Artificial intelligence has emerged as a powerful tool in the realm of healthcare and medicine, and even the treatment of cancer. However, recent studies demonstrate that while AI holds immense potential, it also carries inherent risks that must be carefully navigated. One startup has used AI to target cancer treatments. Let’s take a closer look at the developments.
- UK’s Etcembly uses generative AI to create potent immunotherapy, ETC-101, a milestone for AI in drug development.
- A JAMA Oncology study exposes risks in AI-generated cancer treatment plans, showcasing errors and inconsistencies in ChatGPT’s recommendations.
- Despite AI’s potential, misinformation concerns arise. 12.5% of ChatGPT’s suggestions were fabricated. Patients should consult human professionals for reliable medical advice. Rigorous validation remains crucial for safe AI healthcare implementation.
Can AI Cure Cancer?
In a groundbreaking breakthrough, UK-based biotech startup Etcembly has harnessed generative AI to design an innovative immunotherapy, ETC-101. This immunotherpy targets challenging-to-treat cancers. Additionally, the achievement marks a significant milestone as it is the first time AI has developed an immunotherapy candidate. Etcembly’s creation process. As such, this showcases the AI’s ability to accelerate drug development, delivering a bispecific T cell engager that is both highly targeted and potent.
However, despite these successes, we must proceed with caution, as AI applications in healthcare require rigorous validation. A study published in JAMA Oncology emphasizes the limitations and risks associated with relying solely on AI-generated cancer treatment plans. The study assessed ChatGPT, an AI language model, and revealed that its treatment recommendations contained factual errors and also inconsistencies.
Facts Mixed with Fiction
The Brigham and Women’s Hospital researchers discovered that, out of 104 queries, approximately one-third of ChatGPT’s responses contained incorrect information. While the model included accurate guidelines in 98% of cases, these were often interwoven with inaccurate details. This therefore makes it challenging even for specialists to spot errors. The study also found that 12.5% of ChatGPT’s treatment recommendations were entirely fabricated or hallucinated. So, this raises concerns about its reliability, particularly in advanced cancer cases and the use of immunotherapy drugs.
OpenAI, the organization behind ChatGPT, explicitly states that the model is not intended to provide medical advice for serious health conditions. Nevertheless, its confident yet erroneous responses underscore the importance of thorough validation before deploying AI in clinical settings.
While AI-powered tools offer a promising avenue for rapid medical advancements, the dangers of misinformation are evident. Patients are advised to be wary of any medical advice from AI. Patients should always reach out to human professionals. As AI’s role in healthcare evolves, it becomes imperative to strike a delicate balance between harnessing its potential and ensuring patient safety through rigorous validation processes.