News & Analysis

When AI Starts Listening to You

Today’s AI ideas revolves around using it to listen to chests and accepting it as a teammate

No, we aren’t trying to be facetious here, just because of the polarization that AI has brought about in our workplace between those that are besotted with it and others who think they’ve got enough of it. But, it would be unfair on our part to not report stories that add a few more bytes to what’s an already animated conversation. 

So, here goes – a healthcare company secured $41 million to AI-enable a piece of medical equipment that has existed since 1816 while a software company believes that it is the bounden duty of all humans to accept AI and get them on board as teammates in all future progress that humanity could be eying in the future. 

Eko Health: AI-enabling stethoscopes to detect heart disease faster

The first one is Eko Health, which was founded in 2013, to create state-of-the-art stethoscopes that conform to the digital age. The company sold over half a million of their devices to physicians in the past decade and are now hoping that by AI-enabling it, they have access to massive datasets of chest sounds and ECG information at hand. 

Now Eko Health wants to parse all of this data through multiple algorithms and detect oncoming heart conditions, if any. Wonder what Rene Laennec, the French physician who invented the stethoscope back in 1816, would be feeling now. Because proponents would cheer AI as a lifesaver while opponents might baulk at the possibility of data leaks. 

CEO and founder Connor Landgraf believes that stethoscope examinations are inconsistent and inaccurate resulting in delayed diagnoses or even poor diagnoses. The purpose of the fund raise is to bring high-precision into the process using artificial intelligence so that patients get faster and more accurate diagnoses. 

Further explaining the process, Landgraf said physicians cannot always understand the subtle shifts between health and disease with a traditional stethoscope, while with the AI, they can combine heart sounds with cardiac rhythm assessments and then interpret them using the datasets of past samples. 

So, it is about putting technology into the hands of even a basic care physician or a nursing practitioner. The Eko stethoscope received FDA clearance in April for using AI to detect first signs of heart failure during routine medical tests. Earlier, the company also got FDA sign-offs for detecting heart murmurs via their algorithms. 

The latest Series-D round of funding of $41 million, saw the participation of Artis Ventures, Highland Capital Partners, NTTVC and Questa Capital, would be used to training AI algorithms to detect pulmonary conditions such as asthma and pneumonia and towards marketing and sales efforts to sell the device and the software outside of the United States. 

Asana: Getting people to work side-by-side with AI

Now, coming to the other story around AI, we’ve all heard of Asana and its mobile and web-based work-management platform. The company wants to shift away from the traditional approach of installing AI agents – software that can act autonomously and take up tasks. They believe AI would work better when it works alongside a human presence. Towards this end, they introduced “AI Teammates” in the beta version. 

The company’s head of AI Paige Costello says the choice of name is deliberate and hopes to create a mindset change around how people respond to working with AI. The idea Asana wants to propagate is that the future of work would be people working with other people as well as with AI. 

The company wants us to know that their statements aren’t mere wordplay. The single most important shift that Asana wants to bring about is the understanding that businesses should have a clear idea of what they want AI to do, what it has already done and how much would it likely cost to take the next step. 

The purpose of this shift is to bring about transparency in the AI-led processes and structures in order to help enterprises specific and create customized assistants that can execute core bits of their workflows. Costello thinks this is crucial because existing workflow tools are defined rigidly while AI Teammates allows more flexibility to move work within a company. 

And here’s how he explains the process: As work comes in, AI evaluates it to determine whether it’s ready to shift a gear or move back to the people who were doing it originally so that they could add more data. An example is a ticket raised in a company with some data points missing that AI could send back to the person who submitted it. 

Just as Eke Health sat on a treasure trove of data around human hearts and their sounds, Asana too is sitting on information related to how their customers work and what could be done to create models and train them so that the AI agents do exactly what is required of them. Remember! AI Agents tend to hallucinate quite a bit!