In a recent LinkedIn post, Arun P. delves into the profound ethical implications of advanced artificial intelligence, particularly focusing on the potential risks when AI gains the ability to directly control physical human actions. The post, sparked by news of MIT students developing an AI capable of moving a human hand via commands and muscle stimulation, prompts a critical examination of AI safety and oversight.
Arun P. highlights the groundbreaking nature of the MIT project, acknowledging its engineering prowess. However, he quickly pivots to the inherent dangers, questioning the consequences of AI misinterpretation in a physical context.
“We just gave an AI access to move a human body. Now imagine it misreads the scene. Imagine it fires the wrong motion. Imagine that power in the wrong hands, with no one watching.”
The core of Arun P.’s concern lies in the disparity between AI errors in the digital and physical realms. He emphasizes that while text-based AI mistakes might result in a “bad sentence,” errors in AI controlling physical actions carry far graver, immediate consequences.
The Perils of Physical AI Control
Arun P. argues that the current development trajectory of AI often prioritizes granting more access, control, and autonomy before establishing robust safety mechanisms. This approach, he suggests, is fundamentally flawed when dealing with AI that can interact with the physical world.
“This is the part we keep skipping. We race to give AI more access, more control, more autonomy, and we build the monitoring afterward, if at all.”
The post posits that essential safety features like monitoring and governance should not be an afterthought. Instead, they must be integral to the AI development process from the outset, especially for systems with the potential to influence or control physical actions. Arun P. stresses the need for proactive implementation of these safeguards.
The Critical Question of AI Oversight
A central theme in Arun P.’s analysis is the challenge of accountability and oversight. As AI systems become more sophisticated and integrated into physical control systems, the question of who monitors the AI becomes paramount.
“It has to be the other way around. AI monitoring and AI governance can’t be the cleanup crew. They have to be in place before we hand AI anything that matters.”
Arun P. frames this not just as a technical hurdle but as an ethical imperative. The development of AI capable of direct physical intervention necessitates a parallel, and preceding, development of robust governance frameworks. Without such measures, the potential for unintended harm or malicious misuse escalates significantly.
Introducing Block Convey
To address these critical concerns, Arun P. reveals that he is building a venture, Block Convey, specifically to tackle the challenge of AI monitoring and governance. This initiative aims to provide solutions for ensuring that AI systems, particularly those with physical control capabilities, operate within safe and ethical boundaries.
Arun P. concludes by inviting discussion, underscoring the collective responsibility in navigating the complex landscape of advanced AI. The post serves as a vital reminder for technologists, policymakers, and the public to prioritize safety and ethical considerations as AI capabilities continue to expand into physical domains.
📝 About This Content
This article is based on insights shared by Arun P. on LinkedIn.
📅 Originally posted on June 23, 2026 | View original post on LinkedIn →