Artificial Intelligence vs Human Intelligence: Who Takes the Cake on Indonesia’s Bureaucracy?
The use of technology to improve human life and activity has long been implemented. Nowadays, we see technological innovations beyond what our predecessors could have ever imagined. Take traveling for example. People use to travel by foot or riding an animal of some sort. Then comes the invention of carriages with (again) animals to pull it. Many years later, we now have cars–which is basically an automated carriage if you think about it–, trains, ships, planes, and all other sorts of vehicle I haven’t mentioned. This is only on transportation technology. There are still many examples that I can pull out of my magic pocket, but we can save that for another day because today, we are going to talk about artificial intelligence (AI). Well, I believe a disclaimer is to be noted, I’m not here to explain what an AI is–that’s for the engineers, I’m a politics and government undergrad. Instead, I’m going to (attempt to) compare artificial intelligence with human intelligence and see which one is more suitable for Indonesian bureaucracy.
Some of you may ask, “what’s with the weirdly specific topic?” Well, allow me to shed a little light on that. The topic of using AI in Indonesia’s administrative procedure isn’t something new. Indonesia’s President, Joko Widodo, first shared his big plans on reforming Indonesia’s bureaucracy back in 2019. President Joko Widodo wanted to switch eselon III and eselon IV with AI. Eselon is a rank/position in Indonesia’s civil servant profession. Joko Widodo’s ambition to automate a big portion of Indonesia’s bureaucracy was quoted as “a radical plan” by local news.
Indonesia’s Pekerja Negeri Sipil aka PNS
To help understand how big of a change this plan is, we need to first understand the hierarchy of Indonesia’s civil servant. In Indonesia, we call our civil servant with the term Pekerja Negeri Sipil (PNS). To make things simpler, I would be referring to Indonesia’s civil servant profession as PNS. PNS has two different group divisions. The first one being the division of ranks. While the second one is the division of eselon. The division of ranks shows that there are four ranks in total: golongan I, golongan II, golongan III, and golongan IV. These golongan divisions correlates with how much wage a PNS gets. The higher the rank (golongan) the higher the pay. On the other hand, we have eselon. Unlike golongan, the division of eselon is a structural one. Meaning, a PNS is considered an eselon when they hold a position in the institution’s structure. Eselon is divided into four groups: eselon I, eselon II, eselon III, and eselon IV. As the name suggests, eselon I would be the highest in position followed by eselon II, eselon III, and finally eselon IV. Each eselon work in its own administrative scope. Eselon I does the big thinking. They decide the main policies for the success of the institution’s both long- and short-term goals. Eselon II works to create strategies, implement it, and further develop policies from eselon I. Eselon III is responsible for strategy preparation and realization from eselon II. While eselon IV is responsible in operating eselon III’s strategies and plans.
Now that we have covered the boring part, do you guys see just how BIG of a change President Joko Widodo is suggesting? Completely automating eselon III and IV in all of Indonesia’s institution–which is a lot by the way. This is why this topic has caused quite an uproar amongst Indonesian citizens with the million-dollar question: is AI really the answer?
From Street-Level Bureaucracy to Systems-Level Bureaucracy
Back in 1980, Lipsky wrote about the street-level bureaucracy. These are the bureaucrats that directly provides government services, benefits, and punishments to the public. You might know these street-level bureaucrats as police officers, doctors, fire fighters, postmen, or many more. They are here to serve and solve problems for the public. As Herbert Simon (1947) best implies, street-level bureaucracy highlighted the human aspect of decision making in a bureaucratic process. A fancy word for it: human discretion.
In 2002, Bovens and Zouridis noticed a switch from street-level to screen-level and finally to systems-level bureaucracy. The rise of technology, especially computers, plays a big role in this shift. Bureaucrats are no longer ‘on the streets’ as Lipsky suggests. Information Communication Technology (ICT) tools now bridges bureaucrats and the public. Often time, people no longer need to interact with these street-level bureaucrats. A click on their computer or phones would do the trick. Thus, the name screen-level bureaucracy emerges. As the world makes way for screen-level bureaucracy to flourish, the public service world soon changes into systems-level bureaucracy. ICT tools don’t just help connect bureaucrats and the public, in some cases it began replacing the role of expert judging (Bovens & Zouridis, 2002 in Bullock, 2019).
What the heck is Artificial Intelligence?
Now, let’s talk about AI. As quoted from Nilsson’s book “The Quest for Artificial Intelligence A History of Ideas and Achievements”, AI lacks a universal definition. But Nilsson himself defines AI as an activity devoted to making machines intelligent, and intelligence as a quality that enables an entity to function appropriately with foresight in its environment (Nilsson, 2010). Max Tegmark (2017) on the other hand defines AI as the ability of nonorganic, mechanical entity to accomplish complex goals. Tegmark’s definition is further explored by Justin B. Bullock on his article about AI, discretion, and bureaucracy. According to Bullock (2019), Tegmark’s definition on AI implies that intelligence is “substrate independent” meaning it can exist both as a biological and mechanical process. This goes in line with what Nilsson was talking about in his definition on intelligence. Based on Nilsson’s approach, many things may fall in the spectrum of ‘intelligence’. Animals, to some degree, have their own intelligence. Machines, also to some degree, have their own intelligence. Humans, as the most complex being in existence, can be assumed to be the most intelligent of all.
Continuing the discussion about machines or mechanical entity with intelligence, comes the question: what is considered a machine? Nilsson (2010) answered this question rather poetically. It would be an academic crime if I don’t quote his original words here.
“To many people, a machine is a rather stolid thing. The word evokes images of gears grinding, steam hissing, and steel parts clanking. Nowadays, however, the computer has greatly expanded our notion of what a machine can be. A functioning computer system contains both hardware and software, and we frequently think of the software itself as a “machine.” For example, we refer to “chess-playing machines” and “machines that learn,” when we actually mean the programs that are doing those things. The distinction between hardware and software has become somewhat blurred because most modern computers have some of their programs built right into their hardware circuitry.”
What I get from Nilsson’s answer is that machines are really the hardware and software altogether. The hardware are those parts that you can touch and see. While the software are the programs that makes the machine works.
Now that we know what AI, intelligence, and machines are, it’s time to answer THE question: is AI really the answer for Indonesia? My short answer would be No. Bummer, I know. But hey, I can think of two main reasons why this big AI dream in Indonesia is probably not going to turn out the way it’s planned: education and economy. Both reasons intertwined with one another in a way. I will try my best to explain below.
Education and Economy: A Bundle of Issues
It doesn’t take a data scientist to tell that Indonesia lacks in these departments: education and economy. A June 2018 World Bank report called “Indonesia Economic Quarterly” highlighted one of Indonesia’s biggest problems: its human capital quality. In this report, World Bank argues that Indonesia’s long-term growth potential and quality of life is highly dependent on the quality of its human capital. It has been proven that poor human capital quality leads to low labor productivity, limited contribution of education to economic growth, and lower overall competitiveness. The challenge that Indonesia faces is that its education–the most important factor to ensure a high-quality human capital–is lacking. Although Indonesia had implemented many policies to reform education access and quality for the past fifteen years or so, the results are mixed. Schooling achievements grew significantly. But student learning stays below other Southeast Asia countries.
A 2018 Organization for Economic Co-operation and Development (OECD) Programme for International Student Assessment (PISA) survey shows that Indonesia scores an average of 371 points in reading, 379 points in mathematics, and 396 points in science. For comparison, OECD’s average for each category is: 487 points in reading, 489 points in mathematics, and 489 points in science. Indonesia is around 100 points behind OECD’s average. For further comparison, let’s check out our neighboring countries’ PISA score. In a 2018 OECD PISA survey, Singapore scores incredibly in each category: 549 points in reading, 569 points in mathematics, and 551 points in science. Each score far exceeded OECD’s average. Also in a 2018 OECD PISA survey, Malaysia scores an average of 415 points in reading, 440 points in mathematics, and 438 points in science. Indonesia is roughly 50 points behind Malaysia in all categories. Thailand’s 2018 PISA scores 393 points in reading, 419 points in mathematics, and 426 points in science. These numbers are still higher than Indonesia’s 2018 PISA score. I will provide a graph of each country’s score down below to give a better view and understanding. I will also include other available Southeast Asia countries, such as Brunei Darussalam’s and Philippines’, 2018 PISA score as comparison.
In case you don’t know, PISA is OECD’s program to assess 15-year-olds’ ability to use their reading, mathematics, and science knowledge and skills to meet real life challenges. As per OECD’s report, currently, there are 93 countries participating in PISA. The next PISA assessment is scheduled to be done in 2022.
All these human capital quality and PISA score talks can be summarize like so: Indonesia’s education is below average, be it compared to OECD’s standards or compared to neighboring Southeast Asia countries. A below average education translates into low quality human capital which then translates into low economic level. Now what does all these have to do with our main topic of AI and bureaucracy? Well, first thing first, AI is a mix of much knowledge scattered in many areas. Nilsson (2010) refers to these knowledges as ‘clues’ needed to make machines intelligent. Areas such as philosophy, logic, biology, psychology, statistics, and engineering all inspired and created AI. It is not an overstatement to say that AI is a very complex machine that requires extensive knowledge and funding not only in its creation, but also in its operation. It would take some degree of education and economic freedom to even begin to understand AI.
My subjective view on AI is it looks and sounds (and probably is) like a next level, futuristic technology. Don’t take it the wrong way, I’m all in for technological innovations and improvements! But geeking over an innovation and implementing it are two very different things. Using the newest technology is not as easy as it seems. Besides having to spend probably half of your live savings to access these brand-new techs and youtube-ing the how to, we need to take account the people’s ability-slash-capacity to operate these technologies. I’ve seen people confused on how to start a video call on their phones or check emails on their laptops. This is on an individual level. Take it up a notch to bureaucracy level then you’ll see instances where people struggle to get their queue number in hospitals, struggle to print their boarding passes at the airport, struggle to apply for a service online, and the list goes on. These are simple, real examples that I, you, we encounter in our day-to-day life. Those people who struggled were lucky because there were other workers–human workers–that was ready to assist them. Can you imagine what would happen if these workers were to be switched with AI?
The issue doesn’t stop here. Just like how humans have their own capacities on understanding technology, AI has its own capacities on understanding humans. This is best explained by Bullock (2019) in his 2 x 2 matrix. Bullock argues that two things matters when it comes to implementing AI in bureaucracy: the context of the task and the type of the task. Two assumptions are then added to help identify the tasks. First, tasks that rarely deviate from normal procedures, less complex, and more routine making it more ‘certain’. Second, tasks that often deviate from normal procedure and is more complex making it more ‘uncertain’. Tasks with low complexity and high certainty are seen to be best fitting for an AI switch. While tasks with high complexity and low certainty are best for humans. The ones in the middle are determined by the certainty of the task. A task can be high in complexity but if its high in certainty too, then Bullock suggests it can lean more towards AI. A task with low complexity but low in certainty should lean more towards human.
This matrix Bullock made may help decide which is really best for Indonesia’s bureaucracy. President Joko Widodo suggests changing eselon III and IV, eselon that are responsible for strategy preparation, realization, and operation. These are broad scopes of tasks. It’s high in complexity and most certainly uncertain. These set of responsibilities that the eselon need to do are too unspecific. Therefore, switching all eselon III and IV into AI is not even supposed to be a conversation. It’s too far stretched.
The implementation of AI in Indonesia’s bureaucracy, especially to switch a huge part of human workers in the structure (eselon) is not feasible. Indonesia’s level of education and economy is not there yet to support such huge change. The people lack the basic education to even begin to comprehend these upgrades. Economic level plays a huge part too in terms of access towards better quality education and technology. A low education score, bad economy, and AI is not the best meal combo out there. Not to mention, AI technology itself might not even be able to perform the tasks that are done by these eselon because AI has its own limitations.
To conclude this rather long writing, I believe that human intelligence will always trump a machine’s. It is true that us humans has uncertainty that comes in the form of biases, lack of knowledge, and so on and so forth. But what a machine could never replace is the human touch. The ability to not just help with rationality and logic, but also with compassion, patience, and empathy. These are extremely important factors when it comes to public service. This is what makes a government connects with its people. This is what makes a nation.
References
Atmaka, P. (2021). Beda Golongan dan Eselon PNS, Mau Ikut CPNS 2021 Wajib Tahu. Pendidikan Info ASN. Retrieved from https://pendidikan.infoasn.id/beda-golongan-dan-eselon-pns-mau-ikut-cpns-2021-wajib-tahu/
Bestari, N. P. (2021). Heboh Robot AI Bakal Gantikan PNS, Ini Faktanya! CNBC Indonesia. Retrieved from https://www.cnbcindonesia.com/tech/20211125062103-37-294203/heboh-robot-ai-bakal-gantikan-pns-ini-faktanya
Bovens, M., & Zouridis, S. (2002). From street-level to system-level bureaucracies: How information and communication technology is transforming administrative discretion and con- stitutional control. Public Administration Review, 62, 174–184.
Bullock, J. B. (2019). Artificial Intelligence, Discretion, and Bureaucracy. American Review of Public Administration, 49(7), 751–761. https://doi.org/10.1177/0275074019856123
Febryan, A., Mursid, F. (2021). PNS Diganti Robot: Ambisi Jokowi demi Digitalisasi Birokrasi. Republika. Retrieved from https://www.republika.co.id/berita/r3f1f5409/pns-diganti-robot-ambisi-jokowi-demi-digitalisasi-birokrasi
Franedya, R. (2019). 2020 Jokowi Pangkas Eselon, Gantinya Artificial Intelligence. CNBC Indonesia. Retrieved from https://www.cnbcindonesia.com/tech/20191129063231-37-118900/2020-jokowi-pangkas-eselon-gantinya-artificial-intelligence
Lipsky, M. (1980). Street-level bureaucracy: Dilemmas of the individual in public service. New York, NY: Russell Sage Foundation.
Nilsson, N. J. (2010). The quest for artificial intelligence : a history of ideas and achievements. Cambridge ; New York :Cambridge University Press.
Organisation for Economic Co-operation and Development. (n.d.). PISA Participants. Retrieved from https://www.oecd.org/pisa/aboutpisa/pisa-participants.htm
Organisation for Economic Co-operation and Development. (2018). OECD Programme for International Student Assessment (PISA) Results from PISA 2018: Brunei Darussalam. https://www.oecd.org/pisa/publications/PISA2018_CN_BRN.pdf
Organisation for Economic Co-operation and Development. (2018). OECD Programme for International Student Assessment (PISA) Results from PISA 2018: Indonesia. https://www.oecd.org/pisa/publications/PISA2018_CN_IDN.pdf
Organisation for Economic Co-operation and Development. (2018). OECD Programme for International Student Assessment (PISA) Results from PISA 2018: Malaysia. https://www.oecd.org/pisa/publications/PISA2018_CN_MYS.pdf
Organisation for Economic Co-operation and Development. (2018). OECD Programme for International Student Assessment (PISA) Results from PISA 2018: Philippines.https://www.oecd.org/pisa/publications/PISA2018_CN_PHL.pdf
Organisation for Economic Co-operation and Development. (2018). OECD Programme for International Student Assessment (PISA) Results from PISA 2018: Singapore. https://www.oecd.org/pisa/publications/PISA2018_CN_SGP.pdf
Organisation for Economic Co-operation and Development. (2018). OECD Programme for International Student Assessment (PISA) Results from PISA 2018: Thailand. https://www.oecd.org/pisa/publications/PISA2018_CN_THA.pdf
Simon, H. A. (1947). Administrative behavior: A study of decision- making processes in administrative organizations. Oxford, UK: Palgrave Macmillan.
Tegmark, M. (2017). Life 3.0: Being human in the age of artificial intelligence. New York, NY: Alfred A. Knopf.
World Bank. (2018). Learning more, growing faster. In Indonesia Economic Quarterly (Issue June). https://openknowledge.worldbank.org/handle/10986/29921
