In episode 114 of the AI/Hyperautomation Minute, Aaron Back reviews a recent press release from VERSES AI about its breakthrough in explainable AI research.
Highlights
01:16 — After reviewing the press release and VERSES AI’s research on explainable AI, Aaron noted a few key points:
- Fully transparent explainable AI is not here yet
- Human influence is still a major factor
- There’s still a lot of work to do
01:32 — This surfaces critical points that are outlined in the Acceleration Economy guidebook “The Ethical & Workforce Impacts of Generative AI.” In the guidebook, Aaron posited that explainable AI should provide visibility for both technology teams and business decision-makers.
01:57 — For technology teams, explainable AI should provide visibility into data sources, data usage, data influence, how the AI model can be improved, and AI data security.
02:11 — For business decision-makers, explainable AI should provide visibility into competitive AI opportunities, AI data compliance, AI upskilling opportunities, and AI security.
02:34 — The VERSES AI research paper raised the topic of audits and the complexities of really understanding the AI output. Could it pass a stringent audit, especially in the areas of compliance or for specific industries?
