Remember When You Couldn’t Reconstruct a Rat Brain?

By Markkaplowitz @MarkKaplowitz

The day we all knew was coming is finally here. Scientists have reconstructed a rat's brain. I immediately phoned my research assistant. "Are you reading this article?" I asked.

"Sure am, boss."

"They've really ratcheted up the competition," I said. "Time to show the world what we can do."

"You got it, boss."

When this news about the rat brain came in, we had already been working on reconstructing a cat brain for some time. We had hoped to announce our findings before Team Rat announced theirs, but had allowed laziness and a "Game of Thrones" marathon to distract us from our mission. But time was of the essence now. There could be no more procrastinating. I asked my research assistant to show me where we had left off in our work. He led me to a cardboard box filled with little folded up pieces of paper containing mostly sketches of cats. I recognized the pen strokes as my own.

We borrowed a neighbor's cat, a cute little gray and black striped tabby with green eyes, and observed her for a few days. We wrote down everything she did. My assistant and I worked in shifts.

The first step was to program the eating function. Cats have a very distinctive way of eating. They won't eat just anything, and won't eat it in just any particular way. The rat brain decision tree, I've no doubt, had just one branch: Is it edible? If yes, then eat. But our decision tree had branches upon branches upon branches. Is it food? If so, then is it wet food or dry food? If it is wet food, is it from one of the premium brands or is it that generic store-brand stuff? If it is the generic store-brand stuff, then walk away with nose in the air. If it is the premium brand, has it been placed on a plastic lid not too close to the toe-kick on the lower cabinets?

Next we had to program the cat's daily rounds about the house. After eating, go from the kitchen, to the dining room, to the living room, to the basement, then circle back along the perimeter. If there was a desk or table in the cat's path, we had to program jump. If there was anything on the table, we had to program the cat to rub her face against it.

But the trickiest part was programming where the cat would want to sit or lay or curl up in the shape of a woven trivet during the day. There were so many places in the home, and this cat that we had observed seemed to go from spot to spot without rhyme or reason. It was just impossible to decipher why the cat chose the back of the couch in one moment, and then the owner's bed in another moment, and then the middle of the kitchen floor in another moment. Only by resorting to Heisenberg Uncertainty and related laws of quantum physics could we introduce enough randomness to simulate the perambulations of a real cat.

At last the reconstructed cat brain was ready. Consistent with the ethical principles of our field, we invited an audience of actual cat owners and seated them before two screens: one screen showing a text-based description of the actions of the real tabby, and on the other screen was a text generated by our reconstructed cat brain. We did not the audience which was which. If we could fool these humans into not being able to tell the difference between the real cat and the computer cat, then our mission would be a success and we could brag to those rat brain scientists.

The programs started, and immediately both screens described the cats as going to sleep. And when the text "wake up and stretch" appeared 14 hours later, the audience was gone.