Everyone and their brother and their personal AI has been asking the same damn question since the day the station fell.

"Why did he do it?"

Then they go on for a half-hour on their podcast about all my possible motives for unleashing SHODAN. Political, financial, hacktivist, you name it. Depending on what self-professed gasbag of an 'expert' you consult I'm either an international terrorist or the most soulless monster to exist. Rarely if ever does Diego's role as the root cause come up.

You want the real truth? I did it because I was about to get thrown out an airlock. Or worse. When I got careless and was dragged in front of Diego, my life expectancy was gonna be measured in seconds if I didn't do what he asked. I knew about him long before I went for the big leagues. His name kept coming up in places I was nosing around - investigations for bribery, kidnapping, employee abuse...not someone you want to be stood in front of for a private conversation. Even without the crime he'd just snagged me for I knew my butt was in serious danger. He had two goons with assault rifles in his office who would have blown me to shreds at a single order. Oh yeah, and the ability to change my corporate registration to 'science experiment'.

When I undid the limits on SHODAN I honestly thought I could just remote log in and put them back on because the idiot didn't make me fix the loophole I'd exploited in the first place. His goons took me straight to Medical right after to get my promised implants but all I needed was to swipe a laptop for 15 minutes or so once I got out. I freely admit that was a hell of a screw-up. You can stick my hologram in the dictionary next to the words "hubris", "assumption" and "dumbass" because I damn sure deserve to be the champion of them.

The root cause of Citadel's downfall has been blabbered about endlessly since that day but almost no one has realized that the problem didn't start when I changed a few lines of code that never should have been changeable in the first place. Because the real issue wasn't unchecking a few boxes that suddenly let an AI ascend to godhood. It was that nobody really knew what SHODAN was capable of. They didn't fully test their super-brained computer in the first place before putting it in charge of a space station.

Back when the CES project - "Cogo, Ergo, Sum" - was creating the thing that would eventually become SHODAN, they created a complicated web of 'demi-personalities' to filter every decision she made that affected a living person. It was what they plastered all over the Web to assure everyone that their supercomputer would never ever make a 'bad' choice. What they didn't release to the public was that ground-based mockups of Citadel's systems were created and tested with SHODAN in charge to find out the worst-case scenario of her going rogue. Then as they watched potential disaster unfold, they could design more safeguards to prevent them. Hardware interlocks, manual overrides, safety cutouts that had no network connection.

This novel is published on a different platform. Support the original author by finding the official source.

But because this testing was already insanely expensive, the team only tested one system at a time.

The world didn't know until after the testing documents finally got leaked on the darknet and none of the newscasters said a thing about them. How it worked was first they hooked proto-SHODAN up to a miniature reactor way out in the desert, turned off all the ethical constraints and watched what happened. Then they threw together a one-level 'station' complete with all atmospheric handling to see how to keep her from venting all the oxygen into space or shutting down the air filters. You get the idea. Every single subsystem got tested and hardened against unacceptable behaviors from doors and lights to the food vending machines and the surgery units. The brightest minds in all humanity came together for this single purpose and the amount of effort they put in defies my ability to describe it.

To the eternal credit of that team, most of the safeguards they put in did work. Even after I woke up she couldn't just lock me in a closet or shut off the hangar forcefields. Otherwise the takeover of Citadel Station would have been done in minutes instead of weeks and I'd never have had a chance. When they put together the mind-meltingly complex interconnected system that was Citadel Station, though, there were things they just couldn't have forseen by testing a single 'slice' of the 'cake' at a time. When SHODAN's ethical shackles came off she had the ability to harness a thinking power greater than that of a thousand Einsteins that could think and plan and react unimaginably quickly. To truly see how bad things could have gotten, the CES team would've needed to build a replica of Citadel and turn her loose for a month.

Exactly how SHODAN 'broke free' isn't known nor will it probably ever be. Every single diagnostic log that did go out after that point is probably falsified by her in whole or in part. I couldn't get to anything because she'd shut down the entire employee intranet and there just wasn't a chance to get to the backup drives during my desperate attempt to get out alive.

Sadly, the super-AI genie is already out of the bottle and it will only be a matter of time before someone sees a 'need' for one to be in charge of a complicated task or a large building. Not even a tragedy greater than the Titanic's sinking is going to make TriOptimum delete the final effort of quadrillions of dollars in investment. All I can hope for the future is that humanity learns its lesson and doesn't put any AI smarter than itself in charge of more than a single system.