The author predicts that local AI apps, which operate with the same permissions as the user, will be the future of AI automation. This approach bypasses the traditional permission system of the internet, allowing AI to function more freely. The article concludes by stating that while this shift won't happen overnight, it's already in progress and will significantly shape the future of AI. The author also suggests that a more decentralized permission system, similar to what exists in crypto, could emerge in the future.
Key takeaways:
- The concept of 'having permission' is defining the shape of AI, with companies like Apple and OpenAI building AI agents that require user permission to access and utilize data.
- AI can potentially automate nearly anything we desire, but it requires access to everything, which means it needs permission. However, not all companies are willing to grant this permission due to various reasons.
- Companies like Apple, Microsoft, and Google are leveraging their power to convince app developers to give their AI permission, while others like OpenAI are skipping permission altogether and building AI models that can use websites and apps built for humans.
- Local AI apps are a way to bypass the legacy permission system of the internet and set AI automation free, and this trend is already happening and will shape the future of AI.