“It seemingly unleashed 10 years of efficiencies overnight,” says Joseph Hoefer, vice president of government relations and public affairs at Monument Advocacy. He explains that a task “that would normally take two or three hours to do, with the right prompt and the right data, you can do in 45 seconds.”
For example, “the amount of bias with the data sets is problematic,” Hoefer says. AI algorithms are trained on data from existing publications that are almost always biased, and developers are still devising ways to remove bias from AI outputs. If lobbyists aren’t careful, the biases in AI algorithms can impact the quality and accuracy of their analyses.
AI’s summarizing capabilities can be particularly helpful when it comes to comparing multiple bills. “We would view AI tools as being particularly helpful when it comes to analyzing similar bills focused on a specific issue and their key differences, key similarities, and communities of people that tend to sponsor bills together,” says Andrew Howell, partner at Monument Advocacy.
Read the full story here.