Worrying about the Future of Work
AI technology has changed the way humans work. The latest innovations, such as ChatGPT, has proven to be able to complete tasks that take days, weeks, or years for humans almost instantly.
The appearance of ChatGPT is impressive. The latest innovation from the merging of natural language understanding technology and its processing has proven to already help human tasks. However, its presence has also caused concern, will it take over our work?
Artificial intelligence (AI) technology has changed the way humans work. The latest innovations, such as ChatGPT, which combines natural language understanding (NLU) technology and natural language processing (NLP), has proven to be able to complete tasks that take days, weeks, or years for humans almost instantly.
ChatGPT is a large language model trained on a large text database to replicate human communication. Its abilities have attracted the attention of many people. According to a UBS report quoted by Reuters, ChatGPT has attracted 100 million users in just two months, much faster than Tiktok who took nine months to reach the same milestone.
Microsoft, as an OpenAI investor and the ChatGPT developer, has incorporated the technology in its Bing search engine. This inspired other giant technology companies, such as Google, to immediately develop similar innovations. Google developed Bard and Meta. Mark Zuckerberg, CEO of Meta, announced that Meta would release a new large language model called LaMDa that would help scientists and engineers explore applications for artificial intelligence, such as answering questions and summarizing documents.
Also read:
> AI, A Double-Edged Sword We Must Face
Meta also said that it would make the model available to the public research and take applications from researchers. The model that underlies Google's LaMDA and OpenAI ChatGPT is not public.
Even though it has triggered almost mania to forge new artificial intelligence innovations, the popularity of ChatGPT has also triggered concerns that AI could interfere or even take over human work. One of the founders of Apple, Steve Wozniak, claimed that he was skeptical of ChatGPT. A potential problem that could arise is that the technology could make a terrible mistake as it does not know what humanity is.
However, there are several aspects of human work that have already been replaced by AI. Researcher of intelligence artificial Wahyu Adi Setyanto gave an example, graphic designers can provide text input to AI machines to create a logo. Then, the machine will generate and produce a pixel circuit to form a logo.
In the last six years, the development of artificial intelligence has been extraordinarily fast.
In other aspects, humans tend to consider artificial intelligence as a "collaborative friend" to produce work. This example has been carried out by, among others, programmers who used to write down programming language codes.
"In the last six years, the development of artificial intelligence has been extraordinarily fast, trying to follow the cognitive processes of the human brain. The latest development of artificial intelligence even seeks to combine computer vision innovation that imitates the way the iris in the human eye works. The goal is to produce a better language translation," Wahyu said in Jakarta on Friday (24/2/2023).
According to Wahyu, the concern now is the side effects of artificial intelligence products. In terms of data collection, for example, one issue can be the use of invalid data sources. Then, the process of encoding data when it is carried out by developers who do not have ethical awareness. As a result, the products produced may further strengthen negative sentiment toward minority groups, carry a gender bias and other forms of discrimination.
"To match humans, it seems [the development of] artificial intelligence has not yet arrived at this point. The thing that is a current worry is what if the research teams come from homogeneous groups," he said.
The solution to overcome this potential problem is to create a module card that allows each data source to be described when and what is intended for what.
Mapping
A lecturer at the Electrical Engineering and Information Technology Department in the Engineering School of Gadjah Mada University, Sunu Wibirama, said that the phenomenon of the rapid development of AI technology could not be resisted. Among academics there has been a feeling of apprehension about the extent that artificial intelligence, including large language models, can develop.
"Wanting to forbid students from using it will be difficult. ChatGPT and the like have been implanted in search engines and perhaps derivative products owned by giant technology companies. Their artificial intelligence research funds are also very large," he said, in Jakarta on Thursday (2/3/2023).
Also read:
> Artificial Intelligence Generative in the World of Education
The thing that can be done now is to anticipate. Sunu argued, it is time for Indonesia to map the advantages and disadvantages of AI technology. From there, the types of work or human professions that can be replaced can be documented.
In 2020, the World Economy Forum (WEF) released a list of jobs projected to be replaced by digital technology. However, Indonesia's conditions are different. There are still many workers working in the informal sector.
"If there is no regulation, the government at least should take steps to encourage awareness that the phenomenon of artificial intelligence already exists. The government should then also clarify the steps taken to map the strengths and weaknesses of this technology. It must be now so that there is still time to take anticipatory steps, such as reskilling," said Sunu who also serves as chairman of the IEEE Systems, Man, and Cybernetics Indonesia Chapter.
Regulations
Legal consultant in the field of AI technology Matheace Ramaputra explained that Indonesia did not yet have regulations for artificial intelligence. Existing legal provisions only dealt with the transformation of electronic-based services. There is also the mention of electronic agents in the amendments to Law No. 19/2016 on information and electronic transactions (ITE). However, the mention of electronic agents cannot be said to be AI.
Matheaca believed it was time for AI to be regulated as there were still opportunities for technology that is encouraged to match the cognitive abilities of humans.
Matheace suggested that the Indonesian government also made regulations with two approaches. First, the developer approach. Regulations for developers are needed so that AI innovations take into consideration the significance of the risk that the AI could interfere with or replace the human professions. This approach also aims to protect professions that do not have associations.
Second, professional approach. The most important instrument is a provision that strengthens professions through improvement of the code of ethics. Thus, professions produced increase in qualifications.
"I cannot deny that among the advocate profession, for example, there are unfair advocates. As a result, there emerges initiatives to use artificial intelligence technology to give birth to more equal legal decisions," he said.
Matheace added that in many countries, the government's attention was moving toward regulating AI technology. At the same time, that movement must face pressure so that regulations are not tight and limit innovation. Eventually some countries have made legal guides for the use of AI. The British government, for example, has made a legal guide for the application of driverless vehicles.
Not replacing
Cin Cin Go, IBM Indonesia's Technology and Country leader, emphasized that for IBM, it was clear that the purpose of the AI system was developed and had now been applied is to solve business and community problems by increasing human intelligence. So, AI technology was not replacing human intelligence, but helping humans make better decisions and empowering people to do their best work.
"Work that previously took several hours can now be completed in a few minutes. Automation and technology based on artificial intelligence will allow machines and humans to collaborate smarter," Cin said.
With the faster growth of AI technology innovation, workers' skills must be improved. He added that IBM was helping to support the technological skills training of professionals in Southeast Asia.
Rapid AI growth is also a necessity for us to improve labor skills, especially in fields that require the skills to do new jobs that will be created by AI in the future.
Also read: Artificial Intelligence
IBM recently announced the opening of the IBM Academy for Hybrid Cloud and AI in Batam to develop "ready" labor skills and fill in the existing skills gap. Chairperson of the Indonesian Customer Experience Professional (ICXP) Sri Safitri said the implementation of a customer service chat had been running for the last few years in Indonesia. In reality, the implementation of this technology reduced the number of human workers. The affected workers, by several companies, were diverted to work in other fields. For example, consumer service administration on social media.
"A consumer service chatbot usually only plays a role at the beginning, namely answering general questions. That means, the role of human workers is still needed to serve complex questions, such as complaints. There are companies that eventually want to provide training so that affected employees have the relevant skills," said Sri.
CEO of Kata.AI Irzan Raditya is of the view that the technology behind ChatGPT could actually be used for chatbots so that services to consumers are more accurate. However, this did not necessarily make all forces or the prospective workforce have to master digital technology skills, such as programming code, although he believed it would be useful for the workforce to have those skills.
"Adding new skills is always good. Only I think things that need to be improved by humans now are insight, critical thinking skills and compassion," said Irzan.
This article was translated by Hyginus Hardoyo