Artificial Intelligence (AI) is the process of getting machines – what we call “computers,” may later call “flawless sapiens” – to make decisions as if equipped with human senses, neural nets, and instinctive powers, as if human. AI would probably even like to chop the wood for my fireplace. Where does AI lead?
This stuff is the rage, folks diving into the deep end without checking for water. That is wrong. But so is deliberate ignorance, trying to imagine AI away.
Misinformation often outpaces knowledge. Some say AI means we are making ourselves extinct; others predict we will live forever, remake mankind, and re-tilt Earth.
Some say this perpetually self-teaching, information-sucking tool is wonderful, others a curse. That debate turns on how we use it and whether we can control it.
This last bit is a public preoccupation – for educators, lawyers, doctors, politicians, warfighters, and those wishing to deter wars, anyone who has paused to do some imagining.
Among the cautious were scientist and fiction writer Arthur C. Clarke, as well as theoretical physics pioneer Stephen Hawking.
Wrote Hawking: We should beware of a technology that thinks itself so smart it can self-redesign until it sees itself in charge of those who created it.
There is something ironic, and mildly comical about that view. It sounds like the “fly on the wall” talking to God as creates Mankind, knowing His creation will make big mistakes, and might one day imagine itself being God.
Arthur C. Clarke, famous for inventing “Hal” – a rather self-righteous and survivalist computer in “2001: A Space Odyssey” – had a similar concern. He imagined we could get above ourselves; and invent machines we regretted making.
Wrote Clarke: “The most intelligent inhabitants of the future world won’t be men or monkeys; they will be machines. They will start to think and eventually, they will completely outthink their makers.”
With deference to ourselves, these minds were often right, too often not to think about the things they imagined.
While neither thought machines would “come to life,” that is, have feelings, heart, faith, or the unduplicatable markers of humanity, they did wonder about limits.
So, let’s ask – what could happen? We know this. The ability to ingest nearly infinite amounts of human learning, find connections, apply them to complex problems, make decisions, and exercise faux “judgment” – with human bias – is real. AI can do that, and fast.
What does that mean? Hypothetically, AI might invent, deploy, and suppress a deadly pandemic, stop an asteroid, personalize health to extend life, understand animal talk, and permit emotional communication with an unborn baby. Anything!
The current evolution of AI already allows us to solve wildly complex problems, align actions with preferences, predict and preempt bad things, tailor papers, speeches, legal briefs, and mete out solutions heavy with math, nuance, and emotion.
AI can digest, connect, and apply emotion – calibrate and compensate for reactions to problems, in effect often get there before we can. That could be valuable.
AI can instantly do analysis based on game and probability theory, and if allowed, act. It can assess, execute, track, continue tracking, and act toward an outcome.
We may not be far away – honestly – from the global ability of AI to assess whether a war could be won, how it should be fought, and then (gasp) initiate it.
But is this all real, or just science fiction – like antigravity machines, invisibility cloaks, interstellar travel, the light disappearing into black holes, time-changing speeds, humans in space, satellites in orbit for instant communications, an internet, email, skype, mobile phones, and smartwatches?
Yes, you get the point. The first three are predictions Clarke made, the fourth one Hawking made, the fifth Einstein’s, and the last five, based on satellites non-existent when Clarke wrote – all things he predicted.
All have come true except the first three – and they may one day also. So, should AI be stopped, pursued with caution, or embraced? Like computers, genetic engineering, and nuclear weapons, AI is out of the bottle. The key is how we use it.
In truth, many do not fear it but urge caution. They see AI’s capacity to change life on Earth – for good and bad. Morality, ethics, the sanctity of life and free societies, respect for human control, and humanity’s uniqueness, all need protection.
Bottom line: The rise of AI is another reason to slow down, as we once did with computing, genetic engineering, and nuclear power, all life-changing. These innovations have improved medicine, communication, transportation, and the quality and length of life, but with caution.
Funny enough, some clever soul asked AI to assess whether Einstein would think AI a danger to humanity. The answer that came back was strangely Einstein-like.
The AI suggested Einstein would not have shut it down, would have liked the innovation, would have counseled caution – just the sort of answer “Hal” might have given “Dave” (the human in Clarke’s “2001: A Space Odyssey”).
I guess, as we better understand and use AI, we should keep one eye on who is telling us AI is fine, to keep going, since it may be AI itself.
As for me, I will not be running ANY of my columns through AI, not letting AI fill my tea kettle or pour my coffee, and certainly not chop my wood. It’s probably just a Maine thing, but I am not going to open the door wide, not yet. Sorry, Hal.
Robert Charles is a former Assistant Secretary of State under Colin Powell, former Reagan and Bush 41 White House staffer, attorney, and naval intelligence officer (USNR). He wrote “Narcotics and Terrorism” (2003), “Eagles and Evergreens” (2018), and is National Spokesman for AMAC.
