X Users Obliged Elon Musk’s Ask To Upload Medical Information For His AI Company To Review
X owner and incoming White House cabinet member Elon Musk asked users to upload MRIs, CT scans and other medical information for his artificial intelligence chatbot, Grok to review and some fell for it, Fortune reports.
Musk pitched the idea on X in late October 2024, admitting that Grok is still in early stages, but wanted to see what the response would be. Once images poured in, he would see if the interpretation was spot on. “Try submitting x-ray, PET, MRI or other medical images to Grok for analysis. This is still early stage, but it is already quite accurate and will become extremely good,” he wrote.
“Let us know where Grok gets it right or needs work.”
Some people that issued images celebrated Grok being “spot on” with blood test results and breast cancer detection but others waved red flags against the platform. Josh Sharp, who goes by @showinvestment on social media, pointed out how a broken clavicle was looked at as a dislocated shoulder. “I submitted my Xray to Grok and it said I had a dislocated shoulder. Pretty obvious broken clavicle,” he said.
“Grok still has some work to do…”
Radiologist, Dr. TJ, provided an in depth analysis on an MRI image that he labeled as “too generic” and not a real diagnosis. “Hello Elon, Radiologist here. Just tried with the last MRI image I had (adhesive capsulitis of the shoulder, level: easy, typical pattern). Heading + 1: true; 2: not true; 3: true but too generic; 4-5-6: not true,” he wrote.
“Conclusion: too generic + no diagnosis. Will try again next time!”
Another example is the bot mistaking a mammogram of a benign breast cyst for an image of testicles.
The platform was launched in May 2024 after, through Musk’s tech startup, xAI, raised $6 billion in an investment funding round. That large amount of funding gave the billionaire an abundance of capital to invest in health care technologies but it has yet to be determined how far it will go to address other medical needs. Grok is not the first of its kind as Google’s Gemini or OpenAI’s ChatGPT also allows people to submit medical images.
But while some praise the potential technology advancement, medical privacy experts are sounding the alarm on what artificial intelligence will do with the information. “This is very personal information, and you don’t exactly know what Grok is going to do with it,” Vanderbilt University’s professor of biomedical informatics, Dr. Bradley Malin said, according to the New York Times.
“Posting personal information to Grok is more like, ‘Wheee! Let’s throw this data out there, and hope the company is going to do what I want them to do.”
The Health Insurance Portability and Accountability Act better known as HIPAA protects medical information when shared with doctors or on a patient portal, as federal guidelines protect it from being shared without consent. But protections do not reach social media sites – only for doctors’ offices, hospitals, health insurers and some companies they work with.
RELATED CONTENT: Former St. Louis Prosecutor Admits To Using Public Funds To Cover Personal Court Fees