Other animals
- Predation, starvation, dehydration and disease
- Competition for territory and mates
- Injury, parasites and environmental extremes
- Habitat loss and human-caused threats
Field notes on uncertainty · Edition 01
RISK.
To be alive is to be exposed. To be human is to imagine what might happen next.
From predators to phishing, disease to debt, earthquakes to embarrassment: an interactive field guide to the uncertainties we inherit, invent, share and sometimes insure.
Enter the atlas ↓Move your cursor. The fan of possible futures narrows when you commit and widens when you hover.
01Risk is not certainty
02Risk is not always danger
03Zero risk is rarely possible
The uncertainty engine / interactive field map
✦ 8 interconnected worlds
Explore the risk universe.
Choose a constellation. Follow a question into the atlas.
Turn on cascade mode, then tap one world to set the first failure.
Biological risk
The oldest risks: illness, injury, hunger, and the limits of the body.
01 / The terrain
What can go wrong?
Explore categories rather than a definitive ranking. The same event may involve multiple risks, and exposure differs dramatically across people and places.
✺ EXPERIMENT 07 / THE RISK REACTORINTERACTIVE · NO PERSONAL DATA
The future has branches.
One threat. Three levers. Watch how probability, consequences and preparation change an illustrative risk landscape. This is a teaching model, not a prediction engine.
₹10,000Expected loss before protection
₹6,000Expected loss after protection
₹4,000Illustrative expected reduction
Preparation reduces the severity of a loss in this model; it does not change the event's probability.
Assumes one event, fixed severity, a known probability and a protection measure that reduces severity by the selected percentage. Real-world events have uncertain probabilities, varied losses, costs of protection and dependencies. Expected loss is not a guarantee.
One planet. Different risks.
Humans
- Many of the same biological and ecological threats
- Financial, legal, institutional and reputational losses
- Digital fraud, data theft and infrastructure failure
- Long-term planning, imagined futures and cascading global risks
Important nuance: Animals also navigate social uncertainty, learn from experience and sometimes anticipate future events. The difference is not that animals have no social or cognitive risks; human symbolic systems and technological interdependence create additional kinds and scales of exposure.
How insurance changes the game
Insurance does not prevent the accident. It can transfer some of its financial consequences.
Many policyholders contribute premiums to a pool. Insurers use actuarial estimates, underwriting, diversification and capital to meet covered claims. Large pools of sufficiently independent, comparable exposures can make aggregate outcomes more predictable. But catastrophes, correlated losses, bad pricing and costs can produce losses.
Play with expected loss
Illustrative mathematics, not a quote or insurance recommendation.
Expected annual loss
₹2,000
Probability × severity. An average over many comparable exposures, not a prediction of your personal outcome.
Illustrative premiumExpected loss plus an assumed loading. Actual premiums also depend on coverage, exclusions, risk classification, expenses, taxes, regulation, reinsurance and capital requirements.
Premiums − claims − expenses + investment result = simplified insurer result
Live pool · illustrative
Watch strangers share the bill
Each dot is one policyholder. Run a year and meteors pick who has a loss. Shrink the pool and the results lurch. Trigger a storm and many dots are hit together, which is a correlated loss. Assumed here: the pool starts with half a year of expected claims, and 60% of the loading pays expenses.
0Years run
–Claims last year
–Insurer reserve
0Years in deficit
The law of large numbers improves estimates under suitable conditions; it does not guarantee insurer profitability. Insurance transfers specified financial risks under a contract, not every possible consequence of an event.
The predator has a URL.
Phishing
Messages impersonate trusted people or organisations to steal information or trigger harmful actions. Verify through a separately obtained channel.
NIST guide ↗Account takeover
Stolen credentials and compromised recovery channels can unlock financial and personal accounts. Unique passwords and phishing-resistant MFA reduce exposure.
CISA resources ↗Identity & synthetic media
Impersonation, manipulated audio or video, data breaches and social engineering can exploit trust. An apparently familiar voice is not proof of identity.
NIST framework ↗Ransomware
Attackers may encrypt data or threaten disclosure. Backups, patching and access controls can reduce harm.
Cyber risks ↗Platform dependence
Outages, account suspensions, algorithmic changes and vendor failures can interrupt livelihoods and communication.
Operational riskAI & automation
Misuse, unreliable outputs, concentration of power and poorly governed automation can introduce novel failure modes.
UN global risks ↗859,532complaints received by the FBI's Internet Crime Complaint Center in 2024
$16.6 billionreported losses across those complaints
193,407of the complaints were phishing or spoofing
United States only, and only crimes that people reported, so the true totals are likely different. Source: FBI IC3 2024 Internet Crime Report ↗
Danger is not the same as probability.
How likely?
A plausible event need not be a probable event. Look for a reference class, time horizon and credible evidence.
How costly?
A rare event can warrant attention if the consequences are severe, irreversible or widely shared.
For whom?
Hazards interact with exposure, vulnerability, resources and institutions. Risk is not distributed equally.
What is controllable?
Avoid, reduce, transfer, share or consciously retain risk. No response eliminates uncertainty completely.
What feels scary?
Novelty, vivid stories and perceived lack of control can distort how threatening a risk feels.
What is lost by avoiding?
Taking no chances can also carry costs: missed learning, relationships, enterprise and discovery.
Gut check · three rounds
Which feels scarier? Then see the counts.
Pick the one that feels scarier to you. There is no wrong answer yet.
Totals, not personal odds: far more people drive every day than swim near sharks, so exposure differs. Sources: WHO road traffic injuries ↗, WHO top 10 causes of death ↗, International Shark Attack File ↗.
Go beyond the atlas.
Starting points from regulators, public-health bodies and research institutions. Reports describe different populations and time periods; don't interpret their risk lists as universal personal rankings.
NAIC — Introduction to InsuranceRisk pooling, premiums, uncertainty and the law of large numbers
NIST — PhishingHow deceptive messages exploit trust and urgency
NIST — Cybersecurity RisksRansomware, phishing, account compromise and risk management
WHO — Climate Change and HealthEnvironmental hazards, exposure and vulnerability
IPCC — Key RisksInfrastructure, health, ecosystems and cascading climate impacts
United Nations — Global Risk ReportPerceived systemic and emerging risks; a survey, not a probability forecast
FBI IC3 — 2024 Internet Crime ReportReported cybercrime complaints and losses in the United States
WHO — Top 10 causes of deathWhich conditions account for the most deaths worldwide
WHO — Road traffic injuriesGlobal deaths, risk factors and prevention
International Shark Attack FileYearly worldwide counts of unprovoked shark bites
The aim is not a risk-free life.
It is a life in which we understand enough uncertainty to make room for courage, care and curiosity.
A project in risk literacy, not financial, legal, insurance, cybersecurity or medical advice.