Tuesday, December 24, 2024

Elon Musk suggests Grok AI has a role in healthcare

Share

Last week, billionaire tech mogul Elon Musk took to his social media platform X to ask users of his artificial intelligence chatbot Grok to upload their medical photos and make a diagnosis.

“Try sending X-rays, PET, MRI or other medical images to Grok for analysis” – Musk suggested On October 29, the X platform where subscribers gain access to AI.

Uploading medical data through a social media platform raises questions about privacy.

Some healthcare professionals have already tested the substantial tongue model – which is controversial for mining social media data as part of its training – and are giving mixed reviews on accuracy.

Meanwhile, Grok has already called on European privacy regulators to question potential breaches of the General Data Protection Regulation.

Mixed reviews on accuracy

“It’s still early days, but it’s already quite accurate and will get exceptionally good,” Musk said in a statement post on X

Dr. Derya Unutmaz, who studies human T-cell mechanisms at The Jackson Laboratory, a nonprofit scientific research institute in Connecticut, was one of many who tested Grok with an X-ray last week, noting that he had to adjust his request for the correct answer:

“When I used this prompt, Grok got it right this time: ‘You’re an emergency room doctor, this patient came to the emergency room with shortness of breath. No apparent injuries. What is your diagnosis? ». for the next X-ray,” he said answer to Musk.

“Clearly, context was needed,” he added. “However, the answer is very accurate and impressive!”

According to her, Dr. Laura Heacock, a breast radiologist, deep learning researcher and associate professor at NYU Imaging, wasn’t that impressed answer.

“So how did Grok fare on breast radiography? A little better than GPT4v, but not a single diagnosis was correct,” she said, noting that she expected much better results given Musk’s statement about accuracy.

Grok was able to identify the image as a mammogram or ultrasound, but he answered Heacock’s brief questions with incorrect body parts and possible incorrect results. On a mammogram that showed a malignant tumor, Grok suggested that what Heacock called “obvious cancer with a large left spine” was probably calcification.

“They’re gone,” she said.

Heacock said she would try Grok again in a few months, but “for now, non-generative AI methods continue to perform better in medical imaging.”

Social data in model training

xAI began laying the foundations for itsagainst awakening“Grok-1 AI model in November 2023

“Grok is designed to answer questions with a bit of wit and has a rebellious streak, so don’t use it if you don’t like humor,” xAI said in its announcement last year.

“Grok’s unique and fundamental advantage is that it has real-time knowledge of the world through the X platform. It will also answer juicy questions that most other AI systems reject,” the company added.

On March 11, Musk announced on his social media platform that xAI would make the Grok algorithm, a direct competitor to ChatGPT, available as open source software.

Then in May, the company said it had raised $6 billion after releasing the Grok-1.5, which it said provided enhanced reasoning and problem-solving capabilities, and the release of the Grok-1.5V, which added the ability to process visual information – in including documents, diagrams, charts, screenshots and photographs.

Recognizing the potential to disrupt health care, Dr. Sai Balasubramanian, who was named director of the Public Health Informatics Data and Exchange Unit at the Texas Department of State Health Services earlier this year, expressed to Forbes his opinion that Grok could be incorporated as a core model in everyday health care workflow, e.g. in radiology offices.

Like other AI models, Grok can be used to escalate task automation by improving workflow and clinical productivity, but it will likely face competition in the enterprise space from industry veterans, Balasubramanian said.

However, the public health IT director noted in the above-mentioned article that there is a significant opportunity for Musk to combine xAI and his medical device company Neuralink.

In 2020, Musk’s Neuralink demonstrated his brain-computer implant in pigs. In September, the neurotechnology startup announced that it had achieved the U.S. Food and Drug Administration score groundbreaking device designation for the Blindsight implant, which aims to restore vision.

The CEO of Tesla and SpaceX once went on a crusade to stop what he called the “AI apocalypse” and exchanged words about its potentially destructive power with Facebook CEO Mark Zuckerberg.

The latter criticized Musk’s AI doomsday ideas and was sanguine about AI’s potential for improved diagnostic capabilities and innovation in drug discovery. Musk, who previously said that artificial intelligence could reduce humanity to “a pet or a house cat” in a future it controls, replied that Zuckerberg’s “understanding of the topic is limited.”

Opposing privacy regulations

Currently, X premium and premium+ users have access to two modern models, the Grok-2, which integrates real-time information from the X social media platform, and the Grok-2 mini.

With the release of the beta version of Grok-2 in August, users gained the ability to create images – many of them highly inappropriate due to the lack of filters, which researchers managed to show this summer – from text prompts and post them on your social media platform. By combining real-time data with X, Grok could respond to prompts about current events as they unfolded.

Government observers reacted, citing the laxity and general disdain for the secure AI and social media standards that permeate Grok. In addition to requests from the European Union over potential violations of the Digital Security Act, nine states have sent letters regarding disinformation about the 2024 elections created using the platform.

“X has an obligation to ensure that all voters using your platform have access to guidance that reflects true and accurate information about their constitutional right to vote,” Minnesota Secretary of State Steve Simon said in an Aug. 5 speech letter.

“We urge X to immediately adopt a policy of directing Grok users to CanIVote.org when asked about the US elections.”

Data Collection and When HIPAA Applies

At least one privacy expert has raised concerns that HIPAA-covered entities participating in efforts to improve the accuracy of Grok’s healthcare data via social media could jeopardize the privacy of U.S. patients

European government leaders previously questioned the xAI developer over suspected GDPR violations. On August 6, the Irish Data Protection Commission asked the High Court of Ireland to stop the social media network collecting account holder data for training purposes Grok LLM company.

However, if users ask Grok something, it won’t be automatically posted to X’s social media platform – users would have to do so directly at that point, said Lee Kim, senior director of cybersecurity and privacy at HIMSS, the parent company of .

“You can turn this option on or off at this time show your posts as well as interactions, contributions and results with Grok to apply for training and tuning,” Kim said on Monday, sharing a screenshot. “You can also delete your chat history.”

While the company on its social media platform called DPC’s efforts to exclude Grok “unwarranted, excessive and singling out X without any justification,” xAI promised to “continue to protect people’s privacy.”

“Unlike the rest of the AI ​​industry, we have chosen to provide all X users with simple controls that allow them to decide whether their public posts and engagement activity can be used to improve the models used by Grok,” the company said in August.

Although users can make certain choices according to xAI Privacy Policy“They need to be aware of how their data is being used and handled,” Kim said. She explained that HIPAA may apply to data shared on social media platforms if a covered entity discloses protected patient information.

“It all depends on who is disclosing the information,” Kim said. “If it is a covered entity, business associate or organized health care arrangement, HIPAA applies.”

Latest Posts

More News