Algorithms Cannot Consider

Copyright 2020, All rights reserved

A number of people have been laboring for a generation to create algorithms that can respond to real-world situations. Long ago, in a very modest way, I participated in developing one such “expert system” as we called it in those antediluvian days. I mention this to support the validity of my essay.

Our race to create “artificial” intelligence began as an emulation of human reasoning. Early expert systems sought to understand the reasoning of humans who were knowledgeable and experienced, and then code that into the IF…THEN logic of computer programming. Later, more sophisticated approaches sought to use data itself as a way to achieve knowledgeable responses to new information (that was my microscopic experience in this larger movement). More recently, we have sought to devise algorithms to act as agents to sort, categorize, and connect disparate types of input into logical constructs that help us understand the tsunami of data washing over us on a daily basis.

These efforts are possible because of the Web. A friend of mine met Tim Berners-Lee when he made his first sojourn in the US as a technology evangelist spreading the word about his invention, the World Wide Web. My friend was then an undergrad studying information science. Towards the end of Berners-Lee’s presentation, he whispered to his friend sitting next to him, “This changes everything!” And it did. It removed the traditional frictional barriers between information stored in one place and time from access in another. What had been a mountain range of barriers and inaccessible pools of information was transformed into a flat plain where everything was one degree of separation from everything else. The nature of information access and our ability to understand and use it was changed, and not always in a good way. These changes were the foundation of the flat information space we have today. As it grew, however, the premium of speed became more and more important.

In exploiting online information, we need both power and speed. Power to absorb the data using computational statistics that allow us to address the exabytes of data now available — data that will continue to grow with exponential acceleration. But more importantly, at least financially and even socially, we want speed. Michael Lewis ably described the race of financial “quants” to create a faster trading network in his book, Flash Boys: A Wall Street Revolt. Recognizing that speed was the single most important factor in executing trades on Wall St., there was a race to construct a fiber-optic trading network that shaved nanoseconds off stock market transactions to exploit momentary advantages to capture wealth for the network owners. (This was not a race that yielded a net increase in investment capital that could be used for economic growth, but rather rent-seeking — extracting value without actually increasing overall economic well-being. But that’s a whole other topic.)

When considering the growth of social media, power and speed have become a metastatic agent. Here we see human beings directly in the algorithmic mix as millions jostle each other to put their ideas and express their feelings to the fore. Flattening the access to reach every point on the internet, we have created mass movements of opinion that rapidly turn into toxic flash mobs. As we crowdsource our information, too often it turns into mob-sourcing. Is that just because the barriers to others’ views have been flattened? For sure, that’s a factor. But the rise of the bots — those automated entities that can forge connections of sentiment between millions of online actors — have taken speed and mixed it with algorithmic connections to create a new threat to our freedoms and liberty.

The recent primary election win in Georgia of a Republican supporter of QAnon — what was once only a fringe conspiracy community — reveals a consequential danger in computational speed and algorithmic sentiment analysis. QAnon could not exist before the rise of online social media. In the former analog world, such movements were the exclusive realm of distributors of leaflets at the entrances of subway stations or the mutterings of the guy at the end of the bar — people who were routinely ignored. They did not rise to public prominence because their ideas were too harebrained to gain an audience. Hence, the idea of a murderous child molestation ring operating out of a pizza joint’s basement would never been considered real in the public imagination.

Now QAnon can boast that they will have a public supporter in the US House of Representatives, at least that’s the electoral expectation. Even the president complains of conspiracies that are patently false. Yet, these ideas persist and spread.


Let’s give a brief mention to human cognitive fragility since we all believe in myths, those things we believe are true without any basis in fact. We are all susceptible to confirmation bias. But that is just a contextual reality. Why do tragicomic ideas like “pizzagate” get traction at all?

Because they are not considered. Let’s dwell on that term. To consider means to think carefully about, trying to use knowledge, logic, and the time needed to think. In the age of thinking at the speed of clicks, we are impelled to react rather than respond. If it feels right, it must be true. Ironically, in a time when we have unparalleled access to information, we select to ignore evidence and consideration to follow the mob. And that mob has been generated by social media’s algorithms––sentiment connections that amplify the strongest emotions.

We have no time to think partly because it’s not allowed. The onrush of stimuli is so fast and so seemingly urgent, that our brains are short-circuited. We react with the more ancient parts of our mental anatomy that spur a fight or flight reaction. Seated in our chairs, staring at an illuminated screen, we lack the possibility of flight, so we exercise the other option as the only one available to us. We like, clap, favor, denounce, and support by clicking on icons as a substitute to actual discourse. (In the backs of our minds — remember, that’s the old part, we hope our notions will be picked up and amplified by the multitude.) These actions are the opposite of considered thought.

In truth, they are engineered that way. The algorithms of Twitter, Facebook, Instagram, Snapchat, Reddit, and successful social media platforms are engineered to excite our feelings and prompt the adrenal rush that activate emotion over thought. Their algorithms seek to connect our online sentiment with other players’ profiles and history to deliver connections — however ephemeral — to leverage our attention. All so we will see the ads that flow, blink, and pop across our screens. The thing these companies do not want is considered thought because that would reveal the bankruptcy of the ideas their advertisements rely upon.

As a society, however, we still have some instruments available to us to slow this down. They are time-tested but have not been applied to the realm of thinking at the speed of twitch.

In many states in the US, if you make a major purchase like a house, there’s a “cooling off” period after reaching a sales agreement. It varies, but you get a few days to say, “Actually, we really cannot afford a house with a pool.” You might have to sacrifice some earnest money, but you get an opportunity for considered thought. The same legal friction is built into buying some kinds of investments, when public employees move into the private sector, and even when buyers of refrigerators have a change of heart. We have engineered these safeguards into our lives to protect us from the impulses of our primitive brains.

Let’s do the same in social media. Stop a tweet from going into the viral swamp for a bit. Give the person a moment to consider what they are doing before it becomes irrevocable. (Just ask Jerry Falwell, Jr., if he wished he had a moment to consider the wisdom of his recent Instagram post.) Put the post in a holding container to let the poster draw a breath or two before making it public. (By the way, let’s do the same with “reply-all” emails.) If everyone has the same frictional threshold, then no one is advantaged or disadvantaged.

Our heedless rush to participate or influence online has shorn us from our higher brains where logic and knowledge offer wisdom. We are actually becoming dumber as we travel down our current trajectory. Let’s take some time to consider our situation.



Historian, informatician, novelist, and grandfather. Part-time curmugdeon.

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store
David Potenziani

Historian, informatician, novelist, and grandfather. Part-time curmugdeon.