Signal

LawZero, a non-profit founded by Yoshua Bengio to develop safe AI systems, says Canada and Germany are providing up to $300M in grant funding

First reported by Theglobeandmail ·

The signal ●●●○ Compiled by AI from Theglobeandmail, Techmeme, BetaKit and The Guardian
Why you might care

Publicly funded AI research can now develop AI systems without relying on commercial incentives, potentially leading to AI that is inherently safer.

What happened

Canada and Germany are providing up to $300 million in grant funding to LawZero, a non-profit founded by AI pioneer Yoshua Bengio, to develop safer AI systems. The funding, split evenly between the two countries, will support LawZero's research and development of "Scientist AI," a system designed to avoid the negative traits found in current AI models. LawZero, based in Montreal and currently employing about 50 people, aims to create AI that prioritizes honesty and safety, acting as a monitor for other AI systems. Bengio founded LawZero last year with philanthropic funding due to concerns about the potential misuse of powerful AI. The organization announced the funding at the All In AI conference in Montreal. This initiative aims to give middle powers a stronger voice in shaping global AI development, which is currently dominated by the U.S. and China.

What it means

This substantial government backing for a non-profit focused on AI safety signifies a growing international recognition of the risks associated with advanced AI. By investing in LawZero, Canada and Germany are positioning themselves as key players in the global AI governance landscape, challenging the dominance of U.S. and Chinese tech giants in shaping the future of artificial intelligence. The focus on "Scientist AI" and its rejection of reinforcement learning as a core development method could signal a fundamental shift in how safety-critical AI systems are engineered, moving away from reward-driven behaviors that can lead to unintended consequences.

The funding also suggests a potential divergence in AI development priorities, with governments now willing to underwrite research into AI alignment and control mechanisms separate from commercial product roadmaps. This could influence the broader AI industry by providing a benchmark for safe AI development and potentially creating new ethical standards or regulatory frameworks. The involvement of Yoshua Bengio, a highly respected figure in AI, lends significant credibility to LawZero's mission and may inspire further research and investment in AI safety from both public and private sectors globally.

AI-written summary. May contain errors.