3 minute read
A NO BRAINER
from Block Issue 02
by SiGMA Group
BY KATY MICALLEF, HEAD OF CONTENT AT THE MALTA A.I. & BLOCKCHAIN SUMMIT
e stand on the precipice of a technological revolution that is set to alter the very fabric of our digital world. Change has been tugging on the threads of society for some time now and it is impossible not to feel the deep impact of this intelligence rebirth as it beats the rhythm of a not too distant industrial revolution – sounding reverberations that are both triumphant and cautionary. We must steady ourselves for its impact, for these are also the drums of war.
Mankind finds itself facing a fast-paced renaissance of technological creativity and discovery without precedent; this is a chapter in history for which we must find a natural balance between prudence and freedom. Embracing innovation without the burden of fear has always been a cautionary tale.
Human intervention in our natural ecosystem has thus far ensured our survival, as messy as that may be. Optimistically the hope is for intelligent machines to refine that metamorphosis, building on the best of humanity and elevating it beyond its limitations. From governance to security and from education to research, the vast scale of this evolution is far-reaching and radical in the transformation of its applications. Its ability to connect nations with information has given rise to a truly powerful age of information.
This digital age has bequeathed us an abundance of data. A population bursting at its seams and the rise of globalisation has enabled us to generate vast quantities of data - a tantalising food source for an intelligence system that feeds off knowledge. How we evaluate and generate this digital information will be transformational to the way technology evolves. The advances in software development show rapidly improving machine learning algorithms; the promise of software that can learn by example is immense. It took AlphaGo Zero just three days to train itself through self-play, beating the world’s top player at the strategy game Go and releasing itself from the limitations of human knowledge. It is undeniable, however, that they still require millions of examples for any real learning to occur; humans are still much faster and more efficient at picking up new tasks. It is important to remember that computing power alone is not intelligence. Without organisation and interaction, no process can hope to replicate human brain circuitry.
Such fast evolving intelligence may eventually outpace even the best human brain, something many consider a milestone along a trajectory that will almost inevitably lead to singularity. It brings with it all the opportunities of advanced machine intelligence, including A.I.-driven automation - opportunities that are tinged with both peril and value.
The way we navigate this boon will define humanity for years to come. For some the temptation of power has already been too easy to resist, leaving them knee-deep in trouble as they attempt to wade the murky waters of technology as a tool to curb liberty and engender social compliance.
We can at least acknowledge that even if we do pick the less perfect path, that these decisions have been difficult, that we have battled with our conscience and felt some pangs over our choices, as a quest for evolutionary survival propels us towards our desires. Our decisions have been our own. The question is - will they continue to be so? Should machines gain super intelligence, any deviation in roads travelled could well be our undoing.
There is no such indication that A.I.motivated values would align with our own: when it comes down to it, will A.I.driven singularity chose their own survival or ours?
There is nothing to suggest that they would prioritise our interests above their own, and if our own human example is anything to go by, we should indeed consider this an existential threat to our species. According to blogger and author Sam Harris, we could become little more than ants underfoot – inconsequential but ultimately disposable if we get in their way. It is chilling to think that one day machines could treat us with similar disregard.
In truth though, there are shafts of light amongst the shadows. A.I. projects should inspire hope too. Take Google’s DeepMind, a machine learning research project which has, amongst other things, opened up new avenues in the way we address climate change and energy use.
Technology-triggered evolution in machines has forced us to consider a future that just a few generations ago seemed so out of reach. Whether the hype surrounding it is merited or not is irrespective – it would be naive to ignore the concerns it raises.
And while it is unclear how far along the road to singularity we are, the question of who we want to be as a species has never been more pertinent. The only thing we can be certain of is that ready or not, change is upon us.