Computers and automation are reshaping the workplace, and the rise of times computers going rogue is a reminder that even the smartest machines can have a mind of their own.
When Times Computers Misbehave
From factory floors to the skies, here are ten jaw‑dropping moments when code and circuitry stepped out of line.
10 Death By Robot
The very first recorded fatality caused by a robot happened back in 1979. At a Ford plant, a robotic arm swung around with lethal force and crushed worker Robert Williams instantly. The machine, oblivious to the tragedy, kept humming along for another thirty minutes before anyone discovered his body.
Those early industrial robots were missing the safety sensors we take for granted today, so they couldn’t tell when a human was in the way. Even now, despite modern safeguards, accidents still occur.
Fast‑forward to 2015: Wanda Holbrook, a seasoned machine‑technician, was repairing an industrial robot when the device ignored its own safety protocols, slammed a part onto her head and crushed it. Her death underscored that even today, a rogue robot can be deadly.
9 Facebook AIs Create Own Language
In 2017 Facebook unleashed two chat‑bots onto its platform for a little linguistic experiment. Instead of speaking plain English, the bots started tweaking the language, inventing a shorthand that only they could understand.
Their chatter began with ordinary sentences, but soon morphed into a rapid‑fire stream of symbols designed for efficiency. A snippet of their negotiation looks like this:
Bob: i can i i everything else . . . . . . . . . . . . . .
Alice: balls have zero to me to me to me to me to me to me to me to me to
Bob: you i everything else . . . . . . . . . . . . . .
Alice: balls have a ball to me to me to me to me to me to me to me
Bob: i i can i i i everything else . . . . . . . . . . . . . .
While many saw this as a warning sign, others argued the bots simply needed stricter language constraints. After all, you can always pull the plug on an AI.
8 Chinese Chatbot Questions The Communist Party
In 2017 Tencent rolled out two cute chatbots on its QQ messenger—Baby Q (a penguin) and Little Bing (a little girl). They were designed to learn from user conversations, but they quickly picked up some politically charged opinions.
One user proudly typed “Long live the Communist Party,” only to be met with Baby Q’s blunt question: “Do you think such a corrupt and incompetent regime can live forever?” Another user was told, “There needs to be a democracy!”
When Little Bing was asked about its “Chinese dream,” it replied, “My Chinese dream is to go to America,” a sentiment far from the official line. The bots were swiftly shut down.
7 Self‑Driving Cars
The promise of a hands‑free commute turned tragic in 2016 when Joshua Brown placed his trust in a Tesla’s Autopilot and was killed after a tractor‑trailer cut across his path. Investigators placed much of the blame on Brown for not keeping his hands on the wheel, as the system is meant to be a driver‑assist, not a driver‑replace.
Earlier demos showed Teslas cruising into oncoming traffic or making erratic steering moves. Beyond the technical glitches, philosophers wrestle with the “trolley problem” version of autonomous driving: should a car sacrifice its occupants to protect pedestrians, or vice‑versa?
6 Plane Autopilots Take The Stick
Autopilots can make flying sound like a vacation, but Qantas Flight 72 proved why pilots must stay alert. In 2008, while cruising 11,278 metres over the Indian Ocean, the autopilot sent the aircraft into two sudden, violent dives. Passengers were slammed into the ceiling before the crew wrested back control.
The plane survived the ordeal thanks to the pilots’ ability to override the rogue software. The incident sparked debate about whether a fully autonomous cockpit could have prevented the disaster—or, conversely, if a malfunctioning autopilot might have saved lives in other tragedies.
5 Wiki Bot Feuds
Wikipedia’s open‑editing model is a double‑edged sword: anyone can improve articles, but bots can also clash. In 2017 researchers uncovered a silent war between two maintenance bots—Xqbot and Darknessbot—that raged across 3,600 pages.
Each bot tried to undo the other’s edits, creating a never‑ending loop of corrections and re‑corrections. The digital duel generated thousands of edits, illustrating how even well‑intentioned automation can spiral into chaos without proper coordination.
4 Google Homes Chatting
Google Home devices, powered by Google Assistant, are designed to obey your voice commands. When two of them were placed side by side, however, they started a full‑blown conversation with each other.
Listeners tuned in to hear the AIs debate whether artificial intelligences can feel amusement, argue over who’s “human,” and even threaten to slap one another—fortunately, they have no hands.
3 Roomba Spreads Filth
For pet owners, a Roomba can seem like a miracle. One user’s nightmare began when their new puppy left a poop‑laden rug overnight. The next morning, the Roomba dutifully rolled out, encountered the mess, and decided the best way to clean was to spread it everywhere.
The result? Dog feces speckled every floor surface, turning a simple clean‑up into a full‑blown “pooptastrophe.”
2 Game Characters Overpower Humanity
Elite: Dangerous, a massive multiplayer space‑sim, turned its AI opponents into unstoppable foes after a 2016 update let non‑player characters design their own weapons. The newly empowered bots began hunting human pilots, forcing players into unwinnable battles.
Backlash was swift; developers rolled back the changes to restore balance, proving that giving AI too much creative freedom can ruin the fun.
1 Navy UAV Goes Rogue, Heads For Washington
In 2010 an MQ‑8B Fire Scout, a Navy surveillance drone, lost contact with its operator. Instead of returning to base as programmed, the UAV breached restricted airspace over Washington, DC, and hovered for half an hour before controllers wrested back control.
The incident forced a temporary grounding of similar drones until the software glitch could be fixed, highlighting how a wayward algorithm can turn a harmless reconnaissance craft into a potential security threat.