Unlike hardware launches where users can keep their old device, forced software updates like OpenAI's GPT-4o replacing 4.0 take something away from users. This sunsetting aspect creates a sense of loss and resentment, especially for users who have formed a deep attachment to the previous version, violating typical launch expectations.
Related Insights
Despite access to state-of-the-art models, most ChatGPT users defaulted to older versions. The cognitive load of using a "model picker" and uncertainty about speed/quality trade-offs were bigger barriers than price. Automating this choice is key to driving mass adoption of advanced AI reasoning.
OpenAI's attempt to sunset GPT-4.0 faced significant pushback not just from power users, but from those using it for companionship. This revealed that deprecating AI models isn't a simple version update; it can feel like 'killing a friend' to a niche but vocal user base, forcing companies to reconsider their product lifecycle strategy for models with emergent personalities.
Unlike traditional APIs, LLMs are hard to abstract away. Users develop a preference for a specific model's 'personality' and performance (e.g., GPT-4 vs. 3.5), making it difficult for applications to swap out the underlying model without user notice and pushback.
While Over-the-Air (OTA) updates seem to make hardware software flexible, the initial OS version that enables those updates is unchangeable once flashed onto units at the factory. This creates an early, critical point of commitment for any features included in that first boot-up experience.
After facing backlash for over-promising on past releases, OpenAI has adopted a "low ball" communication strategy. The company intentionally underplayed the GPT-5.1 update to avoid being "crushed" by criticism when perceived improvements don't match the hype, letting positive user discoveries drive the narrative instead.
OpenAI's rapid reversal on sunsetting GPT-4.0 shows a vocal minority—users treating the AI as a companion—can impact a major company's product strategy. The threat of churn from this high-value, emotionally invested group proved more powerful than the desire to streamline the product.
OpenAI's platform strategy, which centralizes app distribution through ChatGPT, mirrors Apple's iOS model. This creates a 'walled garden' that could follow Cory Doctorow's 'inshittification' pattern: initially benefiting users, then locking them in, and finally exploiting them once they cannot easily leave the ecosystem.
In 2013, Google rolled out its significant 'Hummingbird' search algorithm update a full month before announcing it. No users complained because the experience simply improved. This 'ship then tell' strategy is a powerful playbook for consumer-facing AI products, proving an update's value through tangible benefits before users can form negative opinions based on an announcement.
Because AI products improve so rapidly, it's crucial to proactively bring lapsed users back. A user who tried the product a year ago has no idea how much better it is today. Marketing pushes around major version launches (e.g., v3.0) can create a step-change in weekly active users.
While new large language models boast superior performance on technical benchmarks, the practical impact on day-to-day PM productivity is hitting a point of diminishing returns. The leap from one version to the next doesn't unlock significantly new capabilities for common PM workflows.
