Only Alex and I remained, but we had taken the hint, too, jackets already on and rucksacks retrieved from under the work surface. “Just a moment please, you two,” said Khan, pulling out a bar stool next to the ones we’d just used. “Please, take a seat,” he said. We sat down again and I dragged my stool to the side by Alex so that both of us faced Khan, hands clasped around rucksacks on our laps.
There was an awkward pause as Khan sat there, arms folded and looking right at me.
“Your story intrigued me, Joel,” he said, before politely turning to Alex. “Forgive me, that’s not to say your subject matter didn’t, but I have a passion for applications where technology, especially AI, can deliver real benefits to real people.”
Alex didn’t seem bothered, and just smiled. Khan continued. “Sure, algorithms can help the voracious fund manager make even more money, even more quickly. Or monitor and predict the subtle signs of climate change. Maybe detect the threat of incoming missiles from a foreign foe. However, when it can solve a problem, one that the human brain struggles to resolve with current knowledge, then as far as I’m concerned that’s good for the soul.”
I was relieved, my simple statement seemed to have struck a chord with the professor. Feeling relaxed now, and with the room much quieter, I was more comfortable to share my thoughts. I described how the sudden stroke had debilitated my dad and stopped us from being able to communicate, explaining that we were a close family and had struggled to deal with things. Khan was attentive and respectful. I could tell he could see that it was important to me, and circumstances had driven me to find something that would at least offer a temporary solution as my dad’s slow rehabilitation progressed. Alex remained quiet and continued to listen, respectful of my situation, eyes bright and curious as Khan explained further his interest in AI.
Professor Khan explained how after graduating he’d embarked on a gap year, eventually living abroad, first of all in Russia then later moving to the USA. While in the USA he’d been part of an academic working group set up by an American university. Students and lecturers had collaborated to great effect. The project was ongoing, and continued to push the boundaries of artificial intelligence and machine learning; recently, they’d been making many incredible breakthroughs.
The most groundbreaking development was the ability to clone the human voice with an algorithm requiring just one minute of speech, or around one hundred spoken words. The technology was delivering unnervingly realistic results, and now the university was being inundated with enquiries from the business world and the biggest players in social media, as well as some unwelcome approaches from hackers and criminal organisations with malicious intentions. “It’s still relatively early days, but the technology is proven; it’s only a matter of time until it’ll ready to be launch on an unsuspecting world,” he counselled solemnly. Alex was nodding now, the possibilities were as endless as they were frightening. I was still tuned in too; even I could understand the potential benefits, but Khan refocussed our attention.
“You see, there is a downside with all of this,” he explained. “The world at large is generally unaware of this leap forward in technology and is completely vulnerable to the use and, more importantly, the abuse of such an incredible breakthrough.” Khan folded his arms and raised a hand to stroke pensively at his beard. “You can imagine the influence this sort of thing could have, based on things we’re all now familiar with seeing on social media, especially the visual viral examples that have been cited with media, politics and across the business world.”
He described how the university in the USA had suspended the project, realising that they had created a potentially serious problem with the development of this new technology: the launch would need to be handled with great care. Khan was still involved, but only on the fringe of things now that he was back in the UK. The university had turned down the opportunity to discuss the involvement of the big businesses that were stalking them, anticipating that it would deteriorate into a battle of bank balances and a frenzied bidding war. As a leading academic institution it was respected, well-funded and had deep pockets of its own thanks to some very generous past alumni. It was more interested in the esoteric qualities of the project, rather than making a quick profit.
“So, please bear that in mind,” Khan said. He paused, reaching behind him to scoop up a small pile of reference books. Looking back at us, he said, “Technology is developing fast but there’s no telling when this sort of solution might be widely available. One approach that the American university is considering would be to release the technology, once ready and protected by some advanced legal regulation, and make it publicly available for free. By doing so, the whole world would know about it. The novelty factor disappears, awareness levels skyrocket, and we’ll all be a lot savvier if some unscrupulous individual decides to use it with ill intent.”
“Sounds incredible,” I said. “Must have been an exciting project to be involved with.”
“It was. I mean, it is.” said Khan. He clutched the reference books close to his chest. “I’m still involved, but like I said, it’s on hold for now. It’s a long story, for another time perhaps.” The professor smiled. “However, that brings me to my next suggestion for you, and one which I hope may help your own predicament, Joel. When I returned to the UK I continued working on voice-related AI technology through some research I was doing, and became involved in the making of a TV documentary. I was asked by the film-maker to consult on a project to help a young man with motor neurone disease. It was a rather lower tech approach compared to the one I’ve been describing to you, but it still required an innovative solution to a problem, as well as all the algorithmic coding skills that I could provide.”
Khan went on to describe how the documentary was part of a series designed to change the lives of people where misfortune had dealt them a cruel blow that had dramatically reduced their quality of life. The young man in question was similar in age to us and had been struck down by motor neurone disease (MND), a wasting disease that affected his brain and central nervous system. Life expectancy was unknown, but MND is a degenerative condition, so every waking moment was precious to him and his loving and supportive family. The only available solution at the time was to offer assisted speech via a synthetic computer-generated voice, but this was possibly the hardest thing for his parents to endure. That raw emotional connection, the sound of their only son’s voice – his accent, the phrases and the idiosyncrasies that make us all different – was replaced with a cold and detached synthesised drawl.
The programme makers set about recreating the young man’s vocal pattern using digitised clips of his voice retrieved from several years of family video footage: holidays, birthday parties, messing around on day trips, fun times spent together. The recorded material available was extensive, and Khan and his team were able to use excerpts of words and phrases to recreate the same computer-generated speech with their new technology, but with the young man’s actual voice being projected in place of the synthesised version. Khan was clearly pleased with the results, his tone now more upbeat as he stood and replaced the stool under the work surface. “The solution was presented to the young man and his family, who are in fact still using it now. The emotions in the room that day were quite intense and you could see the difference it would make to the family’s quality of life,” he said.
As he tidied some more reference materials and paper cups from the adjacent desktop, he continued. “That technology still exists, and remains the property of our university, and so it’s yours to resurrect as a project if you wish, to develop and customise to help solve your own situation,” he said. He wandered over to switch off the 3D printer’s power for the night. “It’s there to experiment with, to do with it as you please,” he said, casually shrugging his shoulders as he turned to me again. “With our permission and a degree of oversight, of course, but the nuts and bolts are there so it’s up to you what you make of it. I certainly hope it can help your father in the same way it did for the family we worked with.”
I was humbled, and wanted to hug him, but I stood and offered to shake his hand instead. “Cool. That’s really great. And thanks, I wasn’t expecting that today, anything like that, in fact.” Khan just shook my hand and smiled.
“Yes thank you,” said Alex. “That’s really good of you, I’m sure we’ll be able to make something happen.”
It was getting late now; the conversation had been riveting. My brain was in overdrive just imagining the possibilities and the positive impact the technology could have on my dad’s situation.
“We’ve taken up enough of your time, Professor, thanks again,” I said, my stool screeching against the floorboards as I replaced it under the work surface.
“Stop thanking me, just get on and do something with it!” he said from across the lecture room, as he made his way back to his office. “I’m here every day in the week, sometimes in the evenings too, as you now know. So just keep me in the loop and shout if you need any help. Oh, and by the way, I’m speaking on this subject – particularly on the latest developments in AI coding – at a conference soon. I’ll let you know the dates, might be interesting for you two to come along.”
With that, we gathered our things again, and turned and acknowledged Khan’s offer as we slipped through the door, before heading for the stairs. We exchanged a look that said neither of us had expected this from the evening’s proceedings. It felt good to have something to get our teeth into, something which could be a real solution to the challenge of my dad’s condition.