watch on YouTube ↗  |  January 21, 2026 at 12:08  |  3:41  |  CNBC
Speakers
Daniel Newman — Host, Crypto Banter

Summary

Daniel Newman, CEO of Futurum Research, discusses OpenAI's rapid revenue growth to $20 billion alongside its massive compute spending of $30-50 billion per gigawatt. He questions whether OpenAI can build a durable moat amid rising competition from Anthropic, Gemini, Meta, and open-source models. Newman highlights AI agents as a major monetization inflection and praises the ServiceNow partnership as a strong enterprise AI monetization model.

  • OpenAI revenue reportedly reached $20B, but compute costs are $30-50B per gigawatt.
  • Speaker says OpenAI must lower compute costs or charge more for services to be profitable.
  • Rising competition from Anthropic, Gemini, Meta, and open-source models challenges OpenAI's moat.
  • AI agents are framed as the next major AI monetization opportunity.
  • ServiceNow deal is praised for pairing LLMs with enterprise workloads.
  • OpenAI is seen as playing catch-up in the agents race.
  • Interview was conducted at Davos.
Ideas
Daniel Newman Host, Crypto Banter 1:02
ServiceNow deal enables enterprise AI monetization.
The ServiceNow-OpenAI partnership is a great idea because it pairs a large language model with a company that has huge enterprise workloads, allowing incremental compute monetization while lowering the cost to build future gigawatts; this can monetize the user base much like Meta and Google did in past technology transitions.
Daniel Newman Host, Crypto Banter 3:28
AI agents are next monetization inflection.
Agents are a major monetization inflection for AI and a big opportunity to monetize AI at greater scale, functioning as a workforce alongside humans; competition is rising with Anthropic democratizing code generation and OpenAI playing catch-up, but the shift creates a large opportunity.
Up Next

This CNBC video, published January 21, 2026, features Daniel Newman discussing NOW, AI agents. 2 trade ideas extracted by AI with direction and confidence scoring.

Speakers: Daniel Newman  · Tickers: NOW, AI agents