The Gist
- AI still trails human expertise. Scale AI is testing against human-level benchmarks—and models aren’t close yet.
- Collaboration beats replacement. AI may enhance performance in CX roles rather than eliminate them.
- Customer satisfaction rises with hybrid models. When AI supports—not replaces—agents, satisfaction scores improve.
Scale AI is working to bring artificial intelligence up to human-level proficiency by having deeply skilled humans teach it what they know. The company’s contributor network is made up of 12% PhDs and 40%+ masters and professional degree holders who are training the AI in their areas of expertise.
AI Isn't Replacing Humans Anytime Soon
Scale AI wants you to know that your job is probably not at risk from artificial intelligence. Field Chief Technology Officer Vijay Karunamurthy emphasizes that modern-day AI isn’t measuring up to the benchmarks set for human-level performance. Even if AI approaches that level, it will likely complement human workers rather than replace them.
“Even the most powerful AI models still lack the creativity, emotional intelligence and nuanced decision-making that people bring to their work,” Karunamurthy stated. In areas like finance, AI struggles to handle high-stress situations effectively, showing a significant gap between humans and machines so far.
When AI Supports, Performance Improves
Karunamurthy argues that AI can often only outperform humans in specific tasks. When humans work alongside AI, they can achieve better outcomes than when working alone. For instance, in medicine, a doctor or nurse can leverage AI to synthesize a patient’s medical context, leading to enhanced interactions.
Customer Experience Will Still Need Humans — for now
While AI capabilities are improving rapidly, there are still areas like customer service where average employees could be at risk of replacement. If AI can provide immediate answers for customer support, satisfaction scores could dramatically increase. However, the transition may not be painless for everyone. Scale AI's collaboration with the Center for AI Safety revealed that top models can only answer about 20% of complex questions, indicating a wide gap between AI and expert knowledge. The rapid advancement of technology means that predictions about labor implications are challenging, but significant changes are inevitable.
Comments
Join Our Community
Sign up to share your thoughts, engage with others, and become part of our growing community.
No comments yet
Be the first to share your thoughts and start the conversation!