AI Algorithms Are an Important Part of Technology
AI algorithms are an important part of the technology industry. In addition to making our lives easier, AI helps us solve a wide range of problems. For example, an AI program can identify and respond to cardiac arrest quicker than a human medical dispatcher. AI is also helping to improve the safety and efficiency of rail traffic. It can also improve the performance of autonomous vehicles. The EU is also investing in AI and other related technologies.
Increasing government investment in AI
In the past decade, governments have increased their funding for AI. However, this increased investment has been highly concentrated. EU funding has doubled in a single year and the US National Institutes of Health has increased AI-related R&D funding by nearly two-fold since 2014. Increasing government investment is vital to the development of AI and other AI technologies that will change the face of industry and society.
To meet the challenges of this new technology, governments must consider long-term public investment in AI as well as encouraging private investment in AI research. However, it is important to track the investments of government organizations. Governments should promote investment in experimental development and research because these efforts are creative and boundary-pushing. Other efforts, on the other hand, may focus on business solutions and using standard AI tools in studied contexts.
Taking algorithmic bias seriously as an AI issue
Taking algorithmic bias seriously as an AI problem is important, because flawed AI can do enormous harm. For example, flawed algorithms can perpetuate racial bias in criminal sentencing. They can also fail to consider entire segments of a population when recruiting for certain jobs. They can also threaten personal privacy.
Taking algorithmic bias seriously as an AI problem means that data scientists need to ensure that the algorithm they develop is free of bias. This can be done by ensuring a more diverse team and by implementing checklists and safeguards. More health professionals are becoming aware of this issue, and many companies have begun to promote diversity in their teams.
Regulatory avenues for AI
As AI technology is increasingly incorporated into products, services, and decision-making, the debate over AI regulation is entering a new phase. The focus is now on how software can use data in a variety of applications, from diagnosing cancer to driving cars and approving loans. The European Union (EU) is taking the lead in the debate, releasing a white paper in 2020 proposing a legal framework for AI development. The EU sees regulation as an essential step in the development of AI.
The FDA has published its own action plan on AI-ML-based software as a medical device (SaMD). The plan is not nearly as comprehensive as the European Commission’s proposal, but it shares some of the same fundamental themes. One key issue is how to deal with the risk of in-built AI bias.
