A fictional story on real technology with true role models.
By Alex Ruby [~ 10-15 min]
The heart-piercing cry made him cringe. An impatient frequent flyer had barged his way into the boarding line and rammed the turnstile into the head of a little boy whose mother was searching for the tickets. The cry brought Benedikt back to the call that collapsed his house of cards a year ago. Back then for the first time, he had live-tested the voice recognition and analysis of his prototype system Sentalys.
It had been a collaboration with the police. His system had run in the background parallel to the conventional emergency system. When the 911 call came in, it was answered by one of the officers and simultaneously analyzed and processed by his system. A man and his young son were trapped in their car after an accident. Smoke swirled from the hood. It smelled of smoldering plastic. The boy screamed in the background, his calls heart-piercing. Both were rescued, but with severe burns. This could have been prevented. Not by the human-controlled system that had worked like clockwork, but by Benedikt’s computer-based system. If it had done its magic, the two would have been saved faster, less harmed. Why else would it be good to replace human labor by machines?
The system had failed in two ways. First, in recognizing the condition of the man in the car. According to the loudness, intonation and rhythm of the man’s voice combined with the way he structured his answers, the system should have determined the urgency. To complicate things, it had turned out that the man was autistic so his response was more taciturn than expected. Secondly, there were the actions to be taken. After another call came in seconds later, a fire at an engineering company, it would have been imperative to dispatch the available local rescue teams to each emergency according to their abilities. The rescue team that fought the fire in the engineering office had a young officer who held the record in cracking jammed car doors. She would have been able to gain valuable seconds for the rescue team and the man and his boy.
The idea for his system hit Benedikt when he read of someone who had analyzed the tweets of newly split couples to determine whether prosecco and pilsner made them more amiable to reconcile. Only he wanted to use sentiment analysis to create something more meaningful. His system was intended to support emergency call centers in case of disasters like major fires, hurricanes and earthquakes or even terrorist attacks. It should help emergency services to respond faster and more efficiently, and also prevent mass panics before they even started. For example, the largest search engine in China was already processing user data to predict mass panics. 302 million of the 657 million registered Chinese used the map service, which allowed the system to build corresponding movement patterns. After the cataclysmal live test Benedikt had fallen into deep despair. He wasn’t the typical gloomy guy knocked down by challenges, but his most promising investor-to-be had jumped ship immediately only to be followed by his Angel investor a few weeks later. That was the end of his funding.
It was difficult to predict when the system would be advanced enough and ready for another live test. There were several reasons for this. Programming the algorithms for the sentiment analysis was more difficult than expected, because the evaluation of the criteria that influenced a caller’s mood was extremely complex
His leading developer had decided to leave him for another startup that worked on predictive maintenance that fit the typical hockey stick growth model of successful startups. Benedikt was also still searching for a system to collate and structure the enormous amounts of data and to make it instantly available in case of an emergency. Currently, the skills and experience of all rescue workers were spread over thousands of documents and reports. His speech recognition system was based on an open source program of a Silicon Valley giant that also offered a cloud-based version. This had enabled his team to seamlessly integrate it into their development process that was based on cloud storage as well. However, his system lacked the critical amount of voice samples for evaluation.
His situation was incredibly dicey. He needed money. Urgently. He had called on everybody he knew. Family and friends, and friends of friends. Everyone in the startup scene who could possibly give him access to investors. But it was all in vain. His idea was highly praised. However, it was not able to be scaled up in a couple of months, so there was no money. Everybody was infected with gold fever and wanted to find the next Unicorn to cash in fast. Ironically his system would actually have what it takes to become one, just not in such a short time frame. His last hope was a friend from college who had stuck around in Nevada, the new tax haven, after visiting the Burning Man festival and who was willing to open a few doors for him. Maxine, just Max to her friends, worked for one of the most auspicious startups in Las Vegas. They had taken a deep dive into machine learning and natural language processing to teach cars how to self-drive. Benedikt had scraped up his cash and booked a flight to Sin City with the last miles on his Frequent Traveler Card.
He was off to a bad start at the airport in Munich. His flight had been canceled. All passengers had to be rebooked and the staff at the counters were completely overwhelmed. After he landed in Las Vegas a day later, he realized that his suitcase had not followed him. At least he had an emergency kit in his hand luggage together with laptop and charging cables. It could be worse, he thought, and allowed his spirit to be lifted by the profound excitement of a German couple next to him. The two were eager to visit the mecca of gambling.
As he walked out of the airport the dry midday heat hit him like a wall. But the innumerable beads of sweat evaporated immediately in the icy air-conditioned shuttle bus, leaving a thin layer of salt. The same happened when he arrived at the hotel. Exhausted and shivering, he collapsed onto the bed in his room. At least one dream had come true, when he saw the oversized guitar in front and passed glass cases with numerous exhibits of his personal heroes on the way to the elevator. Because Benedikt was not only a founder and IT specialist, he was also a passionate guitarist of a rock band.