ESSAY FRIDAY | Where technology can’t beat humans

A robot helping medical teams treat coronavirus patients, Italy (REUTERS/Flavio Lo Scalzo)

By Chris Chenga

By 5pm on Friday January 10, it had been an entire business week. Five whole days, eight working hours each, had gone by.

Since Monday the 6th at 8am, I was able to do the following; settle the upcoming month’s work bills, evaluate a range of newly offered packages for business such as banking services, internet and communication, organize work formalities, separating filings of proprietary content from shared folders with clients or third-party service providers and signed up and learnt new editing software.

Tech savvy readers would interject, ‘it took you five whole days to get this short list done?’ Well, the fascination here is that not only did I get it done in five days, I did it all without any human interaction! These tasks are by no means trivial. My entire business model and how I do work had been re-arranged, for greater efficiency and productivity.

All the while, I conducted these tasks with relentless curiosity on matters that before the fact, I had complete unfamiliarity with. Yet, at the end, I was fully satisfied and informed, without-any-human-interaction (WAHI).

Regardless of my fascination in these events that opened my year, I remain moderate towards technology. More recently, while pondering on technology’s influence on modern economies, two events took place tilting scales in my mind that for some things, technology cannot serve greater purpose than fellow humans can for one another; definitely not yet!

Luddites and Cargo Cults

Opinions on technology differ greatly, and are often as polasizing as any other subject matter that people can disagree on. Whatever our modern-day attitudes are towards technology, in its broad application, they probably will not become folklore like those of Luddites and Melanesians.

Luddites were an oath conscripted society made up of textile workers in the 19th century in the United Kingdom. As the industrial revolution took hold, Luddites destroyed the new machinery in cotton mills driven by the belief that machinery would ultimately take away their jobs. Luddites further assumed that it was completely not worthwhile to learn how to operate the machinery, as supposedly, training to operate the machinery was merely a bluff by mill owners trick workers into accepting the machinery.

The Luddites actually got a significant movement going that lasted at least five years from 1811 to 1816. Tensions heated to a point where military intervention by British authorities was necessary to protect the cotton mills that were such a big part of the economy.

The Luddites grew influence beyond the cotton mills, inspiring other workers in a broadly agrarian economy to prefer hard labour practices using basic tools as opposed to more physically forgiving means like using ploughs.

In the present day, Luddites have become synonymous with disdain towards technology.

At the contrasting end are Melanesians. This tribe gained its own notoriety in the 20th century during the second world war, though it would later be superseded by the general reference of “cargo cults”. So fascinated by technological advancement were Melanesians that they built mimic representations of aeroplanes, military equipment, and various cargo that would be offloaded by battling sides that landed into their territories on the Pacific front.

After seeing American military set up bases and landing strips to supply its infantry, the Melanesians would go on to build their own aeroplanes from material such as wood planks, sticks, and leaves. One can imagine that not so many of these took off, fired ammunition, or offloaded vital supplies to sustain livelihood.

However, the Melanesians tried regardless, and today they represent an opposite appetite for technological advancement to that of Luddites. 

The decision-making revolution

The fourth industrial revolution has been gathering some excitement for a few years now, especially at the advocacy of different clusters of interests; finance, security, medical, and industrial.

Underlying the enthusiasm of this particular revolution, and making it distinct in its forthcoming, is that it greatly leverages its contributions on computers’ ability do make decisions!

Algorithms and artificial intelligence are buzz words that are at the forefront of this distinctive offering. While the predecessor revolutions enhanced productivity, the impending revolution is nuanced in its quest to enhance decision making by replacing human discretion.  This is an under noticed aspect of the fourth industrial revolution.

Merely perceiving the revolution as improving productivity misses its unique proposition. Indeed, that I can reconfigure my entire business model to improve efficiency isn’t different from any acknowledged pre-existent revolution. By any scale of productivity such advancement fails to surpass the exponential progression of, say, air travel from maritime means, or the enhanced storage capacity of the microchip compared to 1960s mainframes.

Yet, neither jump in productivity offered by those two creations warranted an acknowledgement of a newly distinct revolution. Any fascination and awe to my opening week’s exploits, WAHI, is reserved for the paragons of the third industrial revolution.

What gives emphasis to a fourth industrial revolution is the decision-making offered by new creations. It is important to highlight discretion proposition because, whether intentionally or mere incidence, much of the hype of a fourth industrial revolution is sold on the already existent infrastructure of the third industrial revolution.

One can wonder why that is the case? Is the clear distinction that a fourth revolution offers replacing human discretion with that of programmed computers less palatable? After all, this is really the difference between now and a potential future.

The other distinctive feature of a proposed industrial revolution of now is that it is occurring after globalisation. More ‘tribes’ and societies – even oath conscripted – are potentially empowered with a greater say in the acceptance of this revolution.

The fourth industrial revolution is subservient to numerous legislative barriers, much more than preceding industrial revolutions ever were. In today’s democracies, this implies greater a potential for social participation in its ushering in.

Tension between citizens and governments is good

Tension between technology and legislation should be encouraged and continuously moderated. The arguments swing both ways. For instance, technological progression is often synonymous with human progression.

It is very easy for societies to be caught up by the waves of excitement that bring in a tide of technological offerings. As many inventions are commercial offerings, technology is accompanied by well sponsored PR and marketing campaigns. Many societies, worried about getting left behind by the rest of the world, have a propensity to chastise what they perceive as slow administrations that delay new technologies. 

In 2017, the Reserve Bank of Zimbabwe issued a circular to banks instructing them not to offer their institutional services to virtual currency exchanges and traders. This was unpopular within tech communities in the country. Weeks later, the enforced shut down of a bitcoin exchange called Golix was received with contempt by the public.

The country’s central bank shared the same suspicions held by former President of the European Central Bank, Jean-Claude Trichet, on cryptocurrencies. The suspicions are focused around client security, as well as monetary implications of cryptos on a national economy.

“The blockchain hype has now been subdued into more humble articles…”

The RBZ and Trichet were validated when news broke that clients of a company called Quadriga CX could not access up to US$250 million in their accounts. After several incidents of a similar nature occurred, the tides that once loudly roared for the arrival of blockchain calmed rather noticeably into serene seas.

From being the next sure tech innovation accompanied by PR and marketing on global magazine covers and keynote topics at world conventions, blockchain hype has subdued to more humble articles on the relatively few entities that have incorporated the technology, gesturing for consideration of how the technology could be useful to other industries.

In contrast, sometimes governments are way ahead of citizens and it is citizens who hold suspicions. In May 2018, Quartz Africa ran a story purporting that the Zimbabwean government had signed MoUs for facial recognition software with CloudWalk Technology in China. This wasn’t well received by citizens who immediately grew a consciousness to the potential legality of such technology. The Government of Zimbabwe has since been silent on the matter.

Why focus on computer decision-making?

Some readers may be tempted to chime in with a suggestion that computers already make decisions. Sure, they do. In limited ways, they do. And that is what the fourth industrial revolution hopes to surpass.

Binary decision making isn’t new. Computer devices have long been programmed to say yes or no, choose between one decision or another. So, the discourse around the fourth revolution should be about the extended latitude of decision-making apportioned to computers.

For instance, artificial intelligence. Such consideration would only be necessary after identifying, and reaching exhaustion, in instances where we can no longer rely on human cognitive and sentimental guidance. If so, yes, maybe then we can give A.I. a try.

Yet, two very recent and ongoing events reveal circumstances where computers cannot help us any more than focusing on improving human discretion can.

In March 2020, the Ministry of Finance amended the Securities Exchange Act to include exchange traded funds, ETFs. Market analysts were pleased with the initiative as it showed a movement towards global practices. ETFs now make up 42% of all U.S. fund assets.

“Computers can make better decisions on your portfolio than the human fund manager, but…”

But, is that enough cause to adopt a technology? That they are globally popular? ETFs are technology driven open ended investment schemes that seek to replicate the performance of targeted indices. For example, the S&P500 is an index of stocks. ETFs are part of what is called passive investing whereby you can simply join an ETF, which then uses algorithms to construct a portfolio of stocks for you.

ETFs mainly use smart beta – algorithms and statistical models – to manage the risk appetite of your portfolio. The appeal of ETFs is that computers can make just as good, if not better, decisions on your portfolio than the average fund manager.

Here is the irony; ETFs assumes the efficient market hypothesis. The reason why ETFs track indices, or market determined stock prices, is that they assume a market of highly efficient investors functioning under competent macroeconomic management. You can make your own appraisal here of ETFs in Zimbabwe. I will offer two questions to guide you:

  • are Zimbabwean capital markets efficient in price discovery?
  • is the Zimbabwean economy under competent macro-economic management?

At the start of 2019, Fitch removed Zimbabwean stocks from its indices, mostly due to the kind of volatility that at one point saw the ZSE almost doubling its value over five trading days.

The decision last month to halt the fungibility of certain stocks may be some form of admission by policymakers that price discovery on the local bourse is distorted and market efficiency could not be counted on to close out the arbitrage. Welcome ETFs into Zimbabwe if you will, but technology in this case cannot serve greater purpose than improving human discretion.

Is it not the case that overly welcoming analysts and participants are conducting a cargo cult enthusiasm to technology?   

Humbled by Covid-19

It seems that it is when we are at our most vulnerable that we appreciate human discretion. A report by Capgemini titled ‘Capital Market Trends: 2020’, released at the beginning of the year, highlighted Artificial Intelligence as a means to improving pricing decisions, Machine Learning to boost automation in trading, and Robotic Process Automation as a means to get clients onto trading platforms.

These are practical applications of the fourth industrial revolution expected to have had an effect on capital markets in years to come. Nobody had foreseen COVID-19. If computer discretion was as good as sold, the subsequent carnage on global capital markets would not so desperately yearn for human discretion as it is today.

Larry Fink of Blackrock, in a letter published this week, suggests we are in a crisis of confidence, not a financial crisis.

Howard Marks, who has long been sceptical of an over reliance on computers as expressed in his letter titled ‘Investing Without Humans’ penned in 2011 argued that technology cannot measure confidence.

It surely cannot use algorithms to measure the corporate culture he added. Larry Fink identified culture as a determinant for companies that will see through the Covid-19 pandemic. Computer proponents will remain mute for the time being as investors seek out the best in human discretion to figure out how to ease Covid-19’s contagion on value destruction. 

We can improve human decision making

I think before a fourth industrial revolution, humans should strive to improve our cognitive and sentimental discretion. Instead of using our average as justification to invest in artificial intelligence, perhaps we can perceive our average as a condition that can always be improved.

The Flynn effect refers to the gains in IQ scores for the average human being in the last century. If you scored people in 1920 based on modern intellectual norms, the average human being would have scored an IQ of 70; which is near mentally retarded. If you scored people today on the intellectual norms of 1920, the average human being would score an IQ of 130; mentally gifted. So human beings are getting smarter.

Daniel Goleman wrote a book titled ‘Focus: The Hidden Driver of Excellence’ presented a strong case that most human beings are far smarter and capable than they are today, if only they engage relentless focus. This is backed by both cognitive science and sentimental motivation that makes us commit to a given diligence. The average can be improved.

Perhaps one can remain stubbornly planted on the assertion that there are instances where we can no longer rely on human cognitive and sentimental guidance. That the average application of humans isn’t good enough. Thus, a fourth revolution of artificial intelligence is necessary.

I am not sure that I can ever agree with this. Outperformance has been the consistent in human progression and wealth creation throughout our species history. It remains so. In 2019, UBS published a report titled ‘The Billionaire Effect’ in which studies reveal that over the 15 years to the end of 2018, billionaire-controlled companies listed on equity markets returned 17.8%, almost twice the annualized average performance of the market of 9.1%.

The study identifies three traits that have driven growth in their companies: smart risk taking, business focus, and determination. Moreover, by referencing that the number of women billionaires rose from 160 to 233 in the five years to 2018, the report adds credence to the fact that when humans are offered opportunities beyond our own structural confines; whether social, cultural, or administrative, human capability can shoot ahead of the present mean into the outliers. 

I am not convinced that technology can beat human discretion.

Adherents to such a notion have probably been looking at the wrong representation of human capability. As Rory Sutherland put it, ‘metrics, and especially averages, encourage you to focus on the middle of a market, but innovation happens at the extremes. You are more likely to come up with a good idea focusing on one outlier than on ten averages… it is perfectly possible that conventional market research has, over the past fifty years, killed more good ideas than it has spawned, by obsessing with a false idea of representativeness.’