Andrew Ng, the founder and former leader of Google Brain, supports Google’s recent decision to remove its pledge not to build an AI system for weapons.
“We are extremely pleased that Google has changed its stance,” NG said in a stage-on-stage interview with TechCrunch at the Military Veteran Startup Conference in San Francisco.
Earlier this week, Google removed its seven-year-old pledge from its AI Principles webpage. In addition to deleting, Google has published a blog post written by Deepmind CEO Demis Hassabis. DemisHassabis said businesses and governments should work together to build AIs that “support national security.”
Google made the AI weapons pledge in 2018 after the Project Maven protest. The protest saw thousands of employees protest against the company’s contract with the US military. The protesters were particularly problematic with Google providing AI for military programs that could be used to help interpret video images and improve the accuracy of drone strikes.
However, Ng was confused by the Project Maven protesters, but he told an audience consisting primarily of veterans.
“Frankly, when Maven’s project falls… a lot of you are out of the air, so many of you are willing to shed blood for our country to protect our country.” said Ng. “So how can American companies refuse to help our own services people who are fighting for us?”
NG didn’t work with Google when the Project Maven protest occurred, but it played a key role in shaping Google’s efforts on AI and neural networks. Today, NG leads AI-focused venture studios and AI funds, and frequently speaks about AI policies.
Ng later said he was grateful that two AI regulatory efforts – the rejected California SB 1047 bill and Biden’s overturned AI executive order are no longer working. He repeatedly argued that both measures would slow America’s open source AI development.
According to Ng, the real key to American AI security is to allow the US to technically compete with China. He said that AI drones “will revolutionize the battlefield completely.”
It’s not just the former Google executive who spread the message. Former Google CEO Eric Schmidt is currently spending his time lobbying in Washington, DC, and is competing with China by purchasing AI drones. His company, White Stork, may eventually supply those drones.
Although Ng and Schmidt seem to support military AI use, this topic has split ranks within Google for years.
Meredith Whitaker, now president of Signal, led the Maven protest in 2018 while working as an AI researcher at Google. When Google pledged not to renew its Project Maven contract, Whittaker said she was happy with the decision, saying the company “should not be involved in the business of war.”
She’s not the only Googler to object. Former Google AI researcher and Nobel Prize winner Geoffrey Hinton had previously called for the World Government to ban and regulate the use of AI in weapons. Another long-standingly respected Google executive, Jeff Dean (now a deep-supervised chief scientist), has previously signed a letter opposed to the use of machine learning in autonomous weapons.
In recent years, Google and Amazon have fallen into new scrutiny of military activities, including a project Nimbus contract with the Israeli government. Employees from both cloud providers staged Sit-ins last year to protest Project Nimbus.
The Pentagon and military around the world have new appetites to use AI, the Pentagon chief executive told TechCrunch earlier. With Google, Amazon, Microsoft and other tech giants investing hundreds of billions of dollars in AI infrastructure, many are trying to bounce back their investments through military partnerships.