Google’s UK-based AI unit DeepMind has invented a tool that can write computer code at a competitive level to solve open-ended problems that require critical thinking, logic and understanding of language. The system, named AlphaCode, is a sign of the evolution of AI and would allow the automation of computer programming, a powerful and highly prized skill underpinning much of modern science and industry.
DeepMind tested AlphaCode’s performance in coding competitions hosted by Codeforces, a platform that pits the skills of tens of thousands of human software programmers around the world against one another. It found that it performed with the expertise of an average human programmer.
AlphaCode is not the first AI tool to generate computer code. In June 2020, Microsoft released a similar tool to aid programmers with the help of GitHub and OpenAI. The tool, called GitHub Copilot, works more as an intelligent assistant to experienced professionals by analysing existing code and generating new snippets or autocompleting lines of code rather than as an independent problem-solving entity.
AlphaCode currently cannot beat the best human programmers. Critics have also expressed concerns that code generated by machines must be scrutinised particularly closely for bugs, or unforeseen effects including vulnerability to cyberattacks. However, its ability to solve original problems suggests that AI could complement humans, improve productivity and broaden the horizon of people without coding backgrounds.