Tag Archives: Brain Function

False Positives and Us

Leave a reply

Cognitive scientists often discuss various forms of cognitive bias. Confirmation bias is just one well-known type (see here). Recognizing cognitive biases in all their forms is really important. But that effectively only focuses on symptoms. not the underlying causes or mechanisms of cognitive biases. In order to better overcome them, we also need understand the mechanisms that give rise to them.

As I discuss at length in my book (see here), our brains are essentially pattern recognition machines. Almost everything we do is a form of pattern recognition. And evolution has tuned our pattern recognition neural networks to err strongly on the side of false positives.

Here’s an example I often use to illustrate the importance of false positives. Imagine when we were evolving as animals. There were real tigers in the forest that were a mortal threat to us. Therefore, our neural networks were trained to recognize even the most vague hint of a tiger in the trees as a real tiger. It did not much matter if we imagined a hundred tigers that were only shadows or leaves blowing in the wind. What was critical however, was that we not miss even one real tiger, no matter how cleverly it concealed itself. An extreme bias toward false positives was a gigantic evolutionary necessity.

The result of all of this natural selection is that today we both benefit from and are hampered by powerful neural networks that are tuned to err strongly on the side of false positives. This is particularly acute when it comes to anything that might threaten us or distress us or make us uncomfortable.

This soft-wiring of our neural networks on the side of false positives not only underpins many of our cognitive biases but has huge ramifications in our social and interpersonal interaction.

For example, false positives certainly bias our perception of any *ism that offends or distresses us. If I am sensitive about my hair, I almost certainly detect far more insensitive comments about my hair than are objectively real. This is true of any *ism that impacts us, whether it be sexism, racism, or any other form of bigotry or hostility. And let me be very clear. All these things do exist and do happen, but I’m making the claim that any given individual almost certainly detects many false positives that are not really incidents of it.

This expands on our usual assessment that I am “sensitive” about my hair. Such prosaic sensitivity can be seen as a another symptom of these underlying mechanics. Our understanding of the false positive bias of our neural networks helps us understand how and why this happens and make us better able to accept it in others and defend against it in ourselves.

This is important because our exaggerated perceptions based on false positives have huge repercussions for individuals and for society. They cause us to react negatively in situations where such a response is actually counterproductive. It also exaggerates our feelings of anger and hostility which not only produce unfortunate behaviors and emotions, but those false positives also act as new legitimate “facts” that “train” our pattern recognition brains to recognize even more extreme false positives. Our biased perceptions and our memories of those false perceptions serve to reinforce our biased neural network in a self-reinforcing feedback loop. Soon we see our *ism everywhere, we hear it in every comment, see it in every glance, and respond with depression and anger which make it still worse.

These same mechanisms play a critical role in our one-on-one interpersonal interactions as well. If our friend or spouse says something we find bothersome or offensive, we quickly become attuned to it and start to see it in every nuance of expression and hear it between the lines in every comment. This reinforces our neural network to become even more sensitized toward it, detecting even more false positives. We can soon get to the point where there is nothing you can say, or even not say, to the listener that is not further evidence to support their feelings. We can quickly become surrounded, even paralyzed by all the tigers in the shadows.

Certainly merely being aware of this mechanism of false positive pattern recognition does not eliminate our susceptibility to all cognitive biases, but I think that understanding how our pattern recognition network functions is essential to protecting ourselves against perceptions that are not realistic or healthy. I know that for me, understanding how I am vulnerable to false positives does not immunize me by any means, but it does help me on many occasions to recognize and to push back against my own pattern recognition biases. And this is true even for perceptions or memories that seem incredibly real and compelling. Having some appreciation, and some humility, with regard to how susceptible we are to false positives can have a tremendous impact for the better.

The Impending Doom of Written Language

Leave a reply

Sci Fi and Fantasy are often lumped together, but they are very distinct literary forms. The core difference is not simply whether the subject matter is dragons or space ships, but whether the subject matter is plausible or not. Whether it could become reality. Dragons could be Sci Fi if originating in a plausible manner and if they adhere to the laws of chemistry and physics. Conversely, a space ship becomes fantasy if it jumps through time and performs “science” feats what would consume fantastically implausible amounts of energy. Lots of Sci Fi fans are actually consumers of fantasy every bit as unrealistic as Lord of the Rings.

Really good Sci Fi is not merely plausible, but likely, even predictive. Great Sci Fi is unavoidable, or more aptly inescapable, given our current trajectory.

But even mind-boggling Sci Fi can often reflect a disappointing lack of imagination.

Take for example the obligatory transparent computer screen that we see in every Sci Fi show. Or even the bigger budget full-on 3-D holographic computer interfaces that provide eye-candy in every major feature nowadays. These look cool, but are probably pretty unimaginative. Plausible and likely, but crude interim technologies at best.

Take for example my own short Sci Fi story Glitch Death (see here). In it, I envision a future in which direct brain interfaces allow people to use computers to “replace” the reality around them with perceptual themes. In that future, we skip quickly past archaic holographic technology and beam our perceptions directly into the brain.

But even that only touches the surface. For example, why would a future direct-to-brain technology be limited to flashing words across our visual field and allowing us to hit “virtual buttons” floating in mid-air? To explain my thoughts on this, let’s digress and talk about math for a moment.

Today we have entered a time where math hardly matters anymore. Oh yes, we must of course understand the concepts of math. We must understand addition, division, and even the concepts of integrals and derivatives and more complex algorithms. But we don’t need to learn or know how to compute them. Not really. We have computers to handle the actual manipulative mechanics of numbers. Most of us don’t really need to learn the mechanics of math anymore, even if we use it everyday.

We are already well on the way there with language as well. We have devices that “fix” all of our spelling and formatting automatically. We don’t actually have to produce typographically correct written text. All we need to do is to communicate the words sufficiently for a computer to understand, interpret, correct, and standardize. We are at the verge of being able, like math, to simply communicate concepts, but not worry about the mechanics of language construction and composition.

So, back to my Sci Fi vision of the future of direct-to-brain interfaces and their likely ramifications. Interfaces like the one envisioned in Glitch Death would soon make written language, and perhaps much of verbal language, prohibitively cumbersome and obsolete. Why shoot words across our visual field, forcing us to read, comprehend, process, and assimilate? Why indeed when the computer could instead stimulate the underlying processed and interpreted symbols directly at their ultimate target destination in our brain. We wouldn’t need to actually read anything. We would simply suddenly know it.

In this situation, we would not need written material to be stored in libraries in any human recognizable language. It would be more efficiently housed in computer storage in a language-independent format that is most closely compatible with and efficiently transferrable into the native storage of the same concepts in the human brain.

In this future, all of which is directly in our path of travel assuming we survive our own follies, we deal at basic symbolic levels and tedious processes of math and language become largely offloaded. Forget tools to translate human languages. We will be able to simply discard them for a symbolic language that essentially transforms us into telepathic creatures. And in this form of telepathy, we don’t hear words in our head. We just transmit ideas and thoughts and understanding and experiences with the aid of our computer interfaces. The closest depiction in popular Sci Fi is perhaps the implantation of memories in the 1990 film “Total Recall.”

A real fascinating unknown to me is, how would humans process and interact without language? Do we require at least an internal language, internal dialogue, to function? I have always wanted to be a subject in an experiment to be made to forget all language, say by hypnosis or drugs, and to experience functioning without it. Like a dog might process the world. Technology may inevitably force that experiment upon us on a huge social scale.

It’s not true that “A sufficiently advanced technology is indistinguishable from magic.” Magic would defy the fundamental restrictions of physics and chemistry. That’s how we’d know the difference. A telepathic future facilitated by direct-to-brain computer interface is Science Fiction, not Fantasy.