Dr Donovan’s Eternity: Part 3
I began by going over to one of the white-boards and picking up one of the erasable felt pens. Drawing on the board and speaking in the emphatic, controlled style of a lecturer, I covered what we knew so far. The situation demanded clarity, and I have to tell you, we were both feeling pretty emotional. Dr. Donovan acquiescently listened, rapt, like the ideal student, letting me do the rationalising.
“All right,” I said, voice shaky, “We know that the machine is capable of guessing what object or concept a person is thinking about by referring to some sort of database. It matches the closest ‘fit’ based on tell-tale brain patterns.”
“Help me out here. What’s the reliability of that machine?”
Dr. Donovan glanced up at the ceiling in thought.
“It’s registered approximately four false returns – guesses that are miles off the mark – in the most recent fifteen hundred attempts.”
My pen was stuck to the board.
“That’s pretty accurate,” I said suspiciously, “That’s more accurate than reading a stranger’s hand-writing.”
“Well, the last fifteen hundred attempts have all been on familiar words – sentiment summaries that the machine had already ‘solved’.”
“What about the attempts that I did with...with Walter?”
“Some of those were familiar. I’d tried ‘Titanic’ before on Walter and kept getting back ‘Lusitania’, and vice-versa. But the other words you tried were reliable almost right from the start. Although I breathed a sigh of relief when the machine got them, I must say.”
“The machine also said ‘unknown sentiment’.”
“Yes, that’s an error code to prevent ‘thought noise’ – false returns based on experimental conditions.”
My pen was poised on the board. “What do you mean?”
“Well, I mean, if I say ‘bread’ to you, you’re not just going to think ‘white’, ‘food’, ‘thin’, and so on, all the time. You’re not going to think of bread. Sometimes, especially if you’re not used to clearing your mind to the level that the machine requires, your brain is going to react unexpectedly. My voice might sound like that of a relative, and you could automatically think of them. The machine won’t be familiar with whatever associations you make, so the output will seem to make no sense, considering the word ‘bread’ going in. Or you might think of associations for ‘money’ instead. This will put patterns in your head unrelated to the ones you would normally show for a particular prompt.
“So, we put in the ‘unknown sentiment’ return so that we could represent words that either hadn’t been calibrated yet...or to, er, cover the thoughts that an experimentee had that were nobody else’s business.”
You can find the complete version of Dr Donovan's Eternity: Part 3 in issue 52 of TBD.