AI In Healthcare: Ant Group’s App, Funding Rounds, Growing Use Cases And Woes

Artificial intelligence (AI) has been revolutionizing industries the world over, and the healthcare sector is not a stranger to its beautiful and brilliant applications.

AI is helping medical professionals provide personalized care, enabling faster decision making, developing new medicines, enhancing telehealth, increasing patient access, quickening diagnoses, detecting early signs of cancer (sometimes even before symptoms appear), boosting cure rates, improving operational efficiencies, creating detailed reports and diagnoses with agentic AI, and more.

In addition to this, AI companies and healthcare providers are always working together to improve AI’s tools and services to enhance existing applications and add more use cases.
 

Google DeepMind’s AlphaGenome

Through a blog post published on its website, Google DeepMind introduced AlphaGenome, a new, unifying DNA sequence AI model that processes long DNA sequences to deliver high-resolution predictions.

This new tool can more comprehensively and accurately predict how single variants or mutations in human DNA sequences impact a wide range of biological processes regulating genes.

The idea is to help scientists better understand genome function and disease biology, while driving new biological discoveries and the development of new treatments.

“Our AlphaGenome model takes a long DNA sequence as input—up to 1 million letters, also known as base-pairs—and predicts thousands of molecular properties characterizing its regulatory activity. It can also score the effects of genetic variants or mutations by comparing predictions of mutated sequences with unmutated ones,” reads the blog post.


Using the tool, scientists can predict where genes start and where they end in different cell types and tissues, where they get spliced, the amount of RNA being produced, and which DNA bases are accessible, close to one another, or bound by certain proteins.

Google DeepMind sources its training data from large public consortia including ENCODE, GTEx, 4D Nucleome, and FANTOM5.

The tool has been made available in preview via its AlphaGenome API for non-commercial research, and plans to release the model soon.
 

Ant Group’s AQ

In a push to grow its presence in the healthcare sector, Chinese fintech company Ant Group, which is an Alibaba affiliate, launched a new healthcare smartphone app called AQ.

The app will help users manage their daily healthcare needs with over 100 AI-powered services, including doctor recommendations, medical report analysis, and personalized medical advice, consultations with AI avatars of real medical professionals, as well as connect users to digital services from over 5,000 hospitals and nearly 1 million doctors across China.

AQ is powered by Ant Group’s Healthcare Large Model, which comes with advanced medical reasoning and multimodal interaction capabilities that see it rank high across various benchmarks, achieving first place in the HealthBench and MedBench evaluations.

The model is designed to empower medical institutions and doctors to offer more efficient, accessible, and personalized services to users with AI-enabled solutions.


“Ant Group hopes that through AQ, it can provide everyone with a trusted healthcare manager, advancing inclusive healthcare and bringing every Chinese citizen one step closer to a healthier life,” said Cyril Han, CEO of Ant Group.

Currently, AQ has launched in China, but considering its name is in English and stands for “answer your question,” the company has plans for it to be rolled out in other countries.

Ant Group has remained active in the healthcare sector, working with IT industry leaders to integrate its advanced Healthcare Large Model into various tools, helping hospitals, insurance institutions, and other healthcare firms develop AI assistants, and collaborating with nearly 200 prominent doctors in China to develop AI Doctor Agents for a wide range of purposes.
 

Funding Rounds

Mandolin is an AI automation platform for specialty drug access just announced that it has raised $40 million in a round led by Greylock Partners and participation from SignalFire, Maverick, SV Angel, along with Jerry Yang (co-founder of Yahoo!) and Guillermo Rauch (CEO of Vercel).

The added capital will help the company develop and deploy several thousand infusion centers, specialty pharmacies, and health systems to boost automation in the healthcare sector and “support roughly one million therapy journeys every year.”

The company plans to scale its Agentic Operating System, for which it spent its first year laying the foundation.


Ahead of this, voice AI startup SuperDial revealed that it raised $15 million in a debt and equity series A funding round led by SignalFire and participation by existing investors Slow Ventures, Box Group, and Scrub Capital.

The funding comes with $3 million in venture debt for SuperDial to invest in R&D and go-to-market initiatives.

The company aims to help U.S. healthcare billing and provider organizations save time and money by automating insurer calls with end-to-end AI agents, which handle outbound phone calls—navigating phone trees, waiting on hold, and conducting live conversations with payer reps. When the AI agents can’t complete a task, they redirect the call to a human.


Amid all this, the world continues to battle the ill effects of AI, which is causing harm to the mental and physical health of people.
 

In The Throes Of Woes

Multiple studies have found that regular users of AI platforms may exhibit less cognitive ability than non-users.

In the most recent study titled “Experimental Evidence of the Effects of Large Language Models versus Web Search on Depth of Learning,” researchers from the University of Pennsylvania’s Wharton School observed more than 4,500 participants (4,591 to be exact), and learned that participants that used large language models (LLMs) such as ChatGPT had a tendency to develop shallower knowledge than those who used web search engines such as Google.

Furthermore, LLM users displayed less interest in forming their advice, and the advice they formulated lacked originality when compared to search engine users.

“This shallower knowledge accrues from an inherent feature of LLMs—the presentation of results as syntheses of information rather than individual search links—which makes learning more passive than in standard web search, where users actively discover and synthesize information sources themselves,” reads the abstract.

This follows an experiment conducted by the Massachusetts Institute of Technology (MIT) where participants were divided into three groups and had to write essays. The first group was told to rely on their brains, the second could use Google to search for relevant information, and the third was allowed to use ChatGPT.

During the study, participants wore headsets with electrodes to measure brain activity. It was observed that participants who used ChatGPT displayed less brain activity than either of the other groups, including fewer widespread connections between different parts of their brains, less alpha connectivity (creativity), and less theta connectivity (working memory). 

What do you think about the latest AI developments in the healthcare sector? Do you think health professionals should take heed of the problems that grow out of extended AI use?

Let us know in the comments below!