The Next Y2K Problem

Y2K, short for “Year 2000,” was a potential computer bug caused by how dates were formatted in older software. To save memory space, early computers used two-digit years—like “97” for 1997—which in the new millennium risked misreading “00” as 1900 instead of 2000, potentially disrupting systems that depended on accurate dates (read 101).

Though a kind of panic occurred in 1999, the Y2K issue surfaced in technical literature as early as 1984. Long before it became a global concern, researchers were already flagging the two-digit date flaw. A 1984 book, "Computers in Crisis," outlined how the year 2000 rollover could break financial, governmental, and technical systems if left unaddressed.

In the late 1990s, many feared this glitch could cause widespread failures in banking systems, power grids, transportation networks, and other critical infrastructure. This idea took hold of the public imagination, spawning doomsday predictions, a booming survivalist market, and a massive global push to audit and repair vulnerable systems before the deadline—work that cost an estimated $300B-$500B. 

Because of the extensive preparations, Y2K passed without significant disruptions, however, its legacy endures. The crisis helped modernize global IT systems, accelerated the outsourcing of programming jobs, and exposed society’s dependence on digital infrastructure—prompting long-term shifts in cybersecurity and software maintenance.

The Year 2038 problem is the next potential computer time rollover bug. Many older systems store time as a signed 32-bit integer counting seconds since Jan. 1, 1970. That counter maxes out on Jan. 19, 2038—overflowing into negative time and sending clocks back to 1901, potentially crashing any older software that depends on accurate dates. The Y2K38 bug is also known as the end of 32-bit Unix time and the year 2038 problem.

Y2K and Y2K38

Do you remember the Y2K scare? It is also known as the Millennium Bug. On this Eve of a new year, I am recalling this scare that stemmed from a widespread concern in the late 1990s that many computer systems would fail when the year changed from 1999 to 2000.

Why? Many older computer systems and software programs represented years using only the last two digits (e.g., "1999" was stored as "99"). It was feared that when 2000 arrived, these systems might interpret "00" as 1900 instead of 2000, leading to several problems.

Systems that relied on accurate date calculations could produce errors or fail entirely. For example, financial systems calculating interest rates or loan payments might miscalculate. Concerns arose about critical systems in utilities, transportation, healthcare, and government shutting down. Files or databases might become corrupted due to incorrect data processing.

Probably the greatest concern was in banking and finance where it was feared that miscalculated transactions, stock market crashes, or ATM malfunctions might occur.

Some people predicted power grid failures or water system disruptions, and aviation navigation systems and air traffic control collapsing.

What if there were malfunctioning military systems, including nuclear launch systems?

And so, billions of dollars were spent worldwide to identify, update, and test potentially vulnerable systems. IT professionals worked tirelessly to ensure compliance before the deadline.

What Happened? The transition to the year 2000 was largely uneventful. A few minor issues were reported, but there were no catastrophic failures. It wasn't that there was no reason to be concerned, but the successful outcome is often credited to the massive preventive effort rather than the fears being overblown.

The Y2K scare highlighted the importance of forward-thinking in software development and helped establish rigorous practices for handling date and time in computing.

If you want to start preparing or worrying now for the next similar scare, the Y2K38 Problem (Year 2038 Issue) arises from how older computer systems store time as a 32-bit integer, counting seconds since January 1, 1970 (Unix time). On January 19, 2038, this count will exceed the maximum value for a 32-bit integer, causing a rollover that could result in misinterpreted dates or system crashes.
Impact: This potentially affects embedded systems, infrastructure, and older software. Modern systems are increasingly being updated to 64-bit time representations, which kicks the problem far into the future.

Perceptron

Mark I Perceptron displayed at the Smithsonian museum

Perceptron as supposed to be a machine, but it began as an algorithm and is still used in that context. It is a piece of tech history from the early days of artificial intelligence (AI).

The perceptron algorithm was invented in 1957 at the Cornell Aeronautical Laboratory by Frank Rosenblatt, funded by the United States Office of Naval Research.

Its first implementation was in software for the IBM 704, and it was subsequently implemented in some custom-built hardware known as the "Mark 1 perceptron." It was one of the first artificial neural networks to be produced.

The Mark 1 perceptron was designed for image recognition and used 400 photocells connected to the "neurons." Still to this day, perceptrons are used generically as a basic neural network.

The algorithm's use was not without controversy. At a 1958 US Navy press conference, some of Rosenblatt's comments opened a controversy among the fledgling AI community. The New York Times picked up on this and reported that the perceptron would be "the embryo of an electronic computer that [the Navy] expects will be able to walk, talk, see, write, reproduce itself and be conscious of its existence."

This fear about "machine learning" was, and still is, feared by many people. Giving computers the ability to learn on their own without being explicitly programmed strikes some people as technology gone wild.

AI has met with ethical and technical hurdles. Steven Pinker has said that trying to duplicate the way the human brain works shows that for computers "Hard is easy. Easy is hard." Computers can do very complex calculations. They can calculate the depth of the ocean and project the outcomes of complicated experiments. But they have trouble answering a question such as "Can a shark play baseball?" which even a child could answer.

The Antikythera Mechanism: Astronomical calculator from 2000 Years Ago

a fragment of the Antikythera mechanism

The photo above doesn't look like any part of the history of computers. Do they date back a millennia to a time of analogue devices?

Some claim that the world’s oldest known “computer” is the Antikythera mechanism. This now badly-corroded bronze artifact was found in 1901 in the remains of a shipwreck near the Mediterranean island of Antikythera.

In the 1970s, using radiography, it was determined that the device is a complex mechanism with at least 30 gear wheels (some have certainly been lost).

It is more accurate to say that it is a mechanical calculator than computer because it can't be "reprogrammed."

That doesn't diminish that is is a very impressive Greek astronomical device for its time, which was around 100 BCE.

About the size of a shoe box, with dials front and back faces, it could be used to show using pointers where the sun and moon were in the sky, the current phase of the moon, the 19-year cycle of lunar months, the 18.2 year Saros cycle of lunar and solar eclipses and probably other celestial events.





The Antikythera Mechanism Research Project in Greece is still discovering inscriptions that tell what stars were just becoming visible at different times of year.

Of course, these things were known at the time of the device. The Greek astronomer and mathematician Geminos had written about these things, but to build a mechanism to do the calculatons was extraordinary.

The presumed date of the shipwreck is around 60 BC, and the location of the shipwreck seems to point to the mechanism originating on the island of Rhodes, from where there is a contemporary historic record from the writer Cicero of such devices.