The Friday 15 episode examines whether AI is becoming an oligopoly, focusing on the emerging “protocol wars” between major tech companies and the implications for B2B e-commerce.
Andy and Brian discuss Apple’s $1 billion deal with Google for AI technology and analyze the competitive landscape of AI foundation models.
FAQ:
Q: What are “protocol wars” in the context of AI?
A: Protocol wars refer to the competition between different technical standards for enabling commerce through AI chatbots. OpenAI launched ACP (with Stripe) in September 2024, while Google announced Universal Commerce Protocol (UCP) with partners like Shopify, Target, Walmart, and Etsy at NRF 2025. These protocols enable shopping and transactions directly through AI interfaces like ChatGPT and Google Gemini.
Q: Why did Apple pay Google $1 billion for AI technology?
A: According to the hosts, Apple Intelligence (released in 2024) has been considered a failure. Rather than continuing to develop their own AI technology, Apple chose to license Google’s Gemini AI to power features like Siri. This represents a significant strategic shift and an admission that building competitive AI requires massive resources.
Q: How much are companies investing in AI infrastructure?
A: The eight major players in the AI space are collectively spending approximately $3 trillion on infrastructure, people, and resources. This investment is staggering compared to Google’s entire search business, which is valued at around $300 billion. The massive capital requirements create significant barriers to entry.
Q: Who are the major players in the AI oligopoly?
A: The episode identifies eight primary companies: Google (Gemini), OpenAI (partnered with Microsoft), X/Grok (Elon Musk), DeepSeek (Chinese company), Claude (Anthropic), Perplexity, Amazon (Titan), and Meta (Llama). Amazon and Google are also investing in multiple companies beyond their own AI models.
Q: What is the current market share distribution in AI chatbots?
A: ChatGPT dominates with 68% of the AI chatbot market share (down from 85%), while Google Gemini has grown to 18%. Other players including Perplexity, Claude, and others make up the remaining market share. However, in traditional search, Google still commands 78% while ChatGPT has captured 17% of search queries.
Q: What does “oligopoly” mean in this context?
A: According to the Federal Reserve Bank of St. Louis definition cited in the episode, an oligopoly exists when five firms in an industry control more than 60% of total market sales. This is compared to a monopoly (one company with total control) and a duopoly (two companies, like FedEx and UPS). The AI market already shows oligopolistic characteristics.
Q: How does this affect B2B e-commerce companies?
A: B2B companies need to understand where their customers are searching and how they’re finding products. As search behavior shifts from traditional Google to AI chatbots, companies must optimize for AEO (AI Engine Optimization), decide which protocols to support, and potentially prepare for AI interfaces to become the primary customer experience—similar to how Amazon disrupted product search.
Q: Do B2B companies have any advantages in this AI transition?
A: Yes. Brian Beck notes that many B2B companies have invested heavily in authoritative content about technical applications, use cases, and complex product information. This depth of expertise and content may work to their advantage as AI models prioritize authoritative, comprehensive information sources.
Q: When will AI-powered commerce affect B2B significantly?
A: While B2C companies are racing to adopt these protocols immediately (for consumer purchases through chatbots), B2B adoption is still emerging. However, the hosts emphasize this is “over the horizon” rather than far-off—B2B companies need to pay attention now to avoid being left behind, just as many were slow to adopt Amazon as a sales channel.
Q: How many foundation AI models do the hosts predict will survive?
A: Brian Beck predicts fundamentally three dominant models with excess capacity resold through other providers (similar to the telecom industry structure). Andy Hoar agrees it will be limited—perhaps four to five—due to the enormous infrastructure investment required. A community poll showed opinions split roughly equally between 2-3, 4-5, and 6+ models surviving.
Q: Will AI become regulated like telecommunications?
A: The hosts suggest that if only three major foundation models emerge, some level of regulation is likely. The comparison is made to historical utilities like telecommunications and railroads, where massive infrastructure requirements led to natural oligopolies that eventually faced regulatory oversight.
Q: What should B2B leaders do now?
A: Key actions include: (1) Monitor where customers are searching and how behavior is shifting, (2) Understand and potentially prepare to support the emerging commerce protocols (ACP and/or UCP), (3) Continue investing in authoritative, comprehensive content, (4) Optimize for AI Engine Optimization (AEO), and (5) Consider how AI interfaces might eventually displace traditional e-commerce experiences.
Q: Is this just a temporary competitive phase or a permanent shift?
A: The hosts draw parallels to other infrastructure-heavy industries (telecom, railroads, airlines, film studios) where high barriers to entry led to permanent oligopolistic structures. The $3 trillion collective investment and the technical requirements suggest this will result in a lasting oligopoly rather than a temporary competitive phase.
