Scientific Community Debates AI Funding and CO Poisoning

This collection of letters to the editor highlights pressing discussions within the scientific community, from the reliability of AI in research to innovative funding strategies and advancements in medical treatments.

The Pitfalls of ChatGPT in Scientific Research

A recent article on ChatGPT's tendency to overlook retractions in scientific papers prompted a strong response. Reader Peet-Mati Sööt from Salem, Oregon, expresses that it's unsurprising and bothersome that any reputable researcher would use ChatGPT for technical research. He argues that while generative AI may be suitable for creative tasks, it cannot be trusted for scientific analysis.

Recalling a personal experience from several years ago, he found the popular AI site produced incorrect answers, was outdated, and failed to cite its sources. Sööt suggests that the focus should be on promoting more reliable AI tools. He recommends perplexity.ai as a superior alternative, noting its thoroughness and consistent use of citations, especially in its pro version. The message is clear: scientists and engineers should be guided toward better tools rather than continuing to test the flawed capabilities of ChatGPT for serious research.

• Peet-Mati Sööt, Salem, Oregon

A New Funding Model for the American Chemical Society

In light of the political climate under the Donald J. Trump administration, Martha Dibblee of Portland, Oregon, proposes a new financial strategy for the American Chemical Society (ACS). She suggests that the ACS should actively seek significant donations from private investors and even consider becoming a publicly traded entity on the New York Stock Exchange or a similar platform.

The goal of this approach would be to create a shared security that could bolster the society's bottom line. This influx of income could then be used to fund important research projects that have recently faced budget cuts, ensuring the continuity of scientific advancement.

• Martha Dibblee, Portland, Oregon

An Overlooked Alternative for Carbon Monoxide Poisoning Treatment

Responding to the article “This Engineered Protein May Save Lives,” which discussed a new intravenous protein treatment for carbon monoxide (CO) poisoning, Joseph A. Fisher from Toronto points out a significant omission. The article claimed that oxygen therapy is the only existing treatment for CO poisoning, which he states is incorrect.

Fisher highlights his own work, easily found via a Google search, which includes the paper “Rapid Elimination of CO through the Lungs.” He developed a completely noninvasive method for treating CO poisoning that is as effective as the current gold standard, hyperbaric oxygen. The treatment uses a small, portable pneumatic device that helps a patient increase ventilation, accelerating the elimination of CO from the body while keeping blood carbon dioxide levels stable.

This technology, developed by Thornhill Medical (a company he cofounded), is approved by the US Food and Drug Administration and is now in the public domain. Fisher emphasizes that when considering a new invasive treatment like a synthetic protein injection, readers should be aware that highly effective, noninvasive alternatives already exist.

• Joseph A. Fisher, Toronto