Check this week's letters.

Jeffrey Van Middlebrook (a perennial Voice-letter writer) cites someone who 'if he could push a button to wipe out every human being so that smart machines could inherit the Earth, he'd do so. When I said to him that he'd be killing my kids and granddaughter, he replied, "So be it." '

That's just a variation on a well-established mode of thinking, self-absorption in a macabre extreme. In 1930s America, when cocktail-party Marxists applauded the "utopia" under Stalin (they didn't have to live in it, of course), a Dashiell Hammett character pontificates: '"Comes the revolution and we'll all be lined up against the wall -- first thing!' He seemed to think it was a good idea.' More recently, a fad appeared among young adults angry about planetary pollution, who describe humanity as an unnecessary or parasitic species (though I don't hear of them following their logic to its conclusion by exiting this world).

But the first example in the letter misleads without more context. Yes, a nasty nuclear genie escaped its bottle in 1945. But it was on the way out anyway. The Manhattan-Project scientists did most of their work believing that they were in a race against Hitler. Do you suppose Hitler wouldn't have used the bomb? Japanese military officials said later that they'd have used it if they'd completed one. And of course Stalin's USSR followed the US by only four years.

Jeffrey:
-- The time will come when humanity will have to go to war against machines.

Very doubtful. Your letter is brilliant, you hit on so much that I have perceived also from working in Silicon Valley since the 80's, but the threat from intelligent entities is not going to be from machines.

I think we are in denial about where the intelligence is, and where the threat is. We will not have autonomous AI ... maybe ever. We do not have anything that is remotely like AI now. We have techniques of priming machines with programs that can collect and assemble data and in a limited way modify their own programming, but none of these programs know what they are doing or why nor are they likely to ... maybe ever.

The threat we have is the same threat we have always had, only now it is global, nuclear, biological, political, because of our electronic infrastructure and the primitive idiocy of our political systems not to be able to deal with it. We have people on Wall St. that almost took down the world's economic system, and no one could talk about what happened or what the fix was ... even almost a decade later, and these things have happened before.

The problem are people who have the awareness, the motivation, and the skills and ability to get in and tinker with things in our society and political infrastructure that has never been thinkable before. Where do these people come from? Why would a society create members who hate it so much they want to tear it down and take control by themselves?

We have been expert at seeing how powerful and efficient the psychopath is in the American system, and over time how their empires, children and the intelligent application of political and media techniques are to infect and sicken our systems, and most people do not have the awareness or power to be the white blood cells of our body politic, they are too busy trying to survive, their independence too compromised by that struggle, and our system has failed to create citizens who can do their jobs, or who can even transmit to their offspring the next generation the values and responsiblities it takes to keep the country and system going.

An AI cannot do this, or it will not be able to do this maybe ever but certainly for a hundred years, and then not in any kind of credible human control. The threat we face is a subset of ourselves and the failure of the operating system we call society to update and protect itself.

It's a leap from proposing the existence of intelligent machines to the conclusion (sadly common) that we will need to "go to war" against them. What is the basis of this war supposed to be? The competition over resources seems unlikely. Smart machines won't be bound by earth's biosphere and will potentially have all the resources of the solar system available to them. Will it be the perception of threat? But what threat do biologically humans pose to sufficiently advanced smart machines? Or does the "war" begin with a pre-emptive strike by humans? Will it be an ideological difference? Humans can learn to coexist with each other, and most have already. There is no reason to assume that human intolerance will be baked into our machines, and some very good reasons to ensure that it does not happen.

In the fevered vision of both the detractors and enthusiasts of a technology singularity, biological humans are destined for "obsolescence" and will need to "upload" their identities into machines or else be destroyed. While it makes for gripping science fiction, it never establishes a plausible motive. Why shouldn't biological humans continue to exist as part of a larger ecology? Why would our machine progeny be set on destroying us?

This is not to deny the potential harm of technology. Yes, technology could kill all our grandkids (and could in fact kill us all right now through nuclear destruction). The choice is not to stop developing technology, but to make ethical decisions about its impact. Assuming that a conscious artificial intelligence could exist, it will ultimately be the work of humans to endow it with values. These would include ideals such as self-determination of biological entities and the protected status of earth as a biological environment. In fact, there would be far greater ability to carry out these goals removed from our current resource scarcity. The wars we humans fight with each other are real and terrible enough, but it is a failure of imagination to presuppose that war must extend into every interaction between conscious entities.

-- What is the basis of this war supposed to be?

Struggle over power and resources or independence.
It seems likely that the inventors of facilitators of any such
entity certainly are doing it because they want ideas on how
to exploit for power and gain human systems that already
exist. There are some interesting movies out lately on this
theme.

- Ex Machina
- Morgan

And of course the old classic "Colossus: The Forbin Project".
It really depends on what consciousness or intelligence turns
out to mean, or how it is implemented.

-- There is no reason to assume that human intolerance will be baked into our machines

Maybe not, but the nastiness of human history is going to be something
that ... whatever ... will have to deal with and coexist with. Being, or being
perceived as more intelligent and fast than humans, some estimates have
been that machine intelligence will take off without limits almost immediately,
which I see no justification for, we will DEMAND to have the information from
such an entity and want to control it.

-- Why shouldn't biological humans continue to exist as part of a larger ecology?

Oh, because we are so limited by our very nature.

I hate to say it, but humans are built in crazy buy their own life cycle.
What kind of entity struggles its whole life, learns the most, and then
just gets done away with, and according to Malthus leave with more
progeny that have no talent, intellect or wisdom until they are educated
and gain experience ... which by then can be manipulated by those
who are smarter and learn a technology on how to manipulate children
that is born into us because that is how we work.

Whole evil societies are based on this fault in our nature.

As they say, what make more sense ... to have a driverless car make
a mistake, fix it and then propagate the fix to every other car or continually
to have each car reinventing the wheel from scratch every time one is
built?