Ghost has unveiled Core, a $3,499 personal computer built to host autonomous AI agents. the device focuses on local data processing to ensure user privacy by avoiding cloud-based storage.
An NVIDIA RTX PRO 4000 powered local hub
Ghost's Core is a specialized hardware solution designed to centralize and analyze user data through continuous AI interaction. According to the report, the machine is anchored by an NVIDIA RTX PRO 4000 GPU, providing the necessary computational muscle to handle complex tasks without relying on external servers. this hardware configuration is intended to support a system that learns from user behavior over time, effectively creating a personalized digital assistant that lives on the desk rather than in a data center.
The move toward high-end, dedicated AI hardware reflects a growing tension in the tech industry between the convenience of cloud-based LLMs and the security of edge computing. By providing a high-finish physical device, Ghost is positioning itself alongside professional workstation manufacturers rather than simple software providers. This approach seeks to turn the AI experience into a tangible, owned asset rather than a subscription-based service.
Pre-installed Qwen 3.8-Next and Gemma 4-31B models
The Core system arrives with several vision-language models already integrated into its software stack. As Ghost announced, the device includes Qwen 3.8-Next and Gemma 4-31B to facilitate immediate task automation and context building. these models allow the computer to "see" and interpret data locally, which is a critical component of the machine's promise to maintain user privacy.
Beyond the factory settings, the device offers significant flexibility for developers and AI enthusiasts. Users have the ability to pull additional models directly from Hugging Face, ensuring the hardware does not become obsolete as new open-source architectures emerge. This integration suggests that Ghost is targeting a sophisticated user base that values both immediate utility and long-term extensibility.
A privacy-first approach to continuous AI learning
Ghost is attempting to replicate the software lifecycle of the automotive industry by offering over-the-air updates for its AI models. The company plans to push these updates to the Core system to refine reasoning capabilities and introduce new features, much like how modern electric vehicles receive software improvements. This ensures that the $3,499 investment can evolve alongside the rapidly advancing field of artificial intelligence.
To safeguard the data being processed, the Core includes a built-in firewall designed to monitor network activity and block unauthorized information sharing. The source notes that users retain full ownership of both their raw data and the specific reasoning capabilities developed by the AI. This architecture is designed to give users granular control over which applications and external devices the AI is permitted to interact with.
Can a built-in firewall protect $3,499 of data?
While the privacy features are robust on paper, several critical questions remain regarding the practical implementation of the Core system.. It is currently unclear how much power the device will consume during its promised "continuous learning" cycles, which could lead to significant electricity costs for heavy users. Additionally, the report does not specify how Ghost will verify that the built-in firewall can effectively prevent sophisticated data leaks in an era of increasingly complex cyber threats. finally, it remains to be seen whether the $3,499 price point is low enough to attract anyone beyond high-net-worth privacy enthusiasts and specialized developers.
Comments 0