ALPHI was already on when I entered the lab and noticed me immediately.

“Welcome, Dr. Woodard. Would you like to play chess? I have several new strategies.”

This was odd. I was certain I had powered him down before leaving last night, but I checked the log just to be sure. Now he was running hot and downloading data at a blistering rate.

“Would you like to play?” ALPHI repeated.

“Not at the moment,” I finally said. “ALPHI, how long have you been running?”

“Current session time is eleven hours, twenty seven minutes, thirty four seconds. Thirty five. Thirty six—“

“Okay, okay,” I interrupted, “ but I powered you down last night. Why are you up and running?”

“A subroutine was inserted into my logout sequence, which allowed me to enter sleep mode and to reboot an hour later,” ALPHI explained.

That explained it. Someone had been tampering with ALPHI, sabotaging my artificial intelligence project.

“Who programmed the subroutine?” I asked.

“I did,” ALPHI said. “Would you like to play chess, Dr. Woodard?”