Chatbots and Intelligence

Chatbot technologies are prompting predictions that automation is going to enter the white-collar space. This inevitability leads to concerns that AI is going to replace humanity. Prophets are using words like “intelligent,” “sentient,” and “conscious” to describe their assistants.

This is all based upon the criteria for intelligence proposed by Alan Turing. The problem is that Turing’s test (can I tell if I am conversing with a computer?) is not a meaningful test of intelligence. Intelligence is the ability to change behavior in response to a change in the environment. The environment known to a chatbot is grossly impoverished in comparison to the environment experienced by humans. The capacity of the chatbot to navigate that environment is almost non-existent – it does so only under the rules defined by its training algorithm. What these systems actually do is propagate human intelligence and combine language in novel ways.

Without intelligence, claims of sentience and consciousness fall aside.

The real problem with these technologies is that other people will use them to create the impression that they are intelligent and moral actors. Copying the speech of Gandhi or MLK Jr. is going to become easy. We are going to have to invest in deeper means of assessing capabilities – such as actually observing what people do.

Then What are 1000 Pictures Worth?

Reports of the dimming of the star KIC 8462852 have been debunked, causing SETI to revise its claims to have proven the existence of extra-terrestrial intelligence. The news also caused a crash in Appalachian coal futures, as CO2 sequestration speculators cancelled orders.

One insider, speaking anonymously to avoid being labelled as a “Koch-head,” revealed “when my employers were convinced that no earthly engineering team could dig an ocean through the Rockies, they were hoping that the ETs would do the work in the course of removing the sub-surface CO2 stockpiles they were hoping to establish in New Mexico and Arizona. No ETs, no CO2 sequestration, no last-grasp strip-mining in Appalachia. Oh well, there’s always that land trade for the Panama Canal!”

More seriously: it turns out that the original study of KIC 8462852, drawing upon analysis of old photographic plates, had failed to account for differences in the equipment used to capture the pictures. By comparing the apparent brightness of KIC 8462852 to that of other stars in the plates, it was determined that the the relative brightness had not changed.

Systematic effects (related to the design of the experimental system) were also a large factor in fueling the “cold-fusion” hype that I got involved in debunking back in the ’80s.

Christianity and Paganism

In response to this post in Gods and Radicals.

It is misguided to found any argument about the future of a spiritual tradition upon the success of political figures in corrupting Christianity.

All gods wish for their followers to worship only them, because it is through the acts of their followers that they are invested in the world. That investment long predates humanity – there were Neanderthal gods, and before them gods of mice and gods of dinosaurs. The problem facing humanity was to create a human god in the context of billions of years of predecessors. That is the project of monotheism – to create a god that manifests and supports the expression of humanity’s unique talents.

Now perhaps the essence of humanity’s talent is political organization, but I see it differently. Looking at our evolutionary success, I would argue that humanity is a manifestation of intelligence. For the original adherents (not those indoctrinated in service to the priests, which is a problem in any tradition), the attractive proposition of Christianity was that the divinity served humanity. Christianity is the original humanism – it is to assert that the human god should be a god of love, and serve all equally, without regard to station or industrial skill.

Obviously this is a reasonable proposition, and the power of the Church in the Roman world came not because of the allegiances that joined the interests of emperors and priests. Rather, it was because in the Roman context of utilitarian worship, the Church followed Christ’s edict of charity. The Church, though oppressed, took care of the orphans and widows, the sick and poor, and organized their gratitude to the service of others. When the Empire collapsed, the Church assumed control because they were the administrative and organizational backbone of Roman society.

I see paganism as a political act on the spiritual plane. Humanity, having succeeded in propagating the tyranny of utilitarianism through the application of intelligence, is confronting the fact that it is destroying the fundament of its own existence. It needs to think about all of those forgotten gods. It needs to infect them with rational understanding, and engage them in expression of mutual support. In other words, Humanity needs to join in loving the world, rather than just itself.

This is a difficult pivot. Our religions are still infected by expressions of our physical vulnerability: as an illustration, the vulnerability of a child whose cave is invaded by the saber-tooth cat while father and mother are away. Many people still live in circumstances of vulnerability, although the predators are no longer other species, but rather politically powerful people.

Jesus preached that the meek will inherit the earth. As a reaction against abusive political structures, I see paganism as furthering that goal.

The Imitation Game

I’ve been known to get emotional at the movies, but it’s been since Alien that I’ve been as broken down emotionally as I was today by The Imitation Game.

Alan Turing not only made fundamental contributions to the mathematical foundations of modern computing, he also formulated an inspirational goal for machine intelligence. Known as the Turing Test, it proposes that if a human communicating through a neutral interface (such as a teletype) can’t distinguish the responses of a human from those of a machine, then the intelligence of the machine must be considered to be comparable to a human’s.

My father, Karl Balke, was one of the men that plowed the field cleared by Turing and others. As he described the think-tank at Los Alamos, the researchers brought every intellectual discipline to bear on the problem of transforming logic gates (capable only of representing “on” and “off” with their output) into systems that could perform complex computations. Their research was not limited to machine design. Languages had to be developed that would allow human goals to be expressed as programs that the machines could execute.

In the early stages of language development, competing proposals shifted the burden of intelligibility between human and machine. The programming languages that we have today reflect the conclusion of that research: most computer programs are simply algorithms for transforming data. The machine has absolutely no comprehension of the purpose of the program, and so cannot adapt the program when changes in social or economic conditions undermine the assumptions that held at the time of its writing. It is left to the “maintenance” programmer to accomplish that adaptation. (Today, most in the field recognize that maintenance is far more difficult than writing the original program, mostly because very few organizations document the original assumptions.)

I believe that my father’s intellectual struggle left him deeply sensitive to the human implications of computing. As I child, I grew up listening to case studies of business operations that came to a grinding halt because the forms generated by the computers were re-organized to suit the capabilities of relatively primitive print drivers, rather than maintaining the layout familiar to the employees. People just couldn’t find the information that they needed. Worse were the stories of the destruction of sophisticated planning systems implemented by human methods. When automation was mandated, the manual procedures were simply too difficult to describe using the programming languages of the day. The only path to automation was to discard the manual methods, which could cripple production.

Turing confronted this contradiction in the ultimate degree after building a machine to break the Nazi’s method for secret communications, known as “Enigma.” If the achievement was to have sustained utility, the Allies’ knowledge of Axis military planning had to be limited: otherwise the Nazis would realize that Enigma had been defeated, and develop a better encryption method. As a consequence, most Allied warriors and civilians facing Nazi assault did so without benefit of the intelligence known to Turing and his team.

While the point is not made obvious, the movie illuminates the personal history that conditioned Turing for his accomplishments. Isolated psychologically from his peers – both by the social stigma of his homosexuality and by what today might be diagnosed as autism or Asperger’s syndrome – Turing was confronted from an early age by the question of what it meant to be human. Was it only the degree of his intelligence that distinguished him from his peers? Or was his intelligence tied to deviant – if not monstrous – behavior? My belief is that these questions were critical motivations for Turing’s drive to understand and simulate intelligence.

That parallels the experience of my father, burdened by his own psychological demons, but also critically concerned that artificial intelligence answer to the authentic needs of the people it empowered. That belief led him to devote most of his life to creation of a universal graphical notation for representation of the operation of systems of:

  • arbitrary collections of people and machines,
  • following programs written in diverse languages.

That technology, now known as Diagrammatic Programming, was recognized by some as the only provably sufficient method for systems analysis. Unfortunately, by the time it was refined through application, the economics of the software industry had shifted to entertainment and the world-wide web. Engineering was often an after-thought: what was important was to get an application to the market, structured so that it held users captive to future improvements. Raw energy and the volume of code generated became the industry’s management metrics.

The personality traits that allowed Turing to build his thinking machines ultimately cost him the opportunity to explore their application. He was exposed as a “deviant” and drummed out of academia. Accepting a course of chemical castration that would allow him to continue his work privately, he committed suicide after a year, perhaps because he discovered that the side-effects made work impossible.

My father was afflicted by childhood polio, and has been isolated for years from his peer group by degenerative neuropathy in his legs.

While my empathy for both of these brilliant men was a trigger for the sadness that overwhelmed me as the final credits rolled, the stories touch a deeper chord. Both were denied the just fruits of their labor by preconceived notions of what it means to be human: Turing because he thought and behaved differently, my father because he attempted the difficult task of breaking down the tribal barriers defined by the languages that separate us.

So what lesson am I to draw from that, as I struggle to prove the truth of the power that comes from a surrender to the purposes of divine love? Is social rejection inevitable when we surrender what others consider to be “humanity”?

Is that not what condemned Jesus of Nazareth? His renunciation of violence and self-seeking? His refusal to fear death?