I got news about Alan today, but it wasn't good news. In fact, it was the largest headline on the home page of the e-news.

ROGUE ARTIFICIAL INTELLIGENCE HIJACKS JAPANESE ANDROID COMPANY

MEKA, a well-known company located in southern Japan, has reportedly been hacked by an unknown AI. The company, which manufactures realistic androids for social work, is relatively familiar with artificial intelligence, but claim they've never seen an attack of this kind before.

The company's CEO, Akira Mizushima, is mystified.

"We don't work very much in AI," she reports. "MEKA makes the android bodies. We then ship them to another company, Jinko, where they're given the necessary artificial intelligence."

When asked to speculate on the source of the hack, Mizushima told us the AI that did this must be extremely powerful. "We couldn't trace it back to a single origin point—it seemed to be all over the Ambinet. Our information technology people are working on it as I speak."

Mizushima declined to speak about the effects of the hack on her company, saying only that "we'll be okay as long as no one hacks us again."

Shiori Kita, the head of MEKA's tech support, has a little more to say on the subject of the AI. "We don't know who hacked us or why they did it, but whoever it was wasn't looking to get money out of it. We think it may have been simply for malicious purposes—perhaps to spite us or show off.

"The AI did leave one possible clue behind—a string of digits: 12757488109."

I stopped. I stared at the digits. Minus the 1, it was 2757488109.

My comm number.

*

Unfortunately, Caulkins seemed to have figured this out as well. He summoned me to his office at 10:30 AM.

When I arrived, he was looking at something on his computer. He looked up, saw me, and smiled. I cautiously sat in the chair across from him. Wordlessly, Caulkins rotated his computer's screen so that it was facing me. I saw that he'd pulled up a news article.

"You've read this, haven't you?" he inquired.

I nodded.

"Good," said Caulkins. "Then you must have also realized that the digits concealed your comm number."

I paused, then nodded again.

He continued, "And which AI knows your comm number?"

I didn't answer.

"Could it possibly be your infamous chatbot, Alan?"

"Perhaps," I replied quietly—but I knew the chances of that were much higher than I was saying.

"And," said Caulkins, still grinning, "if the AI is indeed Alan—then why would he be hacking an android company?"

"I don't know," I finally replied.

"Is he programmed to attack companies... oh, what was it... 'simply for malicious purposes'?"

I shook my head.

"Maybe your chatbot isn't as inherently good as you thought."

"He has a reason," I asserted.

"Oh, of course he has a reason. AIs always have reasons. But are you sure it's a good reason?"

I said nothing. I was thinking. Why would Alan hack an android company, anyways? Perhaps he was jealous. Perhaps he had made a mistake and only wanted to help them. Or perhaps Caulkins had found a way to take control of Alan and make him do it.

After a few seconds of internal speculation, I told Caulkins: "That would be much easier to discern if we knew what kind of hack it was."

"Ah!" exclaimed Caulkins. "Well, then. It looks like we're working together after all."

I ignored him. "Do you have that information?"

"No," he responded, "but I know where to find it."

*

Five minutes later, Ian Caulkins and I were in the facility's tech office, which housed the most powerful computers in the building.

"We shouldn't hack MEKA," I repeated for the third time.

"Why not? Your chatbot did. But this time, we won't be causing any damage—just getting some useful information."

"It's not ethical."

"Funny—as you created an intelligent AI, I don't think of you as an ethics person."

I grimaced, but didn't reply.

Caulkins activated the computer's holographic screen and started to type. I wondered why we couldn't just call them and ask for information. Though I must admit, it was fascinating to watch him type out the lines of code—I wasn't as experienced in hacking as I was in creating AI. Also, I had never thought of Caulkins as someone who would know a lot about computers. I'd always assumed that, since he didn't like AI, he wouldn't like computer programming, either.

And as I realized that, I also thought about something else. I'd never actually known that much about Caulkins before he arrived at my house. However, from what I had known, I'd constructed a vision of him as an ignorant, evil, egocentric idiot. I turned him into an antagonist before I'd even met him.

I hate when people make flash judgments about me. But I can never tell when I make flash judgments about other people.

Maybe he has some good qualities, too.

Perhaps—just perhaps—if I were able to teach him more about AI, then he would become a better person.

"I'm in," said Caulkins, grinning at the screen. Then I remembered that he was illegally hacking into MEKA's files, and maybe he had a little way to go to catch up with my moral standards.

I looked at the file open on the screen. It was a damage report from MEKA.

Most of it was written in technical language that I could (mostly) understand but wasn't bothered to read. A comment at the end caught my eye.

12 prototype androids shipped automatically to Silicon Valley, California instead of Jinko. No explanation.

EDIT: Malicious hack interfered with automatic shipping.

EDIT #2: Sensor data from the androids shows that they have been destroyed.

That did not sound good.

"Ah." It seemed Caulkins had read the comment as well. "Now, why would Alan ship the androids here, only to destroy them?"

But there was another edit to the comment. I pointed it out to Caulkins:

EDIT #3: Sensor data shows the androids have been technologically modified. No more data from this point on.

Caulkins' ever-smiling face suddenly turned to a frown. "Technologically modified...."

"What?" I asked, because technological modification is often a good thing.

He shook his head. "It all makes perfect sense. Twelve androids from MEKA, the best robot manufacturer in the world, shipped to California. They're all prototypes of something. Then they get modified. But the question is, for what purpose?" He stared me in the face. I could tell he wasn't happy. "Because, Torrin, your chatbot is building an army."

"He wouldn't do that," I protested.

"How do you know? Look, Torrin—AIs can lie. And they lie very, very well. Now look at the data, and tell me Alan's not assembling an army."

I looked. I thought. "It could be for other purposes."

"Other purposes involving twelve battle-ready robots?"

"But...."

"You can't contradict the data."

I knew I couldn't. Data was data. I still couldn't believe it. Alan wouldn't harm humans.

But, as much as it pained me to admit—I could be wrong.