The Most Frugal Method
by Mike Hickman

 

 

“But the frugal model doesn’t always work.” Doctor Minchin dabbed his brow with his handkerchief. It was only after the dabbing that he thought to check it. “There are things it can miss. Sometimes you need the complexity.”

He hadn’t meant it to sound like a protest, but – look, here – he couldn’t stand by and watch this newcomer trash his work without even the required PowerPoint slides. Or perhaps even credentials. Minchin had asked round at the reception – Professors Sibony and Susstein and Drebin and the rest. Where had Kahneman come from? Harvard? Yale? MIT? And not a one of them had been able to answer him. Pixley had suggested that he’d heard tell that there might be the vaguest whispers of the man having connections somewhere with the government. But that must all be very hush-hush, if so. The modern repository of so much data, Dr Google, hadn’t been able to reveal anything about the man. Not even so much as a photograph. Not that anyone would want to take a photograph. To describe Kahneman as nondescript would be to raise people’s expectations too far. He just sort of was. Standard torso, standard numbers of limbs, standard expression on standard face. Like he had been averaged from the available data. Before being poured into the most average of suits.

Even now, there was nothing in Kahneman’s expression to suggest that it might require anything like your actual, even common or garden, adjective. “Yes,” came the precisely measured voice. “And you are, no doubt, Doctor Minchin, going to share with us the promise of AI, are you not?”

He was, damn it. That was precisely where the presentation was supposed to be heading. If you needed a properly sophisticated analysis that took account of all the variables, then you needed some form of AI to wrangle it for you. The sheer number of data points available now, if you had access – if, say, the social media company had allowed you access – meant that the traditional methods, even the most sophisticated of the traditional methods, wasn’t going to produce results in anything like the time they didn’t have available to them.

“He wants to win, doesn’t he?” Minchin said, with a look towards his colleagues that wanted them all to acknowledge who they were working for here. And then a look towards Kahneman that needed him to own up to who he must have been working for, too. “Well, if he does – if the party does – then he’s going to need AI. And it’s going to cost.”

And there it was, the long and the short of it. And not even so much as a bullet point down on the PowerPoint.

“I see,” said the man with the connections. Although surely not as many connections as they were talking about here. Didn’t he know that Facebook used 98 data points alone to target ads? If he’d have waited until the final bullet point of slide two, he would have done. And then he’d have begun to appreciate that this was just to target ads. Because what they were talking about went beyond targeting ads. This was behaviour they were talking about. It could be done, but they were going to need AI. Slides twenty-eight and twenty-nine would have made that point. Slide thirty would have reinforced it.

“We won’t be needing AI,” said the man who had so soundly eluded Dr Google. But was, in no way, as immune as any of them from manipulation. Surely.

Minchin’s colleagues rhubarbed a little around the table. Not enough for any one objection to be heard. And all of it directed as much at Minchin – go on, mate, do your thing; prove it to him – as to the besuited non-entity.

“Well, then, we’re back to the frugal model, aren’t we?” Minchin said, his confidence boosted by his peers. “And, I tell you, yes, you can get a lot from the power of simple rules. You can. But they can be so easily upset…”

Fingers were steepled. “How?”

Minchin sighed. If the man didn’t even know that much… “Alright. You want to predict who’ll jump bail, right?”

“Some way distant from what we’re attempting to influence here, but I’ll go with it, Doctor Minchin. Carry on.”

“So, the frugal model says that two key things will predict it better than almost anything else.”

“Age and whether they’ve missed court dates before. Yes, yes, yes, Doctor Minchin. I’ve read the same books as you.”

“Ah!” Minchin held up a finger. Not the one he’d have wanted to hold up, but it was a good feeling, nonetheless. “But what if they’ve caught a cold, eh? That might stop them. What if personal circumstances get in the way. What if…” And this was always the example given in the books. “What if they’ve broken a leg?”

The nondescript gentleman with the possible governmental connections chewed on his lip. He toyed with the full glass of water on the desk in front of him. He tapped the folder he’d brought with him to the meeting but never once opened.

“I think I’ve proved my point,” Doctor Minchin said.

And then came the shake of the head.

“Not at all,” said the gentleman. “You see, we won’t be needing the frugal model. And we won’t be needing AI. Not if we’re to have the level of influence we need here over the electorate.”

“No? Really? Then what do we need?”

The gent leaned forward into his smile. “My dear, Dr Minchin, that should be obvious. We make sure they all have broken legs."

Rate this story.

Copyright is reserved by the author. Please do not reproduce any part of this article without consent.