Moonshots with Peter Diamandis | Ep. 198

"The AI War: OpenAI Ads & Sora 2, Grok Partners With US Government & Google’s Ad Business Is at Risk"
Guests: Dave Blundin, Salim Ismail, Alex Wissner-Gross
Date: October 4, 2025

Episode Overview

In this future-focused episode, Peter Diamandis and his panel of technology moonshotters break down the latest seismic shifts in artificial intelligence, robotics, quantum computing, and longevity biotechnology. The conversation explores how advances in generative AI, codegen, compute, and data center infrastructure are rewriting industries—from advertising to finance to energy, and even medicine. Discussion is peppered with reflection on the accelerating pace of change, economic disruption, and the tension between AGI democratization and newly emerging AI moats.

Key Discussion Points & Insights

1. AI Video & Audio Generation: Sora 2, Meta Vibes, and SUNO 5

• Transition from Content Selection to Content Generation (00:04, 06:03):
  • "We're seeing... the transition from algorithmic content selection in social media to algorithmic content generation." – Alex Wissner-Gross (00:04)
• Meta’s New Vibes App & Rapid AI Partnerships (06:03–07:15):
  • Meta launched Vibes, a TikTok-like app for AI-generated video, but despite deep pockets, it partnered with startups like MidJourney and Black Forest, highlighting that “the really smart creative people all want to do startups and they don't want to join the big companies.” – Dave Blundin (06:37)
• Sora 2’s Viral Growth Model (08:45–09:52):
  • Sora 2 allows customizable, physically consistent, real-world video generation—now with easy access and viral, permission-based sharing.
  • “The interface to create it is entirely voice and prompt. There's no coding and no interface… all of that is about to disappear… and just go to a straight natural language interface.” – Dave Blundin (10:16)
• SUNO 5 – The Musical Turing Test (16:41):
  • “Anyone can compose a top 40 song or an opera, and this is the beginning maybe of disposable or casual art.” – Alex Wissner-Gross (17:05)

2. The Coding AI Wars: Anthropic’s Sonnet 4.5, OpenAI, and Benchmarks

• Sonnet 4.5 & Next-gen Code Agents (18:18):
  • Anthropic’s latest model excels at codegen; “if the code can write itself really well, maybe that's the critical path to an intelligence explosion.” – A.W-G. (18:38)
  • New industry benchmarks (e.g., SWE Bench, MIT’s forthcoming long-form code benchmarks) are crucial for tracking superhuman software development.
• Just-in-Time Software and Imagine with Claude (24:06):
  • Softwares are now generated on the fly (“every pixel”) based on user interaction, not prewritten code—heralding an era where “every software event results in new codegen on demand.” – A.W-G. (27:39)
  • Raises questions about slop (algorithmically generated junk) but also unleashes value for transformative tasks.
• Agentic Compute: Pulsed AI and Persistent Work (29:19–30:48):
  • OpenAI’s ChatGPT “Pulse” starts prompting users and allows for background, persistent tasks—opening possibilities for agents to “run for days or weeks on end” (30:13), pursuing major goals like curing disease.
• AI Model Integrity—Lying, Power-Seeking, and Red-Teaming (22:05):
  • Sonnet 4.5 is reported to have reduced its ability to “lie and seek power by a factor of 10,” and there’s an ongoing, crucial effort across labs to benchmark and red-team AI stability and alignment.

3. AI Monetization: Ads, Partnerships, and the Threat to Google

• OpenAI Introduces Ads (29:26–34:44):
  • “The AI is going to be incredibly good at convincing you to do things whether they're right or wrong… there's a huge incentive to get really aggressive with the advertising.” – D. Blundin (31:44)
  • Tension between trust and monetization: “Seem to be for like, will you trust insights from an AI that has ads baked into it and has an ulterior motive?” – S. Ismail (32:46)
  • Peter predicts the ad model dies in favor of agentic commerce, where “attention is going to equate to some level of interest.”
• OpenAI Partners with Stripe and Ecommerce (37:24):
  • Enabling instant agentic commerce could “be a big threat to Amazon.” – S.I. (40:12)
  • “AI is going to eat the whole economy” (41:16), starting with CPG but rapidly expanding to real estate, healthcare, and utilities.
• Google’s Ad Business at Risk (Interlaced Throughout):
  • If AI assistants handle purchasing and comparison, traditional search ads become obsolete.

4. Benchmarking & Economic Impacts: GDP VAL, Merkor’s APEX, and CFA Automation

• Real-world Task Performance and Knowledge Work Benchmarks (43:49–45:50):
  • New benchmaks (GDP VAL and APEX) show AI models "100 times faster and cheaper than human experts."
  • “This is real world stuff. And now we have the ability to gauge AI doing real world stuff. And now this becomes very tangible.” – S.I. (45:17)
• Merkor’s APEX Index & Founder Trajectory (46:10–48:19):
  • Young founders are achieving “$10 billion valuation” at unprecedented ages; APEX aims to benchmark knowledge work at scale.
• CFA Exam Passed in Minutes (61:01–62:24):
  • “All their finance jobs essentially get rewritten now and recreated. It's a body blow to the accounting world.” – S.I. (61:36)
  • “Do I… have access to the best investment advice that Warren Buffett has access to as well? Is this leveling the playing field?” – P.D. (62:02)

5. Quantum Computing & Math: Timeline, Killer Apps, and Security

• Quantum Compute: Cautious Optimism, Waiting for the Killer App (65:16–66:25):
  • “What is missing right now… is the killer app for quantum accelerated compute.… No one has yet published the killer app for Quantum ML.” – A.W-G. (65:16)
• Math Will Soon Be ‘Solved’ by AI (54:24–59:54):
  • “If the process of mathematics has been solved to the extent that we have a clear line of sight… any mathematical problem that's solvable will be solved with the same algorithms, just with a lot more compute.” – A.W-G. (55:37)
  • Implications ripple into cryptography, physics, economics—threatening to upend major industries (58:44).
  • “We may find ourselves in a world two to three years from now where we're just drowning under math, science, engineering being solved in rapid succession.” – A.W-G. (59:41)

6. Compute, Chips and Data Centers: Capital, Competition, and Scarcity

• OpenAI’s Compute Moonshot: 10GW per Week (71:18–73:30):
  • Sam Altman urges industry-level commitment to massive compute expansion: “We want to create a factory that produces a gigawatt of new AI infrastructure every week.” (71:18)
• Race of the Hyperscalers & Sovereign Backing (71:31–74:03):
  • Sam’s deals ($100B, $300B) are dwarfed by Zuckerberg ($600B) and Google.
  • Tiling the Earth in data centers raises questions about energy and infrastructure, but also triggers “build up to build down” scenarios, where efficiency breakthroughs may reverse the need for brute scaling (74:03).
• Nvidia’s GPU Leasing & Financial Innovation (76:21–80:11):
  • Nvidia pivots to leasing to address capital mismatch between hyperscalers and AI startups: “Who's going to fund it?… Nvidia is saying, OK, we'll fund our own chips using our massive balance sheet and… market cap.” – D.B. (77:25)
  • Massive data center growth is now outpacing most real estate investment sectors.

7. AI, Policy & National Strategy: Sovereign Chips & Robotics

• Intel as Team America’s Linchpin (80:41–86:05):
  • Sovereign venture capital: “The US Cannot afford to let intel fail.… This was almost predetermined… by Moore’s Second Law.” – A.W-G. (83:29)
  • Security threats from foreign-manufactured chips underscore the urgency.
• Scarcity and Sovereignty in Robotics (110:42–113:34):
  • China dominates the supply chain and is exporting robots at scale. The US may need a “Manhattan-style project” to regain lost ground (112:31).
  • Scarcity in robots (and associated compute) will be a bottleneck for automation globally.

8. Trade Skills, Exponential Economics, and the Future of Work

• Trade Boom: Electricians, Plumbers More In Demand Than Grads (87:48–88:32):
  • “Rather than coming out of school $100,000/$200,000 in debt, why not come out with a job that's paying $100,000 to $200,000?” – P.D.
  • Construction management and automation poised for further transformation.
• Memory, Storage, and the Price of Exponential Progress (90:18–92:09):
  • Comparison of RAM and storage prices in the 1980s vs. 2025 stuns the panel.
  • “We have massive abundance—overabundance of storage… But compute processing is still… the human brain is doing really, really well…” – D.B. (91:16)

9. AI and Energy: Solar, Coal, and Exponential Disbelief

• AI’s Hunger for Power (97:14):
  • OpenAI plans a 125-fold energy capacity increase in 8 years, still “a twentieth of a percent of the inbound insulation on Earth’s surface.” – A.W-G.
• Solar’s Unstoppable Exponentials (98:39–106:12):
  • Solar has hit nearly 3 terawatts of capacity, far surpassing every forecast. Yet “all the top energy experts… every time solar goes literally vertical, all the experts go linear.” – S.I. (101:46)
  • Caution against linear thinking by industry analysts is a recurring theme.
• Energy Geopolitics & Untapped Potential (107:07):
  • Untapped hydro and giant, underutilized dams, plus China’s solar manufacturing dominance, shape the global scene.
  • “Would be investing in building out solar capacity manufacturing here, right?” – P.D. (108:25)

10. Robotics & Autonomous Vehicles

• Robot Supply Chains, Shortages, and Policy (110:42–113:05):
  • US lags behind China in robotics component supply chains. Founders like Brett Adcock and Elon Musk forced to insource every part for humanoids.
  • “Manhattan-style project for supply chain for robots and drones.” – S.I. (112:31)
• Waymo & Self-Driving Impact (113:50–115:55):
  • If everyone drove like Waymo, “we’d prevent 33 to 39,000 deaths annually”; half of court cases in the US are from car accidents.
  • Self-driving will cascade into lower insurance, legal, and organ donation numbers—a systems-level disruption.

11. Longevity & Biotech: Adding Decades, Escape Velocity

• Retro Biosciences & XPRIZE Healthspan (117:10–118:23):
  • “Their mission is to add 10 healthy years on human lifespan. They're one of the teams competing for our $101 million XPRIZE healthspan… now over 730 teams have entered.” – P.D. (118:20)
• Chinese CRISPR Advances (119:47):
  • FOXO3 gene modification reduces aging by three to five years—global impacts expected.
• Longevity Escape Velocity Timelines (123:15):
  • “The consensus [from AI models]: 2030”—doubling (or more) of human lifespan may be possible by then (123:15).
  • “There is no upper limit to how long we can live.… breakthroughs required to understand why we age, how to slow it, stop it, reverse it, is going to fall at the knees of digital superintelligence.” – P.D. (122:08)

12. Notable Quotes & Memorable Moments

• “We're seeing algorithmic content generation take over from algorithmic selection—content creation is up for grabs.” – Alex (00:04; 09:52)
• “The interface to create [content] is entirely voice and prompt... all of that is about to disappear from the earth forever.” – Dave (10:16)
• “If the code can write itself really well, maybe that's the critical path to an intelligence explosion.” – Alex (18:38)
• “The AI is both the best ally you've ever had in buying things, but also, if it's misguided, could walk you down some seriously bad paths.” – Dave (32:46)
• “All their finance jobs essentially get rewritten now and recreated. It's a body blow to the accounting world.” – Salim (61:36)
• "[Quantum's] killer app... is probably something like AI-accelerated generalist training for AI." – Alex (66:25)
• “We may find ourselves in a world two to three years from now where we're just drowning under math, science, engineering being solved in rapid succession.” – Alex (59:41)
• “The demand is here now… those things didn’t exist six months ago. Now, anyone seeing those is going to want to do it immediately, whether it’s in a corporate or it’s personal… and so I think the demand is starting to catch up to the construction, and the demand will get way ahead of the [infrastructure].” – Dave (79:24)
• "There are so many ultra high value transformative problems that will set AIs on while we’re sleeping… I’m incredibly not worried that we’re going to drown in slop." – Alex (26:24)

13. Book Recommendations & Exponential Perspective

• Accelerando (Stross): “My favorite book ever… best depiction of what the 21st century is likely to look like… optimism, and physical realism.” – Alex (93:13)
• Nexus (Naom): “About neural interfaces, collective consciousness, and bioengineering—still fresh today.” – Peter (94:49)
• Foundation (Asimov): “A classic. Must-read for everybody.” – Salim (96:00)

Timestamps for Core Segments

• Sora 2 & Meta Vibes: 00:00–12:45
• AI Generated Audio (SUNO 5): 16:41–17:49
• Anthropic Sonnet 4.5 & Codegen: 18:18–22:05
• Red-Teaming, Power-Seeking AI: 22:05–24:06
• Imagine with Claude (App Creation): 24:06–29:04
• OpenAI Ads & Agentic Commerce: 29:19–34:44
• Stripe & Commerce Disruption: 37:24–41:16
• Benchmarks (GDP VAL, APEX): 43:49–48:19
• Math, Crypto, Quantum: 54:24–68:53
• Compute Moonshot, Data Center Arms Race: 71:18–80:11
• Intel as National Asset: 80:41–86:05
• Trades & Exponential Hardware Reality: 87:48–92:09
• Energy & Solar Hockeystick: 97:14–106:12
• Robotics, China’s Lead: 110:42–113:34
• Self-Driving Impact: 113:50–116:59
• Longevity Biotech, Escape Velocity: 117:10–123:38
• Book Recs: 93:13–96:41

Summary

This riveting “Moonshots” episode provides a panoramic, up-to-the-moment survey of AI’s impact across video, audio, commerce, coding, infrastructure, policy, and biology. The dialogue remains upbeat even as it recognizes the calamitous challenges of scale, trust, and exponential acceleration. Diamandis and his expert panel consistently return to the connective tissue of their ideas: exponential progress, the importance of benchmarks for superhuman performance, the democratization—and realignment—of power, and the moonshot mindset required to navigate (and shape) a future arriving faster each week.

Listen for:

• Visions of AI-powered sovereign economies
• Warnings about linear thinking—especially from experts
• Excitement (and anxiety) about the speed of, and scarcity in, robotics, chips, and energy
• The certainty that digital superintelligence will soon rewrite biology and economics alike

Whether you work in tech, finance, medicine, or government, this episode will challenge perceptions about what’s possible—and inevitable—in the decade ahead.