Kids’ robot maker shuts down, leaving AI toys lifeless

The closure of AI company Embodied highlights the risks and emotional impact of cloud-dependent AI products, particularly those designed for vulnerable users like children with autism.

The shutdown situation: Embodied, creator of the AI-powered social robot Moxie, announced its closure following financial difficulties and withdrawn funding.

• The Moxie robot, priced at $799, was specifically designed to interact with autistic children
• The device relied on cloud-based large language models (LLMs) for its core functionality, including conversation and question-answering capabilities
• Within days of the company’s closure, all Moxie units will cease functioning, with no refunds offered to customers

Technical implications: The dependence on cloud-based AI services reveals fundamental vulnerabilities in connected smart devices.

• Cloud-based devices require continuous server maintenance and expensive LLM processing to function
• The situation mirrors previous failures, such as the Vector robot company shutdown
• No clear succession plan exists for maintaining or transitioning the service to new operators

Human impact: The shutdown has created significant emotional distress for users and their families.

• Parents must explain to their children why their robot companion will stop working
• Social media platforms show videos of children and adults expressing grief over losing their Moxie companions
• The company provided a “letter from the Global Robotics Lab” to help parents discuss the situation with children

Ethical concerns: The situation raises questions about the responsibility of AI companies and the wisdom of using AI for child development.

• Using LLMs for child socialization, especially for neuroatypical children, presents significant ethical concerns
• The high operational costs of AI services expose the fragility of current business models in the sector
• The emotional attachment users form with AI devices creates vulnerability when companies fail

Silicon Valley reality check: The inability of many AI products to generate sustainable revenue while covering high operational costs exposes fundamental flaws in current AI business models.

• LLM-based products remain expensive to operate and maintain
• The industry’s approach of prioritizing growth over sustainable business models creates risks for consumers
• The potential for emotional harm when AI services shut down has been underestimated by companies

Looking ahead: While some users hope for community solutions like local operation hacks, the Moxie situation serves as a cautionary tale about the risks of deploying AI solutions for vulnerable populations without ensuring long-term sustainability and considering the emotional impact of potential failure.