Super El Niño, Killer AI, and Ten Feet of Horse Manure

Why history makes me skeptical of both doom and utopia.

Every generation seems to get its own prediction of doom.

Today, we hear warnings that a Super El Niño may bring devastating floods, droughts, heat waves, and economic disruption around the world. The risks are real, and organizations such as NOAA and the World Meteorological Organization (WMO) have warned that a very strong El Niño can significantly disrupt global weather patterns. [noaa.gov], [wmo.int]

At the same time, headlines are circulating about artificial intelligence. Some AI researchers have publicly expressed concern that future superintelligent AI systems could pose existential risks to humanity within the coming decade. Evan Hubinger, Alignment Science Lead at Anthropic, recently stated his personal belief that there is greater than a 10% chance AI could "kill all humans" within the next decade if the alignment problem remains unsolved. [cbsnews.com], [forbes.com]

From horse-drawn carriages to artificial intelligence, history reminds us that the future rarely unfolds the way we expect.

When people encounter predictions like these, they often react in one of two ways:

1. Panic.
2. Dismiss everything as nonsense.

I think both reactions miss an important lesson from history.

The Horse Manure Problem

One of my favorite historical anecdotes comes from the late nineteenth century.

As cities grew, transportation depended heavily on horses. More people meant more horses. More horses meant more manure.

Some observers looked at the trends and concluded that cities such as London would eventually be buried under mountains of horse manure. Although the exact numbers often repeated online are debated, the underlying concern was genuine: urban planners faced a growing sanitation problem caused by the expanding use of horses.

The logic seemed sound.

If there are twice as many people, there will be roughly twice as many horses.

If there are twice as many horses, there will be roughly twice as much manure.

Project that into the future and the city becomes unlivable.

The prediction failed.

Not because people became better at handling manure.

It failed because the entire premise changed.

The automobile arrived.

Electric trams spread.

Urban infrastructure evolved.

The future did not continue in a straight line.

The Problem with Straight-Line Forecasting

Many predictions assume that today's reality will simply become more intense tomorrow.

This is often true for a short period.

It is rarely true over decades.

History is filled with examples:

• In the 1950s, many expected fleets of flying cars.
• In the 1990s, people imagined the internet primarily as a collection of static web pages.
• Around Y2K, some imagined computer failures would cripple modern civilization.
• Today, some imagine AI becoming either humanity's greatest servant or its executioner.

Reality usually takes a different path.

The future is not merely an extension of the present.

It is often a complete reorganization of the present.

Does That Mean We Should Ignore Warnings?

Not at all.

The danger of extreme predictions is not that they identify risks.

The danger is when we mistake risk assessments for prophecies.

Consider El Niño.

Scientists monitor ocean temperatures, atmospheric conditions, and climate patterns to estimate the probability of extreme weather events. Their forecasts are not declarations of destiny. They are warnings about increased risk. [noaa.gov], [wmo.int]

The same principle applies to AI.

When AI safety researchers discuss existential risk, they are usually not claiming that humanity is definitely doomed. They are highlighting scenarios that deserve careful study and preparation. In fact, many of the concerns focus on hypothetical future systems capable of self-improvement and autonomy that do not exist today. [thegrio.com], [cbsnews.com]

The existence of a risk does not guarantee the outcome.

What History Suggests

If history teaches us anything, it is that humans are surprisingly adaptable.

Over the last few decades alone, humanity has navigated:

• SARS
• The 2008 financial crisis
• Major natural disasters
• Global pandemics
• Rapid technological disruption

Each event exposed vulnerabilities.

Each event also revealed resilience.

People adjusted.

Organizations adapted.

Technology evolved.

Policies changed.

Life continued.

None of this means future challenges will be easy.

Only that it is unwise to underestimate humanity's ability to respond.

A Systems Administrator's View

As someone who works in technology, I often think of these issues in operational terms.

Imagine a critical production server.

An engineer reports that the server has a 10% chance of catastrophic failure within the next year.

A professional response is not panic.

Nor is it dismissal.

Instead, you:

• Monitor the system.
• Strengthen controls.
• Improve backups.
• Test recovery procedures.
• Develop contingency plans.

You take the warning seriously.

You simply do not assume the failure is inevitable.

Perhaps that is the right way to think about both climate risks and AI risks.

Final Thoughts

The horse manure prediction was not foolish.

It was based on the best information available at the time.

Its failure was a reminder that innovation, adaptation, and unforeseen developments can dramatically alter the future.

Today, forecasts about Super El Niño and advanced AI deserve attention. The risks are real. Experts should continue studying them. Governments, businesses, and communities should prepare accordingly. [wmo.int], [noaa.gov], [cbsnews.com], [forbes.com]

But history encourages humility.

The future has a long record of surprising both the optimists and the pessimists.

Perhaps the wisest position is neither fear nor complacency.

It is informed caution.

After all, the people worrying about horse manure never saw the automobile coming.

And there is a good chance that some of today's most confident predictions will one day look just as quaint.