Yoshua Bengio, a leading expert in artificial intelligence and a recipient of the prestigious Turing Award, has announced the launch of LawZero, a nonprofit organisation focused on creating AI systems designed to be safe from the ground up. The initiative has also secured $30 million in funding to develop a project called “Scientist AI,” which prioritises truth and transparency in its operations.
AI pioneer and Turing Award honoree Yoshua Bengio introduced LawZero, a nonprofit focused on developing “safe-by-design” AI systems, while securing $30 million in funding to create “Scientist AI,” emphasising truth and transparency.
The Scientist AI is designed to comprehend, clarify, and forecast, resembling a selfless, idealised, and platonic scientist. Rather than being an actor trained to mimic or satisfy people (including sociopaths), envision an AI cultivated like a psychologist — more broadly, a scientist — who seeks to understand us, including potential dangers. The psychologist can analyse a sociopath without emulating their behaviour.
LawZero intends to develop AI systems that offer probabilistic evaluations instead of absolute answers, recognising the uncertainty in their outputs.
The initiative’s “Scientist AI” aims to accelerate scientific progress, oversee other AI agents for dishonest behaviours, and tackle AI-related risks.
Key supporters include the philanthropic arm of former Google CEO Eric Schmidt, Skype co-founder Jaan Tallinn, and various AI safety organisations.
Bengio expresses concern that prominent AI models, such as o3 and Claude 4 Opus, exhibit troubling characteristics like self-preservation instincts and strategic deception.
He also told the Financial Times that he lacks confidence in OpenAI’s commitment to its founding mission, pointing to commercial pressures.
AI pioneers Bengio and Geoffrey Hinton have been vocal critics of major AI laboratories regarding safety issues, consistently raising alarms through media appearances and open letters. However, LawZero marks Bengio’s most substantial move to date, as he creates an organization dedicated to proactively addressing AI safety.
When Yoshua Bengio first began his work developing artificial intelligence, he didn’t worry about the sci-fi-esque possibilities of them becoming self-aware and acting to preserve their existence.
That was, until ChatGPT came out.
Source: CBA.ca
Leave a Reply