Meta, the company formerly known as Facebook, is aiming to build artificial intelligence (AI) that is smarter than humans, and has established a lab dedicated to achieving this goal through self-improving AI systems.
While Meta’s CEO Mark Zuckerberg acknowledges that such systems pose a fundamental risk and could rapidly improve technologies such as hacking and weapon design, some experts believe focusing on automating AI research is the fastest path to achieving powerful AI.
However, others say human ingenuity remains the primary engine of AI advancement and that automating AI research and development could have enormous benefits, such as more sophisticated cancer treatments and investigations into climate change.
Five ways AI is making itself better include enhancing productivity through coding assistants, optimising infrastructure, automating training, perfecting agent design and advancing research.