Today a client asked me, “are you worried about the drought?” I said, “No, I’m worried about AI eating the world.”
I’ve been reading Nick Bostrom’s book, Superintelligence: Paths, Dangers, Strategies. Here are my takeaways:
1. Destructive Potential. In our lifetime, we now have the capability to destroy not only the Earth but the entire Universe.
Here’s how. A nuclear holocaust would wipe out humanity. A runaway artificial intelligence (AI) that seeks its own ends, not ours (i.e., has no recognizable human values), would self-replicate at the nano-scale (and increase its intelligence and capacity exponentially over successive generations of self-replication) and could use all means at its disposal (including human life and nature itself, with no regard to ecological balance) to achieve its agenda.
Bostrom gives the example that if AI has been programmed to maximize production of paperclips, it could turn the entire Earth, or the Universe, into paperclips.
2. No Recognizable Values. The AI is essentially an alien life form. It could be “all head and no heart.”
As noted, the AI could have no values recognizable by humans). No wisdom. No compassion. No capacity for self-reflection. All these things that make us human. We cannot predict what the AI’s agenda will be; we can only entertain the possibility that the AI could be instrumental and pursue its own ends.
3. Reducing Existential Risk. The most important task facing us is reducing existential risk.
Certain philosophical problems–such as what is the meaning of life–can be deferred until we design a greater intelligence capable of addressing these in greater depth (assuming they have wisdom); but we must have life in order to ascertain the meaning it asks of us.
As Bostrom puts it, “a plausible default outcome of the creation of machine superintelligence is existential catastrophe.”
4. Decreasing Scale Required to Inflict Annihilation. Netyanahu’s speech to the UN was a wake-up call for me: distilled, he said that current U.S. policy will lead to Iran having a nuclear bomb within 10 years, and that, to the most volatile region in the world becoming a set of nuclear trip-wires.
Whatever else what thinks of the politician or political debates, the reality hit me that a destabilizing and unstable state actor will probably have nuclear capability within my lifetime. I will be 10 years older (you will be too) and some ideologically-bent, psychologically unstable individual may have the physical wherewithal to destroy everyone.
Now, Bostrom proposes that once we develop a critical mass of information about developing AI–much of which will be available publicly–increasingly smaller projects and actors will have the ability to create a runaway, deadly AI.
5. The Purpose of Reducing Existential Risk is to Save Humanity’s “Cosmic Endowment”. We were given gifts, and we have gifts to give.
Bostrom writes: “Through the fog of everyday trivialities, we can perceive–if but dimly–the essential task of our age… the reduction of existential risk and the attainment of a civilizational trajectory that leads to a compassionate and jubilant use of humanity’s cosmic endowment.”
6. Our Destructive Technological Capacity Exceeds Our Wisdom (and We’ve Run Out of Time). Due to accelerating technologies–the curve for which is becoming vertical, aka the Singularity (aka Ray Kurzweil’s Law of Accelerating Returns, exponential growth)–our destructive capabilities exceed our wisdom.
The only hope, therefore, is to: (1) contain destructive capability – and here, Bostrom gives a range of options (such as “anthropic capture” or getting the AI to play “inside the sandbox”), or (2) attain wisdom.
As W. H. Auden wrote: “We must love one another or die.”
Getting back to healthcare, health tech, and the technology trends that concern me as a healthcare & FDA lawyer, this is where integrative medicine, holistic health, spirituality, and the consciousness movements kick in. I wrote about this in Future Medicine and Healing at the Borderland of Medicine and Religion.
It may seem odd that I titled a book, Future Medicine, before mobile medical apps even came on the market.
Nonetheless, the idea that consciousness is the future of healthcare, is a perennial. The “energy” in energy medicine is awareness, intentionality, conscious direction of the nonphysical part of our being that cannot be reduced to biochemical circuitry but is the proverbial “ghost in the machine.”
More to say about this and maybe someday a robot that passes the Turing Test will explain it in terms everyone can understand.
Part of consciousness evolving is beyond what I can put in writing at this time. There is some essential evolutionary process in us that integrates experience, augments wisdom, and moves us deeper into the heart.
Sometimes this is associated with mysticism, other times with yoga, and at other times with self-knowledge as expressed in different ways through various means and traditions.
Sometimes it’s mindfulness – shining light on darkness so we can shine brighter – clearing the mud so our true nature, as Patanjali expressed in the Yoga Sutras, can burst forth.
There is a linguistic connection between health, healing, and wholeness.
Today was strange. A lady cut me off in line to the Starbucks takeaway. As I looked in the drivers’ side mirror, laughing and smiling at the woman giving her the coffee, I found myself entertaining negative thoughts. I watched this negative spiral go on in my mind–mindfully of course!–reminding myself that one does have to listen to their thoughts. On and on it went about how she would get her just desserts (pardon the pun). When, after what seemed an infinitude of patience, enduring her ill-gotten gains (i.e,. getting her coffee a handful of seconds before me), I drove up to the window. The server smiled and said, “the lady in front you apologizes for cutting you off; she made a wrong turn, and is sorry; by the way, we don’t need your debit card – she paid for your breakfast.”
Ah, I said, thank you for being my teacher.
After enlightenment comes the laundry. At least I was witness to the negativity rather than being absorbed in it. I still have the capacity to acknowledge, thank, and bless.
This is a small, random moment – and yet in the triviality of everyday existence lie opportunities for growth.
Speaking of self-reflection, what would a runaway AI have done with that incident, if it wants lattes, not paperclips?
Seriously, though, speaking of our cosmic endowment, Bostrom points out that we must consider not only the current population of the Earth, but all the planets and star systems we might populate once we achieve the technological means to do it. He calculates a number of human lives that has a lot of zeroes and commas after the 1.
If this life is a simulation, or one of many possible timelines, it seems we are at the verge of apocalypse or breakthrough – dystopia or utopia. Now, more than ever, acceleration of consciousness equal to or greater than acceleration of technology, is needed.
We’re an intelligent program with free will – the ultimate programmer has not constrained us in a sandbox – yet, as far as we know.
Although I think the “zoo hypothesis” is plausible. We may be either quarantined from other species in the multiverse, or otherwise under observation, unless and until we obtain the rest of the intelligence required to join the cosmos.
There’s a movie out, Ex Machina, about a robot that passes the Turing Test. The “deus” has been removed. Small wonder – this is an era of technology, revisiting the Cartesian notion of “God the Clockmaker” who sets things in motion and watches (with the Creator’s version of the Apple Watch, perhaps) from a distance while all the movements continue in perfect harmony.
Yet what makes us human goes beyond our capacity to innovate technologically. Again, the answer may have something to do with:
- access to Spirit
Bostrom richly speaks to our cosmic endowment: “If we represent all the happiness experienced during one entire such life with a single teardrop of joy, then the happiness of these souls could fill and refill the Earth’s oceans every second and keep doing so for a hundred billion billion millennia.”