In partnership with
Welcome back to Healthy Innovations! 👋
This week's Deep Dive was sparked by Amy Webb's Convergence Outlook 2026, launched at South by Southwest (SXSW) earlier this month.
Webb is one of the world's leading technology futurists and founder of the Future Today Strategy Group (FTSG) – and her report is essential reading for anyone working in health and biotech.
Over the coming weeks I'll be highlighting the most relevant topics, starting with Living Intelligence.
Let's dive in!
Healthcare news for decision-makers
Knowing the healthcare headlines is easy.
Understanding what they mean for the business? That’s the hard part.
Healthcare Brew is a free newsletter breaking down the forces shaping the healthcare industry—from pharmaceutical developments and health startups to policy shifts, regulation, and tech changing how hospitals and providers operate.
No clinical deep dives. No overstuffed jargon. No guessing what actually matters. Just clear, focused coverage built for the people making decisions behind the scenes.
Join 135K+ administrators and healthcare professionals staying informed, for free.
What is living intelligence?
For years, most health systems have operated in three distinct phases: collect data, analyze what's been gathered, then wait for humans to decide what should happen next.
Living intelligence collapses those steps into one continuous flow. Sensors track biological signals and behavioral patterns in real time. Adaptive AI models interpret what's happening. Then the system acts – adjusting insulin doses, flagging a deteriorating patient, or confirming a pill has been swallowed – without pausing for human approval.
Webb and her team define it as the fusion of AI, advanced sensors, and bioengineering into systems that don't just observe the body but respond to it. The AI most people picture in healthcare – reading scans, flagging anomalies, helping with discharge notes – is already part of an older model. What's emerging now goes considerably further.
Here are three places where that's already happening.
Digital twins of real cells
One of the clearest expressions of living intelligence in action is the race to build virtual cells – AI models that create a continuous, bidirectional feedback loop with biological data. The models ingest cellular information, interpret it, and generate predictions that feed back into research, compressing what used to take months of lab work into seconds of computation.
A research team backed by the Chan Zuckerberg Biohub, working with Stanford scientists, released TranscriptFormer in 2025 – trained on data from 112 million cells across twelve species and 1.5 billion years of evolutionary history. It can classify rare cell types, identify virus-infected cells, and predict drug effects without additional training data. Theofanis Karaletsos, the team's head of AI for science, has described it as a first step rather than a destination, already being superseded by more powerful iterations.
In October 2025, the team announced a collaboration with NVIDIA to scale the work across petabytes of biological data. A companion reasoning model, rBio, followed in August 2025, allowing scientists to interrogate virtual cell simulations in plain language.
Xaira Therapeutics launched its own virtual cell model, X-Cell, in March 2026, and the Arc Institute ran a Virtual Cell Challenge in June 2025 with USD 100,000 in prizes and hundreds of competing teams.
Drug candidate attrition in traditional pharma sits around 96%, largely because feedback loops between biological prediction and lab validation are slow and linear. Virtual cell models begin to close that loop – and even modest improvements in that failure rate translate directly into medicines reaching patients faster.
Skin for robots that senses pain
Living intelligence doesn't only apply to human biology. Researchers at the City University of Hong Kong have developed neuromorphic robotic e-skin – NRE-skin – that mimics the way human skin senses touch and triggers reflexes without routing signals through a central processor.
Most robots today sense contact the way a spreadsheet processes data: information travels up to software, an algorithm decides how to respond, then the robot acts. NRE-skin collapses that into a local reflex. The system distinguishes between light pressure and potentially harmful contact – registering the latter as "pain" – and triggers a protective response instantly, much like the spinal reflex that pulls your hand from a hot stove before your brain has registered what happened.
The healthcare implications are direct. As robots move into hospitals, eldercare facilities, and surgical environments, they need to interact physically with patients safely and predictably. A robot that senses and responds to harmful contact in real time – without waiting for a central system to process the situation – is a fundamentally different and safer proposition than anything currently deployed.
Smart pills that communicate from the stomach
While virtual cells operate in computers, another strand of living intelligence is entering the body itself.
In January 2026, MIT engineers published research in Nature Communications describing a biodegradable smart pill called SAFARI. Inside the capsule is a zinc-cellulose radio frequency antenna that activates only after the pill is swallowed, transmitting confirmation to an external receiver within ten minutes. The biodegradable components break down within a week, leaving only a tiny RF chip that passes naturally through the digestive tract. The medication itself doesn't change – the smart capsule simply wraps around existing pills.
The gelatin-coated capsules (left) and the capsule without the coating (right). The capsule can be broken down and absorbed by the body. Image source: MIT
Medication non-adherence contributes to around 125,000 preventable deaths each year in the US alone and generates over USD 100 billion in avoidable costs annually.
"The goal is to make sure that this helps people receive the therapy they need to help maximize their health," said Professor Giovanni Traverso, senior author of the study.
Priority patient groups include transplant recipients, people being treated for tuberculosis or HIV, those with recently placed stents, and patients with neuropsychiatric conditions that can interfere with consistent medication use. Funded by MIT's Department of Mechanical Engineering and ARPA-H, SAFARI is heading toward human trials.
From episodes to signals
The thread connecting these developments is the same logic Webb identifies across living intelligence systems: healthcare shifting from episodic encounters to continuous, signal-driven intervention.
By July 2025, the FDA had authorized more than 1,250 AI-enabled medical devices, up from 950 a year earlier. Many now adapt in real time – closed-loop insulin delivery adjusting doses based on live glucose data, AI cardiac devices predicting arrhythmias before they occur, remote monitoring platforms flagging deterioration before a clinician has reviewed a single chart.
What distinguishes all of these from earlier digital health tools is feedback. They don't just report what's happening – they act on it.
What this means for healthcare leaders
The FTSG report makes a point that gets underestimated: organizations fixated on AI model quality are often missing the role of sensing infrastructure. Competitive advantage in living intelligence belongs to those who control continuous, high-quality data streams from the body and from clinical environments – not just those with the best models to interpret them.
Medical device companies that own persistent sensing endpoints are better positioned than those selling one-time devices. Healthcare delivery networks acting on real-time signals will increasingly outperform those organized around scheduled appointments.
Governance tends to get deferred too long. The EU AI Act, in full effect from January 2026, classifies medical AI as high-risk, requiring documentation of training data, bias checks, and human oversight policies. The organizations moving fastest are embedding compliance from the start.
That matters more than it might seem. Systems making consequential decisions faster than humans can review them compress the window for oversight. Human override mechanisms are not a feature to add later – they are the design requirement that determines whether any of this is deployable in regulated clinical environments.
What's coming
Virtual cell models will grow more powerful as data volumes scale. Smart ingestible technology will move into human trials. Sensor-fed systems will push further into the home, making continuous monitoring the default rather than the exception for patients with chronic conditions.
The longer-term shift is one where AI stops being a tool that clinicians consult and becomes infrastructure running continuously in the background – sensing, interpreting, and acting across the entire care pathway. Webb's report calls this the move from human-managed workflows to autonomous execution: systems that don't wait to be asked.
The organizations building their sensing strategies, governance frameworks, and data infrastructure now will not have to scramble later. The window for getting ahead of this is open, but it won't stay open indefinitely.
Innovation highlights
🩹 Speeding up healing with AI. Duke University biomedical engineers, working with Nokia Bell Labs, have built an imaging platform that tracks wound healing beneath the skin without a single biopsy. Their custom optical coherence tomography system, paired with AI, automatically maps tissue structure and blood flow in real time as wounds close. Testing it on mice, the team also found that stiffer hydrogels speed up tissue regeneration. Next up: predicting how chronic wounds heal in diabetic patients.
🧠 AI clocks brain age. UC San Francisco and Beth Israel Deaconess Medical Center researchers used machine learning to analyze overnight sleep EEG signals and estimate "brain age" – how old the brain appears compared to a person's actual age. Studying around 7,000 people over up to 17 years, they found each 10-year gap where brain age exceeded chronological age was linked to a nearly 40% rise in dementia risk. The findings, published in JAMA Network Open, suggest wearable EEG devices could one day flag risk early.
🧬 Tiny bots, big dreams. DNA isn't just the molecule that carries your genetic blueprint — scientists are now engineering it into microscopic robots. These nanoscale machines are designed to travel through the bloodstream, target diseased cells, and deliver drugs with surgical precision. Early research also points to DNA robots capturing viruses like SARS-CoV-2, and building molecular computers beyond current manufacturing limits. Major hurdles remain, but the field is advancing fast through smarter design, AI-assisted simulation, and cross-disciplinary collaboration.
Weird and wonderful
🦟 Science's worst job posting. Georgia Tech professor David Hu needed a human target for his mosquito study, and undergraduate Chris Zuo volunteered – promptly suffering what Hu describes as a "full-body massacre" after his mesh suit failed spectacularly.
By the time Zuo had upgraded his protective gear, he'd become a grad student. Over three years, a CDC-provided camera tracked mosquito flight paths at 100 frames per second, logging 20 million individual flights – more than had ever been recorded. Mosquitoes wander aimlessly with no target, do a casual fly-by for visual-only targets, and spiral into a frenzied orbit around warm-blooded, CO2-emitting humans.
Scientists can now predict which body parts mosquitoes prefer to attack. Zuo's suffering, it turns out, was for something.
Image created using Canva AI
Thank you for reading the Healthy Innovations newsletter!
Keep an eye out for next week’s issue, where I will highlight the healthcare innovations you need to know about.
Have a great week!
Alison ✨
P.S. If you enjoyed reading the Healthy Innovations newsletter, please subscribe so I know the content is valuable to you!
P.P.S. Healthcare is evolving at an unprecedented pace, and your unique insights could be invaluable to others in the field. If you're considering starting your own newsletter to share your expertise and build a community around your healthcare niche, check out beehiiv (affiliate link). There's never been a better time to start sharing your knowledge with the world!
