“I need to get some levels. Neema, tell me what you had for breakfast.” A reporter from NPR, he spoke with that perfect radio voice. How long had he been practicing that?
I settled back into my chair with a cup of tea. My Siamese purred her way between us. “Kale smoothie: kale, spinach, banana, some protein powder, almond milk, and a hint of honey.” Pause. “I like to keep it simple.” Smile. I practice too.
My gaze wandered over the coffee table positioned between us, tracing the irregular swirls embedded in the polished grain. It’s always patterns with me. The digital recorder flashed its own pattern, keeping time with the conversation.
“Alright, that’s good. Let’s start. Neema, tell me about Ujao. How about starting from the beginning? You and Austin, meeting at MIT?”
As I sipped my tea, I remembered those early days when Austin and I had first met. I was fresh out of Morgan State, a nerdy kid from Baltimore, and the first in my family with a postgraduate degree. Austin had grown up in academia, the son of Ph.D.s. He had the most amazing magnetism, and people were naturally drawn to him.
“We met in the Media Lab,” I began. “I was working on this idea of a projection, like a Mercator map, that would take all this textual information we could pull down from sources like The Pile, social media, news sites, and plot narrative trajectories across it. You know how we say stories have an arc? It turns out you can plot that if you have the right kind of map.”
I rambled on about seeing those first predictions of possible futures laid out as branching paths on three-dimensional maps of human culture, biases, and beliefs. “Those trajectories were the raw material for our predictions. The quantum computer let us find the most likely options. Prediction is like breaking encryption, where history is a hashed key. That’s why we called the project Ujao—it’s Swahili for upcoming.
“Austin and I balanced each other’s strengths. I worked on improving our models, while Austin handled the media and investors. Everyone used to say that he could sell sand in a desert. Then we had this major breakthrough. Our first big run successfully predicted the China-India War over Aksai Chin last year—to the day. When the news broke about the missile exchange, the government came calling.
“After that, we got so much attention, it just took us to a whole new level. I was really proud of how we created a global team to map out these potential futures.”
I leaned forward. “We became like an international family at Ujao, everyone driven by a common goal. It sounds cliche, but we really were a melting pot of cultures and languages. Each team member brought their perspectives, which let us tease out what the system was predicting. It was never obvious, but the results spoke for themselves.”
I steeled myself for the inevitable question I knew was coming, but it still caught me off guard.
“The leaked company messages and emails . . . How did those affect your company culture, especially with Ujao being such a big deal?” That smooth NPR voice. Sounds great on the radio, but it’s a different story when it’s grilling you.
I took a deep breath. “We thought we were geared up for success, but we just weren’t ready for the sheer magnitude of attention it got us. We thought we could handle it, but we couldn’t.”
I sighed. “Austin tried his best, you know? But he felt the pressure and thought he had to handle everything himself. And with all those leaks, he just got more and more paranoid. It didn’t take long for that to start messing with our team.”
My shoulders slumped. “Austin began letting people go who he thought couldn’t be trusted. He brought in new advisors who convinced him to make changes. And that’s when I started getting pushed out.”
Empathy in his voice, the reporter asked. “And what happened to Ujao’s product during this period?”
“Our predictions became overpriced guesswork. That blend of cultures and insights that supported all that technology became diluted, and it showed in the results.”
The reporter seemed to be preparing to ask a question but hesitated, checking his notes. “We have reliable sources suggesting that Ujao may have fallen victim to a new kind of attack, one that targeted the company’s management. An AI program, embedded within the servers, manipulating channels like email and chat.”
I stared at him, my eyes wide with a blend of curiosity and apprehension.
“An AI program? Hidden inside our servers?” My voice was tinged with disbelief. “But how would such a program go undetected for so long? We had good security. Particularly for data exfiltration. We would have seen something.”
The reporter’s gaze remained steady. “This program had a different goal. It wasn’t about stealing but causing chaos in the company. By changing or creating internal communications, it preyed on people’s weaknesses, especially those in power.”
I recalled Austin’s slow drift into paranoia. “It played on our vulnerabilities,” I murmured, my voice barely audible even to my own ears. “Could it have affected my decision to leave the company?”
The reporter nodded sympathetically. “It’s certainly a possibility. The nature of this kind of attack is that it operates subtly, almost imperceptibly. It erodes trust and distorts perceptions.”
As the words sank in, I felt a mix of emotions flood over me. Anger, frustration, and a strange sense of awe at the sheer elegance of the attack. In a world where technology had become intertwined with our everyday lives, the fabric of our realities could be manipulated.
“So, without firing a shot of any kind, they managed to destroy Ujao and create the impression that our predictions were the result of dumb luck,” I said, my voice tinged with both resignation and an underlying sense of admiration.
I stared into the distance. “An attack like this could work on any big organization that depends on computers for communication, which is pretty much all of them.” I said, picturing a future run by the technology-averse Amish. “It’s not just about business and industry, but governments too . . .”
I hesitated for a moment before turning toward the reporter. “How can you be sure that NPR isn’t suffering from the same problem that took down Ujao? The media seems like a prime target for this kind of attack.”
The reporter’s eyes glanced away. “We’ve begun examining our systems, but the truth is, we’re only just starting to understand what to look for. It’s like trying to find a needle in a haystack, and this needle does not want to be found. Our systems are huge, complex, and interconnected. To find an AI subtly manipulating things from behind the scenes is no easy task.”
A heavy silence fell as both of us recognized that this was something that transcended one company’s rise and fall. How many others might already be infected? What had been an interview about the past shifted into something else—the understanding that our world would never be quite the same.
