Artificial Intelligence: The “Catch me if you can” moment
Looks like we, currently living humans, are going to witness the finale of the race between biological-human-brain and human-created-Artificial-brain. Interestingly, both have fundamentally the same age-old underlying research: ‘reverse engineering the human brain’.
Following are some thoughts around how would that turn out.
We know that machines are already better than human at doing some very specific tasks that can be automated and humans are prone to error — such faster & accurate diagnosis, labelling of pictures etc. As a next step, entire advanced research on Artificial Intelligence is focused towards achieving Artificial ‘general’ Intelligence (AGI), where machines can do things that humans are far superior at, such as sensing (see, hear), general reasoning out, creating & executing plans etc.
Out of many approaches, the most advanced one towards this goal is Artificial Neural Networks, basically an approach of mimicking how human-brain works, or more specifically, how it ‘learns’. Something that is termed as ‘Deep learning’ in AI terminology. A large community of AI practitioners are already solving many problems by applying this approach. But this approach itself is in its early stage, and will evolve with advanced brain scanning techniques trying to figure out what goes on in human brain in many situations.
After achieving AGI, when machines become more human like in performing better at above ‘tasks’, and that combined with its already superior powers of “speed” & “scale” to multi-task, memorise & communicate, machines can very rapidly outgrow beyond super human abilities, with all of its powers evolving exponentially.
This hypothetical state is called as Singularity, and Ray Kurzweil famously predicts that we will experience ‘Singularity’ in approx 30 years from now. If this hypothesis turns out to be positive, current state of human brain will be incapable of even understanding machines and their responses.
As science-fictional as it sounds, it is real-enough to kick-in our survival instinct.
Enter the “Brain-Computer-interface” (BCI) to create, lets call it, the Artificially Enhanced Human, by recording what is going on in brain (either by scanning on surface or implanting electrodes within the brain). Its in early stage of trails, but if the technology evolves at a faster rate, it can create and ‘evolve’ to become “ultimate” smart humans.
Looks like this revolution of ‘smarter’ humans is already on — Most of us are now smarter than we were 20 years back, aided by our smart phones and virtual assistants. But still constrained by the speed & ability to multi-task being limited to our brain’s respective abilities. With implanted computers in brain, that constraint can be tackled.
One type of visionaries are investing in this technology to prepare for survival, while majority of the recent progress and investment is driven by business opportunities around it. Although there is a philosophical and regulatory issues still to be tackled, technology is already being adopted, primarily for medical conditions of brain’s degrading abilities.
As the technology advances, the Artificially Enhanced Humans can hypothetically download a skill, an ability, a personality trait and can become, at will, smarter, sexier or humorous and more importantly, will be able to multi-task, while being super-efficient at each task.
So, machines being trained to have human cognitive abilities and human’s brains are being melded with machine.
This I think is a ‘catch me if you can’ moment. In my opinion, in the race between human and machines, outcome will be Artificially Enhanced Human and everyone wins.