We don’t know when to stop

By juni 20, 2017 Algemeen

Civilization has come a long way, and our understanding of the world and the advancements we have achieved deserve admiration. After nearly four billion years of Darwinian evolution, we have created a separate and independent evolutionary logic. Whereas what lived and died used to depend only on natural selection and random mutation, now at least half of what lives and dies is due to human selection.

One of the benefits of our evolution is intelligence. In its varied forms powers every opportunity we pursue and every problem we seek to solve. Intelligence is what allows us to create forms of governance, cure disease, create art and music, discover, dream and love. Intelligence is also what decides that these things, rather than other things, are worth doing, by translating discoveries into meanings, experiences into values and values into decisions. And at the same time, the world remains in great peril and abysmal disparity. The potential for conflict between (or within) some societies seems omnipresent, or at the brink of eruption. Seemingly we have to step up our own evolution a bit more.

A seemingly simple change 2.5 million years ago — using stone tools to butcher animals — led early hominids down the path to becoming modern humans. From that modest starting point, throughout human history, we created tools that increased our individual and collective intelligence and became extensions of our natural selves. Both our shift to agrarian lifestyles and the Industrial Revolution were marked by major shifts in diet, exposure to infectious disease, and risks of chronic disease.

Culture (agriculture, medicine, technology) now changes much faster than genes and traits can evolve. Have we therefore entered an era in which we are going to be permanently mismatched to a rapidly changing science and technology for as long as we can foresee? Naively, yes one would expect the rate of advance of machine intelligence to outstrip that of biological intelligence. Tinkering with AI and smart machines seems easier than modifying a living species, one generation at a time. But advances in genomics—both in our ability to relate complex traits to the underlying genetic codes, and the ability to make direct edits to genomes—will allow rapid advances in biologically-based cognition.

We are poised for an explosive, generative epoch of massively increased human capability through an explosion of possibilities represented by the simple equation: HI (human intelligence) + AI (artificial intelligence). When HI combines with AI, we will have the most significant advancement to our capabilities of thought, creativity and intelligence that we will have ever had in history.

The biggest bottleneck in opening up this powerful new future is that we humans are currently highly limited in how we can participate in these possibilities. Our connection with our new creations of intelligence is limited by screens, keyboards, gestural interfaces and voice commands — constrained input/output modalities. We have very little access to our own brains, limiting our ability to co-evolve with silicon-based machines in powerful ways. The frontier machine intelligence architecture of the moment uses deep neural nets: multilayered networks of simulated neurons inspired by their biological counterparts. Silicon brains of this kind, running on huge clusters of GPUs (graphical processor units made cheap by research and development and economies of scale in the video game industry), have recently surpassed human performance on a number of narrowly defined tasks, such as image or character recognition. We are learning how to tune deep neural nets using large samples of training data, but the resulting structures are mysterious to us.

Neuroprosthetics. In recent years, research labs around the world have made enormous strides in understanding how the brain works, how to connect it to outside sources and how we might tap more deeply into its potential. The most immediate need for these devices is apparent in the growing number of people living longer lives while suffering from neurodegenerative disorders. These devices — by directly extending HI, including our memory and other cognitive capabilities — could lead to unprecedented longevity of the mind and body.

There is more.

HI -enhancement-. Another rapidly evolving and advancing intelligence besides that of AI and smart machines: our own brain. Unraveling the genetic architecture of complex traits such as human cognitive ability. Recent advances allow highly targeted editing of genomes, and will eventually find their uses in human reproduction. The potential for improved human intelligence is enormous. Cognitive ability is influenced by thousands of genetic loci, each of small effect. If all were simultaneously improved, we can’t imagine what capabilities this level of intelligence represents, but we can be sure it is far beyond our own. Cognitive engineering, via direct edits to embryonic human DNA, will eventually produce individuals who are well beyond all historical figures in cognitive ability. Now our newfound powers take the randomness out of genetics and ensure directed evolution. To modify our own gene code such that we are fundamentally redesigning our species and a vast number of other species. We are redesigning life itself. Will we have better minds, be nicer, lovable, emphatic or just more effective and efficient?

These two threads—HI- enhancement- and AI—will inevitably intersect. Just as AI will be much smarter in 2050, we can expect that the humans who design, build, and program them will also be smarter. Perhaps we will experience a positive feedback loop: better human minds invent better machine learning methods, which in turn accelerate our ability to improve human DNA and create even better minds. Also, once machines reach human levels of our enhanced intelligence, our ability to tinker starts to be limited by ethical considerations. Rebooting an operating system is one thing, but what about a sentient being with memories and a sense of free will?

Does society need time to research, understand, and discuss the consequences, both intended and unintended.

The potential benefits of HI- enhancement- and AI are undeniable. What becomes the new norm as we try to improve ourselves? Who sets the bar, and what does enhancement mean? You might enhance people to make them smarter, but does smarter equal better or happier? Should we be enhancing morality? And what does that mean? But we are the species that never knows when to stop.