As AI model performance converges, the key differentiator will become memory. The accumulated context and personal data a model has on a user creates a high switching cost, making it too painful to move to a competitor even for temporarily superior features.

Contrary to assumptions about user stickiness, consumers of AI models will quickly switch to a better-performing or cheaper alternative. The 22% drop in ChatGPT usage after new Gemini models were released demonstrates that brand loyalty is low when model performance is the key value proposition.

The assumption that enterprise API spending on AI models creates a strong moat is flawed. In reality, businesses can and will easily switch between providers like OpenAI, Google, and Anthropic. This makes the market a commodity battleground where cost and on-par performance, not loyalty, will determine the winners.

Unlike social networks where user-generated content creates strong lock-in, AI chatbots have a fragile hold on users. A user switching from ChatGPT to Gemini experienced no loss from features like personalization or memory. Since the "content" is AI-generated, a competitor with a superior model can immediately offer a better product, suggesting a duopoly is more likely than a monopoly.

Today's LLM memory functions are superficial, recalling basic facts like a user's car model but failing to develop a unique personality. This makes switching between models like ChatGPT and Gemini easy, as there is no deep, personalized connection that creates lock-in. True retention will come from personality, not just facts.

The most advanced AI users are 'polyamorous' with models, using an average of 3.5 different tools. This indicates a mature usage pattern where users select the best model for a specific job rather than relying on a single, all-purpose AI, challenging the 'winner-take-all' market theory.

The LLM assistance space is trending towards "winner-take-most" not just due to quality, but because of user inertia. The vast majority of ChatGPT users are not multi-homing or even exploring alternatives like Gemini, indicating a strong default behavior has been established.

Despite ChatGPT building features like Memory and Custom Instructions to create lock-in, users are switching to competitors like Gemini and not missing them. This suggests the consumer AI market is more fragile and less of a winner-take-all monopoly than previously believed, as switching costs are currently very low.

The perceived competitive advantage of a chatbot's memory is an illusion. Users can simply ask the AI to output its entire conversation history and then paste that data into a rival service, effectively transferring the 'memory' and eliminating switching costs.