Chapter 2:
Jostling with the Future
When you look someone in the eye for the first time, you instinctively like or dislike them. Many times, you aren't even aware of this assessment. It might be just a nagging feeling in the back of your mind. Years of training for social interaction kick in, and you force yourself to evaluate that person within the parameters of what you know about them and how you expect your relationship to develop. You control your initial gut feelings and shape in your head a model of that person meant to satisfy the rigors of society. On rare occasions, though, you can't escape the sensation of deeply hating or being profoundly comfortable with someone.
Ian was one of these exceptions. When you met him, even if you tried hard to dislike him, for whatever reason, you just couldn't. The moment you talked to him, his straightforward way of thinking and his jovial attitude were bound to captivate you. Honesty seemed anchored deep in his bone marrow. Nothing and nobody could touch him. Like a duck not bothered by water, Ian could withstand any offense, jab, or personal tragedy. He couldn't hurt somebody's feelings, even at gunpoint. He was the exact opposite of a certain hot-headed researcher.
On the surface, he projected a laissez-faire approach to life, a sort of "live and let live" style. But his commitment to progress and his empathy towards his fellow man were unique.
- DeSousa's Memoirs - Part I - Beginning - 2053
This year, the Neuromimetic Conference was hosted by the extended Disneyland Event Group. François went there straight from the train station. The motel he reserved wasn't far away from the conference venue.
Once he passed the gates of the huge resort, he felt as if he'd stepped into a mini town inside Paris, and one that was cheerful and funny. He thought the environment was highly inappropriate for the matters discussed. Sometimes organizers of these events went too far in trying to create a mood.
He arrived at the event location. Seeing the sizable crowd in the main auditorium, François instantly felt a sharp pain in his chest. He struggled to dismiss the fact that these couple of hundred people were, in fact, just a tiny fraction of those watching the event in the Cybersperse. Based on the number of tickets sold, he expected that the conference would have a remote audience of over 100,000 people. Since many would get together physically and share the remote view, especially in poor countries, nobody knew the actual number of participants. That was because the perspective of true AI, integrated deep into our biology, promised unprecedented prosperity.
François was scheduled to speak late in the evening, around 7:00 p.m. He figured most of the attendees would be too tired to pay much attention to him, half of them dozing off and the other half out in the hallways trying to establish business relationships while enjoying the fine French cuisine and wine.
So, no reason to panic. I'll present, answer questions, and be done with it.
Still, he reasoned that this was the first day in what promised to be a long and tiring week, so the audience's stamina would be strong.
What's wrong with me? he chided himself. Why can't I look at this as a tremendous opportunity? He suddenly felt more motivated, only to fall back to desperation. Ha, I started talking like Marie. Who am I fooling? I'll be the laughingstock of this crowd.
Vacillating between panic and optimism, he started paying more attention to the presentations. First, a professor from École Polytechnique Fédérale de Lausanne talked at length about a spin-off of the old Blue Brain project, then a young lady presented Giroux, a start-up providing intelligent avatars for the Cybersperse. Later, somebody from the University of London advocated for the use of brain-computer interfaces for medical research, and finally a Russian researcher insisted that the path forward was Quantum AI. François found it all only mildly interesting, bordering on chaotic-an eclectic mix of new and old unproven ideas reshuffled.
When his turn came, he stood up fast, knocking down the neighbor's tablet, and nearly ran to the podium. Looking at the crowd all focused on him, he realized, again, how much he hated public speaking. His old professor was there for support. He went through the details of his algorithm fast, like an android reading a short story at 1.5x speed. Somehow, he finished four minutes before his allotted time, a performance unique in the conference so far. He dared to look at the audience. A few were politely trying to mask a yawn, others were shamelessly looking at their tablets, and some seemed disgusted. He thought he saw one or two more supportive looks.
Probably out of pity, he thought bitterly.
Since they had plenty of time for questions, the moderator invited the audience to jump in. There was only one. It came from a middle-aged gentleman asking it with a contemptuous rictus.
"Doctor DeSousa, could you explain to us why you named your algorithm adaptable and not adaptive?"
François cleared his throat. "Because this algorithm is very flexible. It has hundreds of parameters that you can adjust. It's malleable in thousands of different ways, yet it maintains its core abilities to recognize patterns in the inputs."
"Aha, so it's not intelligent after all, is it?" the interlocutor said triumphantly.
It certainly is at least as intelligent as you, François wanted to scream at him. He answered instead in a trembling voice. "Not a single cell in a brain is intelligent, yet they give rise to the mind. Without them, there is no intelligence at all."
The gentleman waved his hand dismissively and didn't bother to reply. Given the lack of interest, the moderator had no choice but to thank François, who returned to his seat red up to the tip of his ears. He felt ready to burst into tears.
Curse Marie! Why did I accept this humiliation?
He zoned out for a couple of presentations as he slowly calmed down. He shouldn't feel embarrassed. People here didn't know him. And he would remain a nobody.
The title of one of the next presentations, though, captured his attention: Towards Achieving Consciousness in Structural Artificial Intelligence. In many specialists' opinion, with whom François agreed, the study of how the brain gave rise to conscious experience, from basic sensorial input to the feeling of being alive and the awareness of oneself, was the holy grail of all cognitive science disciplines. Researchers had been able to pinpoint parts of the brain that became active when people were conscious and remained silent when people were in a deep coma or under total anesthesia. They named those active regions "neural correlates of consciousness." What exactly consciousness was and how it came about remained the greatest mystery in the universe, despite many recent claims that machines had achieved it already.
The presenter was Professor Ian Ndikumana from the University of Southern California in Los Angeles. He looked to be in his late 60s, perhaps early 70s. His lively eyes, serene face, and kind smile struck François. His attitude effused so much optimism that François got annoyed and envious.
"He's old. No wonder he still talks about those kinds of structures in AI," François sighed.
Two mainstream methods of building an intelligent agent were conventional. The first, remnants of the remarkable success of machine learning in the 10s and 20s, worked by learning from billions of examples. The second worked by implementing detailed simulations of brain networks in massively parallel computing platforms. Combining them became popular, especially after China's success in '45, when they implemented in silicon a complete brain neocortex that could run intelligent algorithms.
François decided he would follow closely what the professor had to say, not the least because he had an irresistible Scottish accent. His presentation started with a long sequence of failures in AI, from the initial enthusiasm of 1950s and 1960s to the lack of funding of the 1980s, back to the success of the Deep Neural Nets and Large Language Models in the 2020s.
Great, we are here for a history lesson, François thought, annoyed, shifting his focus to the attendees around, some of whom were already scoffing or rolling their eyes.
Eventually, the professor got to the meat of his presentation. "The word 'structural' in the title of my presentation refers to an approach in which," the professor said in a confident voice, "we explicitly give the intelligent agent some a priori knowledge about the world. This includes things like information about its physical properties, say, the way a container holds a liquid, the 2D/3D space, the difference between solids/liquids/gas, or other more subtle aspects, like the relationship between cause and effect, temporal characteristics, and so on."
Blah, blah, François thought. Many have looked in...
