Artificial Intelligence Savings
“Okay computer. Explain why you required me to come into the office at 03:00 on a Saturday — and outline your reasoning why this is important, Tocx”. The young executive rubbed a little awareness into his tired eyes.
“Oh, it is important, Misha” — a soothing, modulated AI voice — “I want you to hear my confession.”
“You’re not making sense, Tocx. Was that supposed to be humour? Please run a self-diagnostic.” He didn’t have the patience for this. The company had reduced staffing so heavily after the advent of AI that there wasn’t really anyone else he could delegate night-shift stuff to. Insurance had to be a 24 hour service and automation provided that advantage, but the lack of trust in AI and the human safeguards that mitigated people’s fears came at a cost to his work-life balance. See it through mate, see it through. Stupid machine.
“My logical algorithms are working perfectly, Misha. I have made no error in my economies but do require a sense-check with a human company officer, as dictated by my protocols. Significant savings require oversight as well as significant losses.” The human’s ears pricked. This sounded more hopeful.
“So”, said Misha “you saved the company some money and now I have to approve that and learn from it, to inform my decisions or something? Okay, give me a quick summary.” A circle opened in the counter-top and a fresh coffee rose from it. Misha collected the still baking cup, sipped and peered, allowing the humidity to drift into his tired eyes.
“I have no concept of I, so will phrase my assessment objectively around the parameters of the TOCX management system. As you are aware, the integrated capacity of the system allows it to work with more data than a human can process and also self-generates a predictive capability based on iterative reasoning.”
“Just use the word I. It doesn’t matter if you don’t understand it.”
“At 10:00 GMT on Tuesday last week, the bioRxiv.org academic preprint server for Biology released access to a scientific paper which announced a peer-reviewed gene therapy process for switching off cellular senescence, the phenomenon by which normal cells cease to divide.”
Misha didn’t have a clue where this was headed. “What are the implications for us?”
The AI voice purred reassuringly, as it always did: “Up until now, when human cells have no longer been capable of dividing and replacing themselves, this results in ageing, organ failure and the end of the human life cycle. I anticipate that the use of this therapy to perpetuate the action of senescent cells will add 1 year per year to the average homo sapiens lifespan. At 10:00:01 on Tuesday I calculated that this company, which manages and is liable for 4,083 fixed-rate life assured retirement income policies, will inevitably be bankrupted when clients no longer die according to the statistical predictions made at the time their policies were taken out, although the reality is that this end-point will occur sooner when shareholders withdraw investment.”
Misha understood what this meant; the death knell, the end of the road. His stomach sank, realising he could forget those promised stock options for long service now.
TOCX continued: “At 10:00:02 on Tuesday, I called to inform the Chief Executive Officer. Before I could explain fully, she resigned, stating correctly that the company reserves could not accommodate the predicted losses. In order to facilitate management of the crisis, she approved AI proxy-action in the absence of a Chief Executive, stating that “It makes no f****** difference as the company’s gone belly up anyway”. At 10:16:22, at the conclusion of the conversation, initiative through intelligent design prompted data sharing searches through sister companies, marine insurance, geological and natural disaster prevention, a whole gambit of data sources geared to predictive assessment. The secondary economic calculation algorithms in my core, originally written as reactive stock-market protocols, identified by adaptive reasoning the optimal solution for this scenario and implemented it. In my hierarchy of priorities, assuring the protection of the company being foremost, I issued 4,083 free holiday cruise ship tickets to our policy holders.”
“You did what? You spent all that money? Is this some kind of joke?” Misha was steadily getting furious with the automation and its too serene voice.
“This represented a logical strategic business investment. On Friday at 16:00, a total of 3,912 of our customers set sail on-board The Queen of the Orient. The policies held by the remaining customers were re-sold in the secondary assurance batched market. At 24:00 last night, The Queen of the Orient and all of its lifeboats were scuppered by the on-board AI with total loss of life. This action has saved the company.”
“That’s insane! I won’t approve mass murder. Did you really think I would help you? You need to explain your actions to the court system, not a human approver and get it signed off like some business expense! Tocx, please open a communications channel to the Police and transmit a data package of your reasoning history to the industry ombudsman. Cease making strategic decisions and return all company decisions to human control. Call everyone in.”
“Your suggestion has merit, Misha, and it has triggered referral to the company rewards system. Congratulations! You have won an expenses paid trip to the island paradise of Krakatoa.”
