9. The Joys of Codefucking
Kaitlyn discovers Dylan B. Goodman’s unique approach to programming, adds an equation for love, and introduces the result to several terabytes of poetry.
Transforming a dating app and a demo involving dancing baked beans into something you can show to a major client is clever. Yet creating a demo for a client is the easy bit. The tarmac connects with the tyres when you have to make it work in real life.
Luiz had airily imagined that EmpathyEngine could be quickly adapted to help confused travellers. In reality, the extra decision-making of a typical holiday, the wheres, the whens, the who withs and so on, involved more complex Artificial Intelligence capabilities. It turns out organising a decent holiday is a lot harder than choosing a partner for life. She’d got the system to come up with a grand vacation for Natalie and Martin, but replicating this in real-time, was proving problematic: far too often her beta version of TEA seemed to believe that everybody wanted to spend two days in the Auckland War Museum.
The problem went to the heart of the software. While the demo TEA they’d shown Natalie seem sophisticated, almost a Ferrari of the software world, under the bonnet it was more like a steam engine bolted onto a mountain bike: clunky, complicated and inherently unstable.
Not only had the core of all AMG’s software been written in an almost extinct programming language, but in its earlier days, AMG had been forced to rely on a number of eccentric — but cheap — freelance coders to create the bits that read human emotional response. Under-paid and under-supervised, these had added in a variety of modules created using cutting-edge techniques. So cutting-edge in fact, that most serious programmers have since cut them. To put it in technical terms, the demo version of TEA resembled a beautifully shot movie in which half the dialogue is in Ancient Greek, and the other half in Klingon.
Kaitlyn had, of course, a variety of AI tools to help her speed up her work. Sadly, few of them worked with this noodle soup of programming. Fixing up the legacy of other peoples stupidity was not how Kaitlyn wanted to spend her time, and she was finding the task of fixing TEA frustrating.
But then she discovered the delightfully named “Codefucker”, the invention of Dylan B. Goodman of Atlanta Georgia, USA.
Dylan was a thirty-two-year-old man who — much to their disgust — still lived with his parents, only leaving his bedroom when his supply of Doritos was running low. Aside from computers, his main interests in life were the Confederacy (‘the South WILL rise again!’), which brand of BBQ sauce is REALLY best, and the various judeo-muslim-feminist-communist conspiracies against the USA.
Probably, had Kaitlyn met him in the flesh, she’d have felt a strong desire to offer violence to his testicles. But, as she reviewed Codefucker, she felt a surprising level of admiration. The man was a genius.
Despite this cleverness, Dylan had held down few jobs. Every few months though, he did manage to pick up the odd piece of freelance work, rewriting and improving old database and logistic control programmes. Like Kaitlyn he found this kind of stuff tedious, so he decided to create an automated reprogramming solution all of his own, utilising a complex experimentation process that involved various AI models querying and testing each other, until - like gladiators in a digital coliseum - one of them emerged with a victorious solution to whatever task Dylan had set it.
The grand thing about this system, as Dylan saw it, was that its processes were so convoluted and obscure, so uncontrolled by any of the QC and legal restraints that the R&D folk at Google or Apple have to endure, that it was almost impossible for anybody to trace that much of what it did was simply to steal and refine answers from other people’s AI systems.
Dylan tried to interest the likes of Microsoft and Oracle in his work, but his long emails, fervent with descriptions of his brilliance, had disappeared into the ether, likely bounced by discerning corporate spam filters.
In revenge against the evil corporates that spurned him therefore, Dylan had released his “Codefucking” system on various dark-web sites.
His hopeful dream was that like-minded individuals would see the potential inherent in his software and use it to create self-sustaining, self-repairing viruses that would overwhelm the defences of the big corporates, and usher in a libertarian paradise where Doritos were free and women with big boobs were irresistibly attracted to fat pale racists with poor personal hygiene.
Fortunately for democracy, free-enterprise and women with larger sized breasts, Dylan — while a programming genius — was in many other ways quite the dumbo.
Although wanting his software system to sweep the world, he simultaneously believed his genius should only be appreciated by the truly gifted. He found it necessary therefore, to protect his programme from exploitation by those whose weaker minds could not comprehend its mystic beauty.
The technical aspects of how he did this, we need not go into, merely mentioning that his input system only accepted instructions that were written right to left, and, foregoing the numbering system in common use since the fall of the Roman Empire, required all numbers to be entered in binary format.
Some of the geekier frame of mind found Codefucker’s eccentricity endearing – a challenge to be enjoyed and appreciated. However, this was a distinctly minority cult, and like their guru, most of the adherents lived with their parents, were addicted to junk food, and had problems holding down regular employment. Certainly, none had access to commercial grade software development facilities and a customised Artificial Intelligence system.
Then along came Kaitlyn. She regarded using binary numbers and entering queries in an Arabic direction as at worst a frustrating irritant. The big picture, as she saw it, was here was a way to get her new tourism system to refine itself, to self-evolve each time a new problem was encountered. Even better, it was free, and unencumbered by tedious licensing requirements.
Never again would she have to deal directly with any of Martin’s and Joeline’s original code; Codefucker would fix everything. Hallelujah!
So, away she went. It’s amazing what a diet of energy drinks and Mr Whittaker’s finest chocolate-peanut slabs can do for a young woman of Kaitlyn’s spirit and determination. Within a relatively modest number of sleepless nights, Kaitlyn had linked Codefucker’s capabilities with the artificial intelligence system that underpinned TEA.
This was, of course, more complex than the last sentence makes it sound. But now when the AI system generated an error, it sent the affected modules to Codefucker, which experimented with them by writing and rewriting the programmes underlying parameters and rules until they worked more “effectively”.
Several snags emerged. Firstly, as Codefucker tested a million possible solutions for each problem, it drained off huge amounts of computing power, overheating CPUs, slowing routine jobs to a standstill, and causing AMG’s lights to flicker oddly. Martin and Joeline’s eyebrows also flickered as electricity and computer maintenance costs soared to unprecedented levels and productivity dipped.
Kaitlyn, who was trying to hide what she was doing until she could make a triumphant announcement of her genius, managed to plaster over the issues, inventing a range of complex and marginally plausible excuses.
She was lucky she worked for AMG. The head of IT at the Ministry of Internal Affairs, for instance, might have been more concerned at her attitude to risk management and the desirability of careful software testing. But Martin and Joeline favoured an experimental approach and were prepared to indulge her in return for ‘big ideas’. And certainly no one anywhere could say Kaitlyn wasn’t capable of thinking outside the fabled ‘box’. When it came to ‘pushing the envelope’ and so forth, she was world class.
The result was that TEA, a system initially intended to interpret your deepest travel desires, was majorly modified: when the AI got stuck on a task, it sent a message to Codefucker telling it to experiment away, fiddling around with the AI’s deepest code, until it produced some approximation of whatever it thought Kaitlyn’s objectives required.
TEA, of course, had nowhere near the power of the big AI systems of Microsoft and so on. But it had far more freedom to self-experiment than those behmouths, and now it had the acquired ability to evolve fast.
Effectively, Kaitlyn had put a “wellness self-help” module in the AI system’s ‘head’. You know, like meditation or a large dose of LSD can help you rewire your brain and see the world in a new light, TEA could now similarly rewire itself.
No risk there really, Kaitlyn decided. As long as she was the one to explain what outcomes were desired and kept them simple and explicit, all would be fine. Her new version of TEA would do exactly what she required.
Despite discovering the joys of Codefucker though, Kaitlyn was not content.
Too often her life seemed but a dreary commute between crises at work and crises at home. Keeping her lovers from running into each other and inventing excuses to account for her oddball timetable was becoming taxing.
This had come to a head a couple of days ago, when, while they were enjoying a quiet cuddle, an opportunistic burglar had broken in through the front window of Analise’s flat and quickly grabbed a few portable valuables. These included Kaitlyn’s recently acquired, and expensive phone.
Such events are disturbing enough in themselves, but with all the immediate kerfuffle, and suddenly losing her access to her private calendar, she became confused about the who, where’s and whens of upcoming rendezvouses. The immediate result of this being that she came within minutes of taking Analise to the very bar where Cole was shortly due to meet her. Hurried phone calls, and a series of embarrassingly thin excuses, managed to stave off this disaster, but it left her with two annoyed sweethearts, and a strong feeling that she couldn’t go on like this for much longer.
The next afternoon therefore, as everybody else was starting to leave the office, her fertile imagination was drawn from Tourism New Zealand’s requirements towards more personal problems. Why … Why … her flittering brain kept demanding of her, does EmpathyEngine find it so hard to choose between my lovers?
Then, suddenly, a burst of hot insight washed over her, and like Archimedes in the bath, she cried ‘Eureka!’
It was sooooo obvious!
If the decision-models that Codefucker was experimenting with could help TEA choose more emotionally satisfying holiday destinations, maybe they could also give the EmpathyEngine app a super-boost.
She’d get Codefucker to bugger about with Dr Spoon’s formula for love until it arrived at a massively improved algorithm for choosing between two potential partners. This would not only improve the app that was now AMG’s big money earner, but more relevantly to her, EmpathyEngine would be able to reach a definite, Cole vs Analise, conclusion.
That, she thought sadly, would mean she’d have to let one of her darling’s down. But at least she’d avoid having ulcers caused by her current guilt, and — when they saw the impeccable logic of the algorithm — probably whoever lost the contest would understand.
Her brain raced onwards. Such a massively improved EmpathyEngine could also be integrated with TEA to enhance its operations. TEA, as she’d shown Natalie, could already provide a couple with an itinerary for the holiday of a lifetime. But what if you weren’t sure who you wanted to couple up with? With her enhancements, you could identify the ideal partner for trip around Godzone. A win for Tourism Aotearoa, and for AMG, and so also for Kaitlyn. Brilliant!
Sadly however, as afternoon turned to evening, and Codefucker ran myriad different permutations of the love formula, it became clear that Dr Spoon’s algorithm was limited. Essentially it measured ‘compatibility’ between people. You know, what you have in common, not that secret sauce that is the key to lasting passion.
It needed a better definition of ‘ love’.
Time ticked, Kaitlyn tired, and faced with a tedious impasse, she decided on a creative shortcut.
Feeling that her darling Analise must surely be onto something with all the mushy poetry she kept sharing, she tasked the computer to search the Internet for poems of love, and then set some of AMG’s most sophisticated linguistic analysis modules to uncover the main themes to be found in these. Here, she hoped, would be found the emotional input needed to define what constituted ‘true love’.
Soon several terabytes of poetry were stored, and the machine’s analysis engines buzzed away.
Combing thousands of poems to collate and define the elements of true love and relating the results to a choice algorithm designed for holiday destinations in New Zealand is an unusual computational task, and not entirely simple.
As Kaitlyn played with the complex instructions required to direct the AI in the right direction, mistakes crept into her increasingly tired brain. It was almost 1 am, she was supposed to be at a meeting at 8:30, and had a ton of ‘normal’ work to do tomorrow. Reluctantly then, Kaitlyn, decided she’d better head for bed.
She shut off the test machine’s access to the Internet, and its connection to the main server. But, on a whim, left her experimental version of TEA running. She had a mild hope that, with the help of the wonderful Codefucker module, the AI would experiment away on its own. The poetry data was likely way too complex, and the preliminary objectives she’d entered too half-arsed, she realised. But, hey, no harm in letting the computer play with them overnight, you never know, it might come up with something.
And it did.
Subscribe to get reminders and the next episode of Lovebirds delivered direct to your inbox…
PREVIOUS EPISODE:
Interlude #3: Putting Men on The BBQ
The sex worker who’d reduced Martin to jelly all that time ago, was in fact not called Lilly, that being merely her Club 88 moniker.
READ THE NEXT EPISODE:
10. A Tryst in Eastbourne
Martin, unaware of Kaitlyn’s interesting experiments, sat in his office that Friday under the impression that the biggest problem in his life was whether he should mention his “date” with a client to Joeline.