Podcast Summary: The AI Podcast
Episode: Anthropic Passes OpenAI in Revenue, Cerebras Files IPO, Sora Shuts Down
Date: April 20, 2026
Episode Overview
This episode unpacks a sequence of major news stories shaping the AI industry as of April 2026, including:
- Anthropic overtaking OpenAI in revenue and what it means for the industry
- Major disruptions at OpenAI, highlighted by the shutdown of Sora and executive departures
- Cerebras making a move to go public, directly challenging Nvidia’s dominance
- Stanford’s AI Index: rapidly closing performance gap between US and Chinese models, and a leap in AI agent capabilities
- The surprising boom in app releases, defying predictions that AI would “kill” the App Store
Throughout, the host combines reporting with sharp personal insight, making this a must-listen for anyone following AI’s competitive landscape.
Key Discussion Points & Insights
1. App Store Boom Defies AI ‘Death’ Predictions
[01:27–03:55]
- Contrary to widespread predictions, AI hasn’t killed the App Store—app launches are up 60% year-over-year and 104% in April 2026 on iOS.
- “App launches are up 60% year over year and over a hundred percent so far in April. So we're going to break down why. And spoiler alert, it is because of AI.” (Host, 01:27)
- Productivity and utility apps are driving this surge.
- The host attributes this to ‘vibe coding’—non-programmers using tools like Claude, Replit, Cloud Code—which allow anyone with an idea to build and publish apps with minimal technical background.
- Apple exec Greg Joswiak is quoted: “The rumors of Apple stores death have been greatly exaggerated.”
- Host's analysis: Tools enabling non-developers are driving the “gold rush” in app creation.
- Offers a plug for the host’s own “AI Hustle school community” for learning to scale businesses with AI.
2. OpenAI: Sora Shuts Down & Executive Exodus
[03:56–06:49]
- OpenAI announced the shutdown of Sora (text-to-video tool) on April 26, with the API following in September.
- Sora lost momentum after peaking at 1 million users, dropping to under 500k, and was incurring unsustainable compute costs ($1m–$15m/day).
- On the same day the shutdown was announced, three top execs exited: product chief Kevin Whale, Sora head Bill Peebles, and the enterprise CTO.
- “Three exits in the same 24 hours. What I think is happening right now is that OpenAI is making a very hard pivot. They're focusing on coding and enterprise, which makes a lot of sense. That's where Anthropic is absolutely eating their lunch right now.” (Host, 05:41)
- Host sees this as a strategic refocus toward coding and enterprise tools, areas where OpenAI is under pressure from Anthropic.
3. Cerebras Files for IPO; Challenges Nvidia
[06:50–09:20]
- Cerebras, an AI chip company, is set to IPO at a $23 billion valuation, following a $10B deal with OpenAI and an AWS agreement.
- The company positions itself as a direct competitor to Nvidia, claiming to have won the “fastest inference business” from OpenAI.
- CEO Andrew Feldman:
- “Nvidia didn't want to lose the fast inference business at OpenAI. And we took that from them.” (Host quoting Feldman, 08:51)
- Cerebras' financial highlights: $510 million in 2025 revenue, $237 million net income (some one-time items included).
- Host’s analysis: This represents the “first real crack in Nvidia's monopoly at the frontier.”
4. Stanford AI Index: US vs China & AI Agents
[09:21–11:14]
- The US holds a 2.7% lead over China's best models (down from 17–32 points in 2023) despite vastly greater investment.
- “China is doing this in…for way less money. U.S. private AI investments was about $286 billion last year. In China it was about 12.4 billion, roughly, you know, 23 times less. And their top model is under 3 points behind. So that is pretty insane.” (Host, 10:08)
- AI agents (autonomous software bots) now succeed at 66% of real computer tasks (up from 12% a year ago).
- The host observes personal success rates of “like 85%” for common tasks:
- “It feels like we've moved forward a ton. …It is just insanely impressive. So you know, I think this is basically something that everyone should be trying right now.” (Host, 11:00)
5. Anthropic Surpasses OpenAI in Revenue
[11:15–15:40]
- Anthropic hit a $30B annualized revenue run rate, overtaking OpenAI’s $25B for the first time.
- Growth has been rapid: $9B (Dec 2025) → $20B (Mar 2026) → $30B (Apr 2026).
- “March to April, $20 to $30 billion. That's a $10 billion jump. It's taken years to get here…this is mind boggling. This is insane.” (Host, 12:47)
- The host reports a complete personal shift from ChatGPT to Anthropic’s Claude models, citing task automation capabilities and integrated agents that do non-API work (e.g., scraping, reporting, scheduled workflows).
- 80% of Anthropic’s revenue comes from business customers; OpenAI remains more consumer-focused (900M+ weekly active ChatGPT users).
- Anthropic’s business customer base doubled in two months (from 500 to 1,000+ customers each spending $1M+ per year on Claude).
- Anthropic spends 4x less on model training compared to OpenAI, drastically improving unit economics.
- Massive infrastructure deals: 3.5GW of next-gen TPU capacity (Google/Broadcom) for 2027, plus 1GW for 2026—“roughly the output of three plus nuclear reactors dedicated to running Claude.”
- Strategic implications:
- The long-held assumption of OpenAI’s insurmountable lead is broken.
- Anthropic is now the “enterprise AI default,” and has relevant Microsoft integrations.
- Host: “It turns out I think having the best coding model and being the kind of enterprise default is way more valuable than being the consumer default. At least while enterprises are the ones actually paying, you know, full freight for AI.” (Host, 14:41)
- Host personally pays $200/month for Claude code but estimates $30,000/month in value.
- OpenAI is predicted to fight back by refocusing on core strengths, setting up dynamic competition ahead.
Notable Quotes & Memorable Moments
-
On the App Boom:
- “The rumors of Apple stores death have been greatly exaggerated.” (Host quoting Apple’s Greg Joswiak, 02:45)
- “Vibe coding is working. Clog, cloud, code, replit. All of these tools have gotten so good that people with ideas and no engineering background are shipping apps right now.” (Host, 03:12)
-
On Sora’s Demise & OpenAI’s Pivot:
- “Then on basically the same day, the shutdown timeline locked in. OpenAI lost three of their major executives…So this wasn't fantastic that 3 exits in the same 24 hours.” (Host, 05:18)
-
On Cerebras vs Nvidia:
- “Nvidia didn't want to lose the fast inference business at OpenAI. And we took that from them.” (Host quoting Andrew Feldman, 08:51)
-
On US-China AI Competition:
- “Top model is under 3 points behind. So that is pretty insane.” (Host, 10:13)
-
AI Agents’ Leap:
- “AI agents went from about 12% success on real computer tasks a year ago to 66% now.” (Host, 10:29)
-
Anthropic’s Rise:
- “Anthropic is literally doing things for me. Like as we speak, I have agents on my, you know, side panel that are gathering data, building out websites, doing reports, doing things that there is no API for.” (Host, 13:30)
- “They're making more money with smaller training runs. And if that holds, basically the unit economics on the Anthropic side look way better than for anyone else on the frontier.” (Host, 15:01)
- “That's roughly the output of three plus nuclear reactors dedicated to running claude.” (Host, 15:27)
Key Timestamps
- 01:27 – App Store & app launch trends
- 03:55 – OpenAI’s Sora shutdown & executive departures
- 06:50 – Cerebras IPO and challenge to Nvidia
- 09:21 – Stanford AI Index takeaways (US vs China, AI agents)
- 11:15 – Anthropic outpaces OpenAI in revenue; deep dive on why
- 15:40 – Wrap-up and final thoughts
In Summary
This episode captures a pivotal moment in AI’s evolution: Anthropic’s explosive, enterprise-driven ascent; OpenAI’s shakeup and narrowing focus; the shattering of Nvidia’s AI chip monopoly; and unnerving signs that China is catching up with far less investment. Meanwhile, the supposed death of the App Store has been replaced by a creative boom, as AI democratizes software creation. The host’s firsthand experiences and strong optimism make this episode both credible and energizing for listeners wondering where the AI arms race is heading.
