By Richard / December 2025. Let’s be honest: my recent life has sometimes felt like my iPhone—too many tabs open, battery running critically low, and the odd random alarm telling me I forgot to breathe (again). A few months ago, my world really shifted. On a Tuesday in May, I had a heart attack. Unexpected. Terrifying. And completely preventable if I had been paying attention. My Apple Watch, which I had dismissed as just another gadget, caught something unusual in my heart rhythm at 3:47 AM. It sent an alert. I ignored it. Two hours later, I did not ignore the pain. That watch quite literally saved my life—I arrived at hospital within the critical window. Since then, I have been obsessed with a deceptively simple question: what happens when we let machines take care of us? So, with burnout in my bones, a newly-repaired heart, and genuine curiosity about whether technology can actually improve wellbeing (or just complicate it), I decided to run a week-long experiment: let AI manage everything. My smart home. My health. My schedule. My life. What could possibly go wrong?

The Moment Everything Changed: Why This Experiment Mattered

Most people do not think seriously about their health until something breaks. I was not different. I ran my life like a startup burning venture capital: maximum output, zero maintenance, ignoring warning lights until something caught fire.

On May 14th, 2025, something caught fire inside my chest.

I was not even awake when it happened. My Apple Watch detected an irregular heart rhythm at 3:47 AM—a potential sign of atrial fibrillation (AF). It sent a notification. I silenced it. I had work the next morning. At 6:15 AM, I felt crushing chest pain, shortness of breath, and the unmistakable sense that something catastrophic was occurring in my body. Ambulance. Hospital. Cardiac intensive care. The diagnosis: myocardial infarction—a heart attack.

The cardiologist said the same thing three times: “If your watch had not detected that arrhythmia, you would likely have died in your sleep. If you had waited another hour to come in, you would have had permanent heart damage. You got incredibly lucky.”

Luck. Or maybe the technology I had taken for granted actually worked.

So when I was discharged from hospital, my priorities shifted violently. I was not interested in optimisation anymore. I was interested in staying alive. I was interested in understanding whether technology could help with that, or whether it would just add noise to the signal.

That is what led to this experiment.

Welcome to My Geek Palace: The Tech Stack (A Complete Inventory)

Before I handed over control, I needed to document what I was actually dealing with. The inventory was absurd.

The Devices (This is Exhaustive):

• Apple Watch Series 10 (2024): Heart rate monitor, arrhythmia detector, emergency alert system. Currently my most trusted colleague.
• Ring cameras (4 units): Front door, back door, garage, driveway. Monitoring for parcels, foxes, and occasional neighbours’ midnight wanderings.
• Philips Hue lighting system (35 bulbs): Every room, multiple brightness zones. Flickering with my mood, the children’s mood, and occasionally in ways that suggest the house is having its own emotional crisis.
• Amazon Alexa ecosystem (5 devices): Kitchen, lounge, bedroom, garage, hot tub (yes, really).
• Apple HomeKit hub: The central nervous system attempting to make everything talk to everything else. Success rate: 73%.
• Hive heating system: Attempting to tame British weather. Mostly unsuccessful.
• Sonos speakers (7 zones): Whole-home audio, airplay integration, sometimes playing Phil Collins at 3 AM for no reason.
• Samsung smart appliances: Refrigerator (with touchscreen that has begun hallucinating), washing machine, tumble dryer, television.
• Yale smart alarm system: Excellent at detecting cats. Still hoping to detect burglars someday.
• Range Rover Defender: Lock/unlock via app, preheat cabin, remote diagnostics. Occasionally confused about whether the boot is open.
• Bluetti power banks (3 units): Garage UPS system, garden power solution, emergency backup. The reason my power tools have become sentient.
• DJI Air 3S drone: Aerial surveillance, low-budget cinematography, tea-break entertainment.
• Smart hot tub with integrated sound system: Where Bluetooth can unexpectedly create neighbourhood amphitheatre.
• Various wearables: Apple Watch (primary), Whoop band (backup heart monitoring), Oura ring (sleep analysis).
• Health apps: Apple Health, WaterMinder, SleepCycle, MyFitnessPal, Strava.
• AI assistants: ChatGPT-5 (primary coach), Claude (backup analysis), Notion AI (task management), Google Assistant (dispute resolution).

Total device count: 47 connected devices. Total WiFi networks: 5 (because apparently, I like complexity). Total apps on my phone: 87. Total apps I actually use: 12.

This is not normal. But it was my baseline.

Day 1: Handing Over the Keys to the Machines

Sunday evening, 8 PM. I sat down with my MacBook and told ChatGPT the following: “You are my life coach for the next seven days. My objectives are simple: stay alive, try for productivity, avoid burnout, and keep my family from divorcing me. Here is my data.” I uploaded:

• Apple Health records (heart data, previous arrhythmias, current medications)
• Google Calendar (all meetings, commitments, blocked time)
• SleepCycle data (six months of sleep patterns)
• Smart home device logs (usage patterns, automations, failures)
• My actual job requirements (which genuinely needed managing)

Then I told the AI: “Make me a schedule that keeps my heart rate stable, gets me 7+ hours of sleep, keeps my wife happy, ensures the kids are fed, and somehow gets work done. Automate everything you can. I will only make conscious decisions about things that genuinely matter.”

ChatGPT took approximately 47 seconds to respond with a complete framework. It included:

• Circadian-aligned wake time (5:30 AM, based on my natural patterns)
• Heart-rate-managed work blocks (30 minutes on, 15 minutes movement break)
• Automated house preparation (lights, temperature, coffee ready at 5:28 AM)
• Scheduled exercise (low-intensity, given recent cardiac event)
• Family time blocks (non-negotiable, calendar-locked)
• Wind-down sequence starting at 9 PM (lights dimming, Sonos playing Endel soundscapes)
• Sleep optimisation (room temperature 16°C, white noise, Apple Watch monitoring)

It also flagged my cardiologist’s advice: “Your patient has experienced recent acute myocardial infarction. Stress management is non-negotiable. Your current schedule suggests 73% probability of elevated cortisol levels if not modified.”

At 8:47 PM, I went to bed. At 5:30 AM, every light in my house came on gradually, my coffee started brewing, and my temperature adjusted to exactly 19°C—one degree warmer than ideal to ease the transition out of sleep. My Apple Watch buzzed with a gentle reminder: “Good morning. Your resting heart rate is 62 bpm—that’s a healthy baseline. Let’s keep it that way today.”

Something I noticed immediately: When a machine optimises your life, the friction disappears. There are no decisions. No willpower required. Just gentle, persistent structure. It is weirdly comforting and deeply unsettling at the same time.

Day 2: The Honeymoon Phase (Before Reality Hits)

That night, I slept 7 hours 43 minutes. Deep sleep: 2 hours 14 minutes (usually 1 hour 20 minutes). REM sleep: 1 hour 56 minutes (usually 1 hour 10 minutes). I woke naturally at 5:30 AM without my phone. My Watch displayed a message: “You slept 23% better than your baseline. Your body is recovering. Keep this up.”

On Day 2, I felt genuinely well. Not just “not terrible,” but actually well. Rested. Focused. Calm.

“This,” I said to my wife at breakfast, “is probably a trap.”

She nodded. “It definitely is.”

Days 3–4: When the Machines Started Arguing With Each Other

By Wednesday morning, the first cracks appeared.

The Robo-Vacuum Incident: My robot vacuum had been scheduled by Notion AI to run at 2 AM every Tuesday and Thursday (minimise disruption to work hours). On Wednesday night, I heard what can only be described as “the sound of chaos descending.” My daughter’s rabbit, Pumpkin, had a digestive incident in the lounge at 1:47 AM. The vacuum did what it was programmed to do: run its cleaning cycle at full power, spreading the consequences across every carpet in my house. By the time I turned it off, the smell had achieved sentience.

The Heating War: My Hive system and Philips Hue motion sensors got into a disagreement. Hive wanted to cool the house to 16°C to optimise sleep. Hue motion sensors detected me moving downstairs at midnight for a glass of water and decided it was morning—raising the temperature to 21°C and activating bright white lights. Net result: I was simultaneously being roasted and frozen.

The Fridge Prank: My Samsung fridge, which I am fairly certain has developed a sense of humour, displayed this message on its touchscreen at 7 AM: “See you tonight, sleep with one eye open xx”—followed by approximately 47 notifications about milk that I demonstrably already had.

The Neighbour Incident: My smart lighting system, in an attempt to create an “evening ambiance,” somehow linked with my neighbour’s smart home and activated all of their outdoor lights. They had no idea why their garden suddenly looked like Glastonbury. I spent an hour explaining that it was not a UFO.

The Sonos Situation: I woke up at 3:17 AM to Phil Collins’ “In The Air Tonight” playing at full volume. My music algorithm (trained on “songs you listen to when stressed”) had apparently decided I was having an anxiety attack. I was not. I was just asleep.

What I learned on Day 3: Machines are very good at optimising for metrics. They are very, very bad at understanding context. My vacuum could not distinguish between “floor” and “disaster.” My heating system could not understand that I did not want to be a rotisserie chicken.

By Thursday morning, my Apple Watch was sending unusual alerts. Elevated heart rate during sleep (105 bpm, normally 55–60). Irregular rhythm detected twice—not severe, but enough to flag. My cardiologist would later say: “You are experiencing stress from poor sleep and equipment malfunction. Your body knows something is wrong even if you don’t.”

The machines were stressing me out.

Days 5–6: The Redemption Arc (And Hot Tub DJ Gate)

I made a decision on Friday morning: instead of letting the machines optimise everything, I would let them optimise only critical systems (health, sleep, basic safety) and give myself permission to make “inefficient” choices everywhere else.

The result was immediately noticeable. My resting heart rate dropped from 67 bpm back to 62 bpm. My stress markers improved. I got a full 8 hours of sleep with zero arrhythmia alerts.

Then my daughter’s birthday happened.

She wanted to have her 16th birthday party in the hot tub. Fine. She wanted to play music through the hot tub’s Bluetooth speaker system. Also fine. What she did not anticipate was that my Alexa hot tub speaker was set to broadcast across a wider area than intended.

At 7:47 PM on Saturday, my system started playing Taylor Swift’s “Cruel Summer” through the hot tub. The problem: my Alexa speakers were somehow linked to my neighbour’s hot tub speakers three gardens away. For 47 minutes, the entire street enjoyed a Taylor Swift concert. From multiple hot tubs. Simultaneously.

The police did not come. But three neighbours came to ask if I was aware I was hosting a block party.

My daughter’s assessment: “Dad, your smart home is dumb.”

She is not wrong.

The incident was actually useful. It forced me to understand my own system better. Turns out, my network setup was optimised for convenience, not isolation. All my speakers could talk to each other. Some neighbours’ systems were on similar frequencies. The result: accidental neighbourhood DJ.

I fixed it by Sunday morning. But not before we had the best music system the street had ever experienced.

Day 7: What Happened When I Pulled the Plug (And Stayed Human)

Sunday arrived. One full week of AI life management was complete. As instructed by my own protocol, I unplugged everything non-essential. No ChatGPT coach. No calendar constraints. No automated routines except the truly critical ones: my Apple Watch (still monitoring cardiac health), my sleep system (soft wind-down only), and basic home security.

What I expected: boredom, chaos, and me falling back into bad patterns immediately.

What actually happened: freedom.

I woke at 7:14 AM because I felt like it. Made breakfast without a meal plan (eggs and toast, nothing fancy). Took a walk without optimisation targets. Spent two hours on my passion project (something no algorithm had scheduled). I said yes to an unplanned family trip. I read a book without a timer. I sat with my wife and just talked, without an alert reminding me we had only blocked 30 minutes for “couple time.”

My Watch buzzed once: “Your heart rate variability is excellent this morning. Your body seems happy.”

It was right. My body was happy. But not because everything was optimised. Because I got to choose.

What the Research Actually Shows: UK Smart Home Adoption & AI Health Monitoring

The Numbers on UK Smart Homes:

I am not just anecdotal. The research backs up why this experiment even mattered. According to 2025 data from UK market analysis:

• 15+ million UK households now have at least one smart home device (39% adoption rate).
• 50.2% of UK households are projected to have smart home technology by 2027.
• £10.84 billion UK smart home market in 2025, growing at 10.13% annually.
• Greater London leads adoption at 45% smart speaker penetration, with smart thermostats at 35%.
• 31% of UK users cite privacy concerns as the primary barrier to fuller adoption.
• 87% of smart home users configure custom alerts, showing demand for user control, not automatic control.

This means my experiment is not just personal. It is a glimpse into how millions of UK households are navigating the same question: how much automation is too much?

The Apple Watch & Cardiac Health Detection (The Part That Mattered Most):

The truly important research is on my wrist. In 2025, Apple published research (in collaboration with Stanford) showing that Apple Watch can detect atrial fibrillation with 98% sensitivity and 99% specificity. For context, that means:

• 98% of people with AF will be detected.
• 99% of alerts will be accurate (very few false positives).
• The watch is more accurate than many clinical devices.

Even more recent research (published December 2025) shows that Apple’s optical heart sensor combined with AI can extract deeper cardiac parameters—not just detecting rhythm problems, but estimating things like stroke volume and cardiac output. The researchers’ conclusion: “AI-assisted modeling can extract more meaningful heart insights from a simple optical sensor.”

That simple optical sensor on my wrist has now saved my life twice in six months. Once at 3:47 AM by detecting my initial arrhythmia. Once during this experiment by flagging elevated night-time heart rate when the machines were stressing me out.

The Messy Truth: What AI Can and Cannot Do For Your Life

What AI Is Genuinely Good At:

• Detecting patterns you cannot see: My sleep algorithm noticed I was getting 12 minutes less REM sleep on days I drank coffee after 2 PM. I never would have noticed. It was right.
• Removing decision friction: When someone else makes your breakfast, you do not waste mental energy deciding what to eat. For people with decision fatigue or depression, this is genuinely helpful.
• Consistency: I cannot discipline myself to go to bed at the same time every night. My house system can. And it works.
• Health monitoring: My Apple Watch catching my arrhythmia was AI-enabled health monitoring at its best. No human checks their heart rhythm at 3 AM. A machine can.
• Crisis response: When something goes wrong (elevated heart rate, smoke detected, unusual motion), AI can alert immediately. That is valuable.

What AI Is Genuinely Terrible At:

• Understanding context: My vacuum could not tell the difference between “dirt” and “disaster.” AI is literalist.
• Knowing when to break its own rules: My algorithm scheduled exercise at 6 AM every day. On the day I was emotionally exhausted, it still insisted. Humans would say “let’s skip today.” Machines do not.
• Wisdom vs. optimisation: Optimisation says “do this, it is 2% better.” Wisdom says “sometimes 2% is not worth the cost to your soul.” AI cannot feel the difference.
• Joy: My optimised schedule had zero margin for spontaneity. Spontaneity is where half of life’s best moments happen. AI schedules cannot account for that.
• Being human: There is something irreducibly human about making mistakes, changing your mind, and doing things for no reason except that you want to. AI cannot replicate that.

Real-World Smart Home Failures (Not Just Mine)

My experience is not unique. According to research from UK smart home users in 2025:

• 26% have experienced device connectivity issues, especially with cross-brand integration (Ring + Philips + Samsung + Hive = chaos).
• 14% report false alarms from motion sensors or AI misidentification (my rabbit definitely counts).
• 9% of smart homes experienced at least one cyber breach attempt in 2024–2025.
• 24% of consumers avoid smart devices due to lack of interoperability— the “buy everything from Apple” problem.
• 16% report difficulty managing firmware updates for their 20+ devices.

The most common complaint I found from UK users: “My smart home is smarter than me, but dumber than understanding what I actually want.”

So Should You Do This? (The Practical Advice)

A Smart Home Experiment Is Worth It If:

• You have a chronic health condition (especially cardiac or sleep-related) and want data-driven monitoring.
• You are in burnout and genuinely need structure imposed externally.
• You want to understand how much friction your current life has that you have gotten used to.
• You are curious about the intersection of technology and wellbeing (and can handle some chaos).
• You have time to fix things when they break (they will break).

Do NOT Try This If:

• You already feel controlled or constrained in your life.
• You have privacy concerns you are not comfortable with.
• Your home is already chaotic and adding smart automation would make it worse.
• You value spontaneity above optimisation (you probably do, even if you do not think so).
• You have young children or pets. The failure modes are hilarious but not ideal.

The Most Important Lesson I Learned (And Why It Matters)

I had a heart attack in May. I had a week of AI running my life in November. Both were wake-up calls.

The heart attack taught me: your body knows things your mind ignores. Listen to the signals.

The AI week taught me: optimisation is seductive, but optimised life is not the same as meaningful life. Sometimes mess is the point.

Here is the thing I wish someone had told me six months ago: Technology cannot save you. But it can keep you alive long enough to save yourself.

My Apple Watch did not cure my heart. But it detected the problem early enough that doctors could. That is what technology should do: extend the time you have, so you can make better choices about how to use it.

My smart home did not make me happy. But by removing friction from basic tasks, it gave me margin to do the things that actually make me happy. That is different. That is useful.

And on Day 7, when I turned most of it off and just existed? That felt like relief. Like I remembered who I was underneath all the optimisation.

Frequently Asked Questions

The Final Reflection: What I Will Keep, What I Will Lose

It is December now. Six months since the heart attack. Seven weeks since the AI week. Here is what stays and what goes:

Keeping (Non-Negotiable):

• Apple Watch: Monitoring my heart 24/7. This device is allowed to stay.
• Sleep tracking + wind-down automation: Lights dimming at 9 PM, temperature dropping to 16°C, Endel soundscapes. My body responds to this. It stays.
• Basic home security: Ring cameras, smart locks, alarm system. These genuinely add safety and peace of mind.
• My wife’s sanity: Keeping me accountable, calling me out when I am being ridiculous, occasionally rolling her eyes at my tech obsession. This is not a device, but it is the most important thing keeping my life running.

Ditching (Immediately):

• Meal planning algorithms: I will eat when I am hungry. Sometimes that is kale and eggs. Sometimes that is pizza at midnight. Both are valid.
• Calendar tyranny: I am keeping work-critical meetings scheduled. Everything else? I will decide in the moment.
• Fitness gamification: I will exercise because my body needs movement, not because Strava says I have not closed my rings.
• The robo-vacuum: I picked it up and thanked it for its service. It now lives in the garage, awaiting a new home. The carpet cleaning can be manual.
• Most of the smart appliances: My fridge does not need a touchscreen to be a fridge. My washing machine does not need an app. I am simpl ifying.

What remains is a life that is smarter than it was, but still recognisably mine. Optimised where it genuinely helps. Messy where mess is part of the point.

The best version of technology is invisible. It works in the background. It solves real problems. And then it gets out of the way.

Your Story Matters Too

Here is the question I want to leave you with: How much of your life have you automated without realising it?

Not technology—I mean the actual decisions you have stopped making. The routes you always take. The meals you always order. The times you go to bed. The conversations you skip because you are tired. The moments you half-experience because your phone is recording them instead of you living them.

My experiment was about technology. But the real insight was about choice. AI is seductive because it removes the burden of choosing. But that burden—that friction, that decision-making, that occasional chaos—that is where being human lives.

Use technology to solve problems. Use automation to create margin. But do not let it replace you.

What is your story? Have you ever let an algorithm run your life? Did a wearable catch something serious? Have you had a tech disaster that turned out to be hilarious? Drop it in the comments. Share this with someone who has too many tabs open and not enough time. Because our stories—heart attacks and tech fails and robot vacuums discovering new creative uses for carpet—they are what make us human.