How To Prepare Faster For Looming AI Regulation: Turn Defense Into Offense
AI is powerful, and using it comes with great responsibility. My recent article in Forbes shares five ways AI compliance also forms a more effective AI culture.
AI is powerful, and using it comes with great responsibility. My recent article in Forbes shares five ways AI compliance also forms a more effective AI culture:
PUT HUMANS IN THE LOOP. The first and most important player in the AI game is the human. Since humans create algorithms and are biased, AI inherits that bias. So make AI a team sport with data scientists and analysts working together. Teamwork helps reduce risk and bias in AI and helps business teams use it more effectively.
HIRE HYBRID SOCIAL SCIENTISTS. Don't over-rotate your data science team with data scientists. Instead, add hybrid social scientists steeped in the domain with a background in philosophy, history, psychology, or linguistics and basic literacy in data, data storytelling, and data science. They provide an essential counterweight to technical teams.
BAKE DECISION OBSERVERS INTO YOUR AI LIFECYCLE. Employ what Nobel Prize-winning psychologist @Daniel Kahneman calls "Decision Observers" as you create, assess, screen, deploy and evaluate algorithms. Decision Observers are external observers and collaborators trained to understand risk and bias.
DON'T FORGET THE DATA (ENGINEERS). Don't delegate data engineering to data scientists. Data engineers ensure AI uses the correct, clean, compliant data. So bake them into your data science teams, too.
EMPLOY AI ANALYTICS TO MONITOR, MANAGE, AND REFINE AI-BASED DECISION MAKING. AI analytics help teams monitor algorithms, the recommendations they make, and the impact of decisions made based on their suggestions. Add business analysts to AI teams to ensure visual analytics and dashboards ensure transparency into AI's use.
Collaborative AI cultures safely seize AI's power. So, put humans in the loop; hire hybrid social scientists, bake the habit of governance into your AI lifecycle, don't forget the data engineers, and employ AI analytics.
Questions? Comments? Ask them here on LinkedIn.
