Tech CEO Warns: Why We Should Slow Down AI Adoption Now

Introduction

The healthcare industry stands at a crossroads with the integration of artificial intelligence (AI). While some professionals are eager to embrace this technology, others remain skeptical. Sarah M. Worthy, CEO of DoorSpace, shares her insights on the cautious yet strategic adoption of AI in healthcare.

Hesitation and Skepticism

“We’re all becoming less trustworthy of AI outcomes as we use it more and more,” says Worthy. The skepticism stems from the unreliable nature of large language models (LLMs) that can provide false information and contribute to societal biases. In life-or-death situations, such unpredictability is dangerous.

• Key Concerns:

– False information
– Societal biases
– Overworked clinicians

Overcoming Barriers

Healthcare leaders must demand ethical AI with robust safeguards. Worthy suggests focusing on non-patient-facing applications for now:

• Short-term Solutions:

– Automate operational tasks
– Manage workforce data efficiently
– Organize siloed data for better decision-making

“Bad data in, bad reports out,” she warns about unorganized data leading to negative ROI on AI investments.

Administrative Focus

Worthy envisions using AI to streamline administrative tasks rather than clinical ones:

• Reduce paperwork burden on physicians.
• Use data analytics to remove inefficiencies.
• Automate low-value tasks like data entry and report creation.

This approach aims to free up valuable time for healthcare professionals while maintaining high-quality care standards.

Future Clinical Applications

Worthy sees potential in specific clinical applications where AI augments rather than replaces human effort:

1. AI Scribing Tools:

– Record doctor-patient interactions.
– Improve EHR note accuracy without disrupting face-to-face consultations.

1. Radiology Enhancements:

– Faster scan evaluations.
– More accurate diagnoses through augmented tools.

She emphasizes that successful implementations share a common trait: they enhance clinician capabilities rather than replace them.

“I don’t foresee AI replacing doctors and nurses anytime soon without disastrous consequences.”

References

1. What LLMs Can Do For Radiologists And The Radiologist Shortage
2. John Halamka On AI’s Future And What It Can And Can’t Do In Healthcare
3. Ambient Voice GenAI Poised To Reduce Burnout Says Stanford Digital Health Director

For further details or inquiries:
Follow Bill Siwicki’s HIT coverage on LinkedIn: Bill Siwicki
Email him at: bsiwicki@himss.org

Note: The HIMSS AI in Healthcare Forum is scheduled for September 5–6 in Boston—