Dennis Miloseski

Jony Ive x OpenAI — reimagining computing, with profound implications for healthcare. The acquisition of Ive’s startup io by OpenAI isn’t just a bold tech move, it signals the dawn of a new era in how humans interact with intelligent systems. With Jony Ive’s legendary design vision and Sam Altman’s AI engine, we’re moving toward a world where computing is not confined to screens or keyboards, but embedded in intuitive, ambient, and human-centered experiences. For healthcare, this means AI that feels less like software and more like a trusted partner. Imagine intelligent clinical assistants that are as empathetic in design as they are advanced in function — transforming how we diagnose, treat, and care for patients. This isn’t just evolution. It’s reinvention. The future of healthcare will be shaped by the tools we feel not the ones we use. https://lnkd.in/egtREvy8

2

A few weeks ago, I had the opportunity to participate in a Hackathon at Claremont Graduate University. My team's challenge was regarding AI in Healthcare and the work we did there enlightened and inspired me to write this article. I believe that the pace of AI innovation is staggering but trust is built when AI is transparent, explainable, and respectful of human judgment. In healthcare, especially, it's not enough for a model to be accurate, it must be understandable, equitable, and aligned with clinical workflows. We need AI that communicates uncertainty, invites oversight, and earns its place in decision-making. That kind of AI doesn’t replace humans, it amplifies trust in every interaction. This is the kind of AI we are building at AI Squared. #AI #HealthcareAI #HumanCenteredDesign #EthicalAI #RealTimeInference

33

If the product or service you're building isn’t seamless, it’s worth asking: where might things break down? Launching a medtech device, a consumer wearable, or even a warehouse automation system without designing the physical and digital experience together is a recipe for misalignment. When no one owns the full experience, things fall apart. And too often, it’s the digital layer, critical in connected systems, that gets scoped too late or treated like an add-on. This doesn’t happen because teams lack skill. It happens because they're siloed. The teams that succeed work differently: they integrate early, think across systems, and build digital as the backbone, not the afterthought. Big thanks to Christine Pillsbury for the great article, definitely worth a read for anyone building connected experiences today. #ProductInnovation #ProductExperience #Innovation

6

Harry Blumsack and Angelo Pennati have lived the problem they are solving. From the minute we met, they have been 'all over it' and deeply understood that 'compliance' does not sit squarely in compliance, and it slows down everyone on the team (especially high-cost engineers). If you believe that our 'devices will be our doctors' in the future, every line of code matters to the FDA and to the safety of the user. Only Infera avoids the tradeoff between speed and safety and enables faster and safer builds. (Thank you Allen Gannett for being the earliest believer.)

14

5 Comments

𝗦𝗮𝗺𝘀𝘂𝗻𝗴’𝘀 𝗖𝗘𝗦 𝟮𝟬𝟮𝟲 𝗩𝗶𝘀𝗶𝗼𝗻 𝗖𝗼𝗻𝗳𝗶𝗿𝗺𝘀 𝗪𝗵𝗮𝘁 𝗪𝗲 𝗛𝗮𝘃𝗲 𝗟𝗼𝗻𝗴 𝗦𝘂𝘀𝗽𝗲𝗰𝘁𝗲𝗱 At CES 2026 in Las Vegas, Samsung didn’t just show new products, they clearly articulated a strategic AI vision grounded in real ecosystem advantage. Samsung emphasized that 𝙞𝙩 𝙨𝙝𝙞𝙥𝙨 𝟱𝟬𝟬 𝙢𝙞𝙡𝙡𝙞𝙤𝙣 𝙘𝙤𝙣𝙣𝙚𝙘𝙩𝙚𝙙 𝙙𝙚𝙫𝙞𝙘𝙚𝙨 𝙚𝙫𝙚𝙧𝙮 𝙮𝙚𝙖r across mobile, TV, and home appliances, a scale that any competitors can match. What this means is simple but profound: Unmatched multi-device connectivity — Samsung isn’t just selling stand-alone gadgets, they are building an interconnected daily intelligence ecosystem. � AI integrated across every category, powered in part by Google’s Gemini, moving beyond novelty features to core user interactions. A living, personal AI companion, not a tool, but an experience layer that evolves with how people live, work, and interact with technology. This is the moment where Samsung’s long-held thesis — that its breadth of devices and ecosystem gives it a unique position to capitalize on AI adoption — didn’t just remain an industry belief, it was publicly declared as a core strategy. What they presented with Gemini is likely just the opening act. The real play starts when that AI becomes the connective tissue across all devices, experiences, and daily life, 𝙥𝙤𝙬𝙚𝙧𝙞𝙣𝙜 𝙫𝙖𝙡𝙪𝙚 𝙞𝙣 𝙬𝙖𝙮𝙨 𝙣𝙤 𝙤𝙩𝙝𝙚𝙧 𝙗𝙧𝙖𝙣𝙙 𝙘𝙖𝙣 𝙤𝙛𝙛𝙚𝙧. The complete presentation on the first comment 🎥

13

1 Comment

👉 What if every design choice could protect your brand from day one? Genichi Taguchi did more than sharpen engineering rigor, he reimagined quality as a profound act of care and accountability. 🔍 “Quality is not made on the line, it’s built into the design.” In the ’50s, Taguchi dared to challenge the status quo: Why wait to catch defects, when you can design variation out completely? His revolutionary Loss Function transformed quality from a binary pass/fail mindset into an emotional and financial storytelling tool. It showed that every deviation—no matter how small—costs real value to people and reputations. — 🎯 His insight was human at its core: Products must endure real-world chaos—stress, noise, friction—because people rely on them every day. That’s the heart of Robust Design: building resilience into products so that they don’t just work in the lab, they thrive in real life . 💡 What this means for you: When quality starts in the design phase, every part, process, and decision becomes an expression of responsibility—to customers, teammates, and society. 🧭 Investing in robustness is smart engineering—it’s your brand’s promise in action. It’s how you turn chance failures into steadfast trust, and how you make cost reduction a byproduct of excellence. 🗣️ “Cost is more important than quality, but quality is the best way to reduce cost.” — Genichi Taguchi ⸻ Build it right. Build it to last. Build it for them. 👉 Which quality pioneer inspires your work—and your values? ♻️ If this speaks to you, feel free to repost and carry Taguchi’s spirit forward. #TheGurusOfQuality #Taguchi #RobustDesign #QualityByDesign #LeanSixSigma

20

10 Comments

Building next-horizon AI experiences lays out design challenges and 4 principles for creating effective AI-native experiences. Companies are investing heavily in AI, yet many aren’t seeing meaningful gains. The challenge isn’t the technology itself, it’s designing experiences that allow AI to scale and create real value. The organizations that pull ahead will rethink how humans and intelligent systems work together. Design isn’t just a finishing touch; it’s a core driver of performance. This framework helps technology leaders, designers, and product teams shape AI experiences that are understandable, governable, and built for impact. https://mck.co/4bk61RF

2

1 Comment

Related to #CES2026 - accessibility innovation is more urgent than ever but it’s not new. In 1959, post-war Germany Mercedes design a vehicle for amputees. This car had no clutch pedal with the driver shifting gears with their hands. 67 years old lessons that good design builds for all users and innovation is not only about new tech but eliminating barriers that hold back users. Fast forward to today - as innovations in hearing (wearables, ambient infrastructure - LE Audio, Auracast etc) resemble consumer tech, we are moving ahead rapidly and yet learning from innovators of the past. Keen to hear what other think and particularly from folks working on innovation in general or specifically for persons with disabilities or seniors. #MakingMarketsWork #Accessibility #InclusiveDesign #Innovation #AssistiveTech #CES2026 Mercedes-Benz AG Arun Gore Shashaank Awasthi Srishti Garodia

7

OpenAI’s new mystery device, designed with former Apple designer Jony Ive, is rumored to be pocket-sized, context-aware and screen-free. It’s being called a “third device”—meant to coexist alongside your laptop and phone, quietly soaking in your environment. Now imagine this: a patient walks into your clinic, plops down on the exam table, and while you’re reviewing vitals, their AI device perks up and says: “Based on your elevated heart rate and that wheeze, I’ve diagnosed you with... stage 7 space flu. Here are 14 Reddit threads and a Youtube video discussing how you can treat this by drinking bleach.” It’s funny—until it’s not. These kinds of ambient, always-listening AI tools raise big questions for healthcare professionals. HIPAA doesn’t take a day off just because your new gadget has a British accent and sleek design. If this “third device” starts tagging along into clinics, hospitals, or occupational health screenings, we’ll have to figure out where privacy ends and convenience begins. We can't use Amazon Alexa devices in-clinic because they are always listening, and never HIPAA compliant. Smartphones start to beg the same compliance questions. Always-on, wearable AI will force this issue to the forefront. Separate from patients, the prospects of a clinician having an always-on AI device listening to all of their patient interactions--well that's just terrifying. As technology evolves, so must our awareness of how it intersects with patient care, compliance, and common sense. https://lnkd.in/ghf3_5-6

2

I’m excited to announce the launch of the embedded Smart on FHIR app for Lyft, powered by VectorCare and now available directly inside the Epic EHR showroom. For the first time, hospitals can connect Lyft’s nationwide transportation network to the Epic workflow using our SMART on FHIR framework. Care teams can request, schedule, and track rides directly from the patient chart without switching systems, calling vendors, or juggling portals. This closes one of the most persistent gaps in care. Transportation delays extend length of stay, slow discharges, increase ED boarding, and drive up no-show rates for appointments. The problem has never been demand. The problem has been that transportation workflows live outside the EHR. Now they live exactly where care happens. With the Lyft SMART on FHIR App, Epic users can: Order rides from inside the patient chart. Auto-populate demographics, locations, and appointment times through FHIR. Track ride progress and arrival status in real time. Support timely discharges, follow-up care, and outpatient visits. Improve access for patients who depend on reliable transportation. The outcome is simple. Discharges move faster. Emergency departments decompress. Patients get to appointments that historically would have been missed. Care teams spend less time chasing logistics and more time doing meaningful work. This is what it looks like when transportation becomes part of the care pathway instead of a separate administrative burden. The Lyft SMART on FHIR App is available now for all Epic health systems. We will expand to additional FHIR-enabled EHRs next, bringing this capability to more organizations nationwide. I am proud of what our teams have built, and even more proud of what this means for patient access, flow, and operational efficiency across healthcare. #Healthcare #PatientCare #PatientLogistics

22

4 Comments

Thanks Aaron Ferber! Just finished up a talk at HLTH on "Human Centered Design Principles in the Age of AI". Well attended, and well received. Couldn't have picked a better partner in Aaron Ferber @ Ascension, who has a background spanning startups and design consulting (IDEO) to mission driven health. We covered: 1. Funnels vs. Anti-funnels. Aka don't make patients download an app or you'll lose 70% of them. 2. Give value to get value. Make sure you have a "guest" experience - airlines & ecommerce figured this out long ago. 3. The power of "defaults". Aka when to use opt-in vs. opt-out and how radical a difference it made in say organ donation rates. 4. Test everything. You're wrong. You just don't know specifically how you're wrong until you test. Send your engineers into hospital/health facilities and have the sit in the waiting room for a couple of hours. 5. Using AI for a human touch. I.e. how to personalize at scale, what you know about education or reading level to alter the way you talk to patients. Full video in the next week or two. For now, here's what I said above in video format for those who didn't open this message :-) #HLTH25 and #HLTHUSA

138

8 Comments

With no affiliation, sharing this as I know it will be good. Peter Merholz's ideas have been instrumental in my shaping of design teams. As iF Design point out here - the right environment around the design team is key to producing a successful team. The role of the leader is to create that environment.

14

2 Comments

I’ve spent the last decade leading design, research, and innovation at places like Bose and the startup world before that. And while it’s easier than ever to build something, it's as hard as ever to know if it’s the right thing. That’s why I'm building RightThing.io. I wanted to see if I could bridge the gap between human intuition and large-scale validation. I went from zero coding to building our first sprintware tool called Scout that can run a thousand interviews in 24 hours. Here's a glimpse into it.

87

11 Comments

Generative AI in the Smart Home: Will It Obey as Well as It Promises? The new wave of generative-AI assistants has arrived with force. Alexa+, Gemini for Home, and the upcoming redesigned Siri promise natural conversations, context, memory, and human-like interaction. The “speaking well” part has advanced significantly. But the practical question remains: do they already obey as well as they should? Based on recent analyses (The Verge, Reuters, About Amazon, Android Central, 9to5Mac), today’s answer is: not yet. These assistants are learning to understand nuances “I’m arriving,” “make the house cozy,” “I’m going to cook” but they struggle precisely where it matters most: executing commands with precision, speed, and predictability. Reports highlight persistent issues: • routines that fail to trigger; • devices appearing “offline”; • ignored commands; • instability in the Google/Nest ecosystem; • higher latency in generative models (5–6 seconds in on-device tests). The reason is simple: LLMs were designed for flexible dialogue, not deterministic behavior. They interpret intention, but their responses may vary. In home automation, predictability is everything. Additionally, the market is still battling interoperability. The latest version of the Matter standard (1.4.2) focuses on stability not new features because platforms still cannot “speak the same language” reliably. Companies are moving slowly for strategic reasons: • security risks increase with generative AI; • cloud dependency introduces delays; • future monetization requires flawless launches; • Apple, in particular, avoids missteps in privacy and execution. Will the smart home improve? Yes. But we are still in the middle of the transition. Today we have a curious combination: assistants that talk like humans but obey like beta-stage devices. The definitive leap will come when three layers converge: 1. Local generative AI (near-zero latency), 2. Deterministic, auditable actions, 3. A fully stable Matter ecosystem. When that happens, we will witness the true revolution: a home that understands, decides, and executes with precision not just one that speaks well.

4

You can always build something you’ve already sold. But you can’t always sell something you’ve already built. When I was at IDEO, I watched David Kelley say “yes” to client projects that sounded impossible - six-week timelines, half the budget, specs that defied physics. He’d nod, smile, and say, “We can make that happen.” Then we’d start sketching concepts and building prototypes in the shop, and before long, we had figured out how. Kelley knew it was easier to build toward a committed outcome than to sell an idea no one had asked for. I still see this play out with founders. Many fall in love with the thing they’ve already built - the clever piece of tech, the whiz-bang widget - and then realize nobody’s asking for it. The best founders uncover hidden demand and build products that bring to light opportunities that would otherwise remain unseen. That order matters. Start with the problem or opportunity, then design the solution.

184

29 Comments

AccendoWave - A Pain #Data Company Omdia’s latest #research shows that the consumer #VR market continued to decline in 2024. Headset sales volumes fell by 10% in 2024 to 6.9 million units, down from 7.7 million in 2023. The market is expected to decline further in 2025, in line with Omdia’s earlier projections. The outlook reflects a challenging period for the consumer VR industry ahead of a forecasted rebound in 2026.   “The Apple Vision Pro’s highly anticipated debut failed to spark the resurgence that the VR industry hoped for, with momentum fading and significant developer interest still lacking ten months on,” commented George Jijiashvili, Senior Principal Analyst at Omdia. “Meanwhile, early sales data for the new entry-level Quest 3S reflects no significant boost to Meta’s headset sales, reinforcing Omdia’s view that VR will remain a niche market as long as it remains in its current form.” https://lnkd.in/gCEiH8yk

5

1 Comment

Ensuring the focus of AI is on the end-user is essential to optimize the utility of this technology as innovation continues. Listen below for an expert opinion.

4

70% of 510(k) submissions receive a deficiency letter on the first review cycle. For Human Factors submissions specifically, that number climbs to 90%. If you're leading a MedTech product through FDA clearance, those aren't the exception. That's the landscape you're operating in. The instinct when a deficiency letter lands is to start fixing things immediately. In our experience, that's exactly what you should not do. Jumping into remediation before you understand the root cause means you’ll fix the wrong things and face a second deficiency letter. Fix the right things in the wrong order, and you'll waste months. Either way, you're back to square one. In our latest article, we lay out the five root causes behind the vast majority of Human Factors deficiencies — from poorly structured URRAs to cross-document inconsistencies that undermine an otherwise credible evidence package — and a structured Phase Zero diagnostic framework for understanding precisely where your submission failed before you spend a dollar on remediation. Whether you're staring at a deficiency notice right now or building the submission that you hope will never get one, this is the diagnostic framework worth adding to your toolbox. Read the full article here: https://hubs.ly/Q046FQXR0 If your team is navigating a 510(k) deficiency and needs expert eyes on the evidence package, reach out directly. We're here to help you get to clearance — efficiently, and the first time. #MedTech #FDA #510k #HumanFactors #RegulatoryAffairs #MedicalDevices #UsabilityEngineering #FDAClearance #ProductDevelopment #HealthcareInnovation #THRIVE

2

Institutions don’t fail all at once. They fail when design can’t keep up with change. In Davos this week, I’ll be joining the inaugural gathering of the Institutional Research Network, a new initiative convened by The Digital Economist. The focus is not another set of conversations but the patient work of connecting research, markets, and governance so ideas can take institutional form and hold under real-world pressure. Some initiatives are about visibility. This one is about durability. #Davos2026 #InstitutionalDesign #IRN #TheDigitalEconomist

18

1 Comment