Do You Treat Employees Like Fixed-Program Computers?

Computers didn’t always work they do today. The first ones were what we now called “fixed-program computers,” which means that, without some serious  and complex adjustments, they could do only one type of computation.

Sometimes that type of computer was superbly useful, such as when breaking Nazi codes during World War II (see the bombe below). Still, they weren’t much more programmable than a calculator, which is a kind of modern-day fixed program computer.

The brilliant mathematician John von Neumann and colleagues had a different vision of what a computer should be. To be specific, they had Alan Turing’s vision of a “universal computing machine,” a theoretical machine that the genius Turing dreamt up in 1936. Without going into specifics, let’s just say that the von Neumann model used an architecture has been very influential up the present day.

One of the biggest advantages associated with Turing/von Neumann computers is that multiple programs can be stored in them, allowing them to do many different things depending on which  programs are running.

Von Neumann architecture: Wikimedia

Today’s employers clearly see the advantage of stored-program computers. Yet I’d argue that many treat their employees and applicants more like the fixed-program computers of yesteryear.  That is, firms make a lot of hiring decisions based more on what people know when they walk in the door than based on their ability to acquire new learning.  These days, experts are well paid largely because of the “fixed” knowledge and capabilities they have. Most bright people just out of college, however, don’t have the same fixed knowledge and so are viewed as less valuable assets.

Employers aren’t entirely in the wrong here. It’s a lot easier to load a new software package into a modern computer than it is to train an employee who lacks proper skill sets.  It takes money and time for workers to develop expertise, resources that employers don’t want to “waste” in training.

But there’s also an irony here: human beings are the fastest learning animals (or machines, for that matter) in the history of, well, the universe, as far as we know. People are born to learn (we aren’t designated as sapiens sapiens for nothing), and we tend to pick things up quickly.

What’s more, there’s a half-life to existing knowledge and techniques in most professions. An experienced doctor may misdiagnose a patient simply because his or her knowledge about certain symptoms or treatments are out-of date. The same concept applies to all kinds of employees but especially to professionals such as engineers, scientists, lawyers, and doctors. In other words, it applies to a lot of the people who earn the largest salaries in the corporate world.

Samuel Arbesman, author of The Half-Life of Facts: Why Everything We Know Has an Expiration Date, stated in a TEDx video, “Overall, we know how knowledge grows, and just as we know how knowledge grows, so too do we know how knowledge becomes overturned. ” Yet, in our recruitment and training policies, firms often act as if we don’t know this.

The only antidote to the shortening half-life of skills is more learning, whether it’s formal, informal or (preferably) both. And the only antidote to a lack of experience is giving people experience, or at least a good facsimile of experience, as in simulation-based learning.

The problem of treating people like fixed-program computers is part of a larger skills-shortage mythology. In his book  Why Good People Can’t Get Jobs , Prof. Peter Cappelli pointed to three driving factors behind the skills myth. A Washington Post article sums up:

Cappelli points to many’s unwillingness to pay market wages, their dependence on tightly calibrated software programs that screen out qualified candidates, and their ignorance about the lost opportunities when jobs remain unfilled…”Organizations typically have very good data on the costs of their operations—they can tell you to the penny how much each employee costs them,” Cappelli writes, “but most have little if any idea of the [economic or financial] value each employee contributes to the organization.” If more employers could see the opportunity cost of not having, say, a qualified engineer in place on an oil rig, or a mobile-device programmer ready to implement a new business idea, they’d be more likely to fill that open job with a less-than-perfect candidate and offer them on-the-job training.

The fixed-program mentality should increasingly become a relic of the past. Today, we know more than ever about how to provide good training to people, and we have a growing range of new technologies and paradigms, such as game-based learning, extended enterprise elearning systems, mobile learning and “massively open online courses” (aka, MOOCs).

A squad of soldiers learn communication and decision-making skills during virtual missions: Wikimedia

With such technologies, it’s become possible for employers to train future applicants even before they apply for a position. For example, a company that needs more employees trained in a specific set of programming languages could work with a provider to build online courses that teach those languages. Or they could potentially provide such training themselves via extended enterprise learning management systems.

The point is that there are more learning options today ever before. We live in a new age during which smart corporations will able to adopt a learning paradigm that is closer to that of stored-program computers, one that they’ve trusted their technologies to for over half a century.

Featured image: A rebuild of a British Bombe located at Bletchley Park museum. Transferred from en.wikipedia to Commons by Maksim. Wikimedia Commons.

Talking Drums and the Depths of Human Ignorance

It’s a small but genuine annoyance. I’ll be listening some “expert,” often a professor, being interviewed for a radio show or podcast. If the idea of cognition comes up, they’ll state as a fact that humans are far more intelligent than any other animal on the planet. And, almost inevitably, one piece of evidence they’ll point to is communication. There’s the assumed inability of other animals to communicate with as much sophistication as we do.

Now, they might be right about these things, though obviously we’d need to define intelligence and communication to even establish a working hypotheses. What irritates me, though, is the certainty with which they make their claims. In truth, we just don’t know how we stack up in the animal kingdom because we still live in such a deep state of ignorance about our fellow creatures.

The Talking Drums

When I hear such claims, I sometimes think about the talking drums. For hundreds of years, certain African cultures were able to communicate effectively across vast distances. They did this right beneath the noses and within the hearing of ignorant, superior-feeling Europeans.

In his book The Information, James Gleick lays out the story of the talking drums in Chapter One. Via drums, certain African peoples were able to quickly communicate detailed and nuanced messages over long distances well before Europeans acquired comparable technologies. At least as far back as the 1700s, these African peoples were able to relay messages from village to village, messages that “could rumble a hundred miles or more in a matter of an hour…. Here was a messaging system that outpaced the best couriers, the fastest horses on good roads with say stations and relays.”

It was only in the 19th century that the missionary Roger T. Clarke recognized that “the signals represent the tones of the syllables of conventional phrases of a traditional and highly poetic character.” Because many African languages are tonal in the same way Chinese is, the pitch is crucial in determining the meaning of a particular word. What the drums allowed these peoples to do was communicate complex messages using tones rather than vowels or consonants.

Using low tones, the drummer communicates through the phrases and pauses. Extra phrases are added to each short “word” beaten on the drums. These extra phrases would be be redundant in speech, but they can provide context to the core drum signal.

Enormous Chasms

The technology and innovativeness of the talking drums is amazing, of course, but what’s especially startling is the centuries-long depth of European ignorance about the technology. Even once some Europeans admitted that actual information was being communicated across vast distances, they could not fathom how.

Why? Sure, racism no doubt played a part. But the larger truth is that they simply didn’t have enough information and wisdom to figure it out. That is despite the fact that we are talking about members of the same species and, indeed, a species with very little genetic diversity.

Here’s how the Smithsonian Institution reports on this lack of diversity:

[C]ompared with many other mammalian species, humans are genetically far less diverse – a counterintuitive finding, given our large population and worldwide distribution. For example, the subspecies of the chimpanzee that lives just in central Africa, Pan troglodytes troglodytes, has higher levels of diversity than do humans globally, and the genetic differentiation between the western (P. t. verus) and central (P. t. troglodytes) subspecies of chimpanzees is much greater than that between human populations.

On average, any two members of our species differ at about 1 in 1,000 DNA base pairs (0.1%). This suggests that we’re a relatively new species and that at one time our entire population was very small, at around 10,000 or so breeding individuals.

For Europeans to remain so ignorant about a technology created by other members of their own barely diversified species tells us how truly awful we are at understanding the communication capabilities of others. Now add in the exponentially higher levels of genetic diversity between species. For example, the last known common ancestor between whales and human existed about 97 million years ago. How about the last known ancestor between birds and humans? About 300 million years ago.

These timescales represent enormous genetic chasms that we are not remotely capable of bridging at the moment. We are still in the dark ages of understanding animal cognition and communication. So far, our most successful way of communicating with other animals is by teaching them our languages. So now we have chimpanzees using sign language and parrots imitating our speech patterns.  African Grey parrots, for example, can learn up to 1,000 words that they can use in context.

Yet, when these species do not use human language as well as humans, we consider them inferior.

If We’re So Bloody Bright…

But if we as a species are so intelligent, why aren’t we using their means of communication? I’m not suggesting that other animals use words, symbols and grammar the way humans do. But communicate they do. I live in Florida, which is basically a suburbanized rainforest, and have become familiar with the calls of various birds, tropical and otherwise. One of the more common local denizens is the fish crow. I hear crows that are perched blocks away from one another do calls and responses. The calls vary considerably even to my ignorant, human ears, and there are probably countless nuances I’m missing.

Are they speaking a “language”? I don’t know, but it seems highly unlikely they’re expending all the vocal and cognitive energy for no reason. Their vocalizations mean something, even if we can’t grasp what.

Inevitably, humans think all animal communication is about food, sex and territory. But that’s just a guess on our part. We assume that their vocalizations are otherwise meaningless just as many Europeans assumed the talking drums were mostly meaningless noise. In short, we’re human-centric bigots.

Consider the songs of the humpback whales. These are extremely complex vocalizations that can be registered over vast distances. Indeed, scientists estimate that whales’ low frequency sounds can travel up to 10,000 miles! Yet, we’re only guessing about why males engage in such “songs.” For all we know, they’re passing along arcane mathematical conceits that would put our human Fields Medal winners to shame.

On Human Ignorance

The point is that we continue to live in a state of deep ignorance when it comes to other our fellow creatures. That’s okay as long as we remain humble, but we humility is not what people do best. We assume we are far more intelligent and/or far better communicators than are other species.

Yet, consider the counterevidence. Just look the various environmental, political and even nuclear crises in which we conflict-loving primates are so dangerously enmeshed. It hardly seems like intelligence. Maybe the whales and parrots are really discussing what incapable morons humans are compared to themselves. With that, mind you, it would be hard to argue.

Featured image from Mplanetech. 11 January 2017

Thinking About Thinking

What is thinking?

There has been a tsunami of articles related to cognition. How does your pet think? How (or should) we build thinking machines? How can you think more effectively? How can intelligence itself be boosted? Etc.

This got me thinking about thinking, so I became involved in several social media discussions on how we should view the thinking process. Below is a short definition I’ve arrive at, one that potentially includes cognition among many animals as well as, perhaps, computing devices today and/or in the future:

Thinking is the process of assimilating sensory information, integrating it into existing internal models of reality (or creating new models derived from old ones),  generating inferences about the past, future and present based on those models, and using those inferences as more input that can be assimilated into internal models via continuing feedback loops.

This is succinct but I’m sure it oversimplifies things. For example, infants are likely born with a certain amount of “hard-wiring” that allows them to interpret the world in basic ways even before they’ve developed many internal models about how the world works.  Still, I’d argue that this definition gets at what we mean by thinking, whether it relates to bugs, birds, elephants or hominids.

What’s the point? Well, cognition is quickly becoming the name of the game in modern society in nearly any discipline you can name: learning, artificial intelligence, information science, bioethics, research, analytics, innovation, marketing, justice, genetics, etc.

A lot of what we will be doing in the future is trying to answer hard questions about thinking:

  • What (and how) do other people (e.g., customers, employees, citizens, etc.) think?
  • How can we make learning more efficient and effective?
  • How can we make machines that are better at solving problems?
  • How can we understand what is in the minds of criminals so that we can reduce crime and make better decisions in our justice systems?
  • How should we view and treat other thinking animals on the planet?
  • How do we know (or decide) when machines are thinking, and to what degree is thinking different from consciousness?

To have better discussions around these and similar questions, we’ll need to develop better and more understandable cross-disciplinary definitions of terms such as thinking, consciousness (which seems to be a kind of attention to thinking), and comprehension. A lot of progress comes from our growing ability to create thinking machines, but we also seem to be getting considerably better at understanding human cognition as well. The next couple of decades or so should be interesting.

(Note: I wrote a version of this post nearly a decade ago.)

Author: Solipsist;
From Wikimedia Commons.
Featured image source: Robert Fludd. From https://commons.wikimedia.org/wiki/File:Robert_Fludd,Tomus_secundus…,_1619-1621_Wellcome_L0028467.jpg