Key insights
- Mark Zuckerberg addressed concerns about Meta using employee computer activity to train AI models. He emphasized that the data is anonymized and not used for surveillance, but solely for AI training. This initiative highlights Meta's focus on AI development and its potential competitive advantage. The market impact is slightly positive, reflecting potential long-term gains from AI advancements.
Investing.com -- Meta Platforms Inc (NASDAQ: META) CEO Mark Zuckerberg addressed employee concerns about the company tracking device activity to train artificial intelligence systems, according to audio leaked by More Perfect Union and posted on the group's social media.
During an recent internal meeting, an employee questioned Zuckerberg about employee device tracking following recent announcements. The CEO explained that Meta is training AI models by having them observe employees using computers and performing tasks.
Zuckerberg said the company's AI development relies on three components: research and architecture, infrastructure, and data quality. He stated that AI models learn by watching people perform tasks, and Meta employees represent a higher quality data source than typical contract workers used elsewhere in the industry.
"If we're trying to teach the models coding, for example, then having people internally build tools or solve tasks that help teach the model how to code, we think is going to dramatically increase our model's coding ability," Zuckerberg said.
The CEO emphasized that no humans are reviewing individual employee activity, and content is stripped out where possible. He said the data is not being used for surveillance or performance tracking, but solely to train AI systems.
Zuckerberg acknowledged the company could have communicated the initiative better. He said Meta faces a tension between transparent communication and protecting strategically sensitive information from competitors in what he described as one of the most competitive fields in history.
The CEO indicated similar AI training initiatives using employee data may continue if this approach proves effective.