Will AI make your job obsolete? | Sunday Observer
‘Rise of the machines’

Will AI make your job obsolete?

26 March, 2023
The Matrix films explored a world where humans are “farmed” by machines and kept in a dream-like state via a virtual reality program
The Matrix films explored a world where humans are “farmed” by machines and kept in a dream-like state via a virtual reality program

Artificial Intelligence (AI) has hit the mainstream, but there is an enormous amount of scepticism around this new technology. Oftentimes AI conjures up images of dystopian futures where killer robots stalk the streets of abandoned cities, hunting the last of humanity. But that is just Hollywood. Encounters with AI are benign as your Spotify and Netflix recommendations. In fact, people are widely working with and experimenting with AI applications to do research and make art.

With AI, we will see a dynamic shift in our workplaces not seen since the advent of computers and the internet. Machine Learning, the technology behind self-driving cars and ChatGPT, could potentially make millions of jobs obsolete. However, we hear graduates using ChatGPT to generate entire research papers, leaving academics concerned about widespread plagiarism. Traditional artists claim that AI rendered images could signal the end of creativity.

In a recent video, British journalist and author of ‘Fully Automated Luxury Communism’, Aaron Bastani said that AI will not put humans entirely out of work. “This doesn’t mean (AI) will take every job in professions like coding, journalism and accountancy.

Rather a single person equipped with powerful new tools might replace a whole team. An advanced version of ChatGPT could automate tasks like debugging and translating code; narrowing entry level coding positions. There would still be coders, just not too many”.

To understand more about the AI phenomenon we spoke to Thamalu Piyadigama. He is a senior data engineer at hSenid Mobile Solutions and researcher in physics, linguistics, philosophy and economics.

Simple steps

He says conventional technology only addresses tasks that we can break into definite simple steps, but AI can automate complex tasks we don’t know how to fragment.

“For example, we don’t know how to prepare a complete instruction set to identify images of cats or to translate a document. AI can solve those problems. We don’t program AI models. AI models learn from data. Although this works amazingly well, we have an interpretability crisis. Modern AI models are black boxes. We can’t explain how they give answers”.

In the case of occupations, Piyadigama said that new technologies replace some jobs and create others.

“As with modern industry, we can replace many jobs with mundane and time-consuming tasks such as processing and restructuring data, labelling, classification and documentation.

It can also replace mediocre designers, content creators, music directors, and translators. But AI can’t replace the entire creative domain. For example, prominent artists can guide AIs and build on top of the output for even better results. AI will be another tool for them.”

In the meantime he says that teachers, sales persons and similar jobs will not disappear as human interaction is central to those jobs.

“Driving is another skill that will become redundant shortly with the rapid development of self-driving vehicles. It will take time to replace doctors and financial workers. Nobody likes to accept medical or financial guidance without a rationale.

Our tendency to seek explanations prevents these jobs from being automated sooner. Anyway, that will change when the accuracy becomes attractively higher. If the success rate of a doctor for some surgery is 75 percent while the success rate of a machine is 99 percent despite nobody in the world knowing how exactly the machine works, what will you choose?”

However, decades ago futurists’ like Arthur C. Clarke predicted that machines would do all the hard work, but critics of AI says that it is more poised to replace white colour jobs.

Piyadigama agreed that people expected quite the opposite before this new wave of AI, where machines would do all the hard labour, and humans will be limited to creative mental work.

“Machines would take over both. It has already started. Robots have been running entire lights-out factories like FANUC in Japan since 2001. As Mara Pometty says in her recent book, ‘Umani Artificiali’, the real problem is not machines becoming more human but humans becoming more like machines.

Thamalu Piyadigama

We rely so much on the targeted campaigns and data selected for us by AI-backed search engines. We change our behaviour to align with these machines and who owns them. Conformal bias, indirect propaganda and less human lifestyles would be the real problem”.

But, he says AI is still not on par with human perception. “AI can work with collected data, but building AIs that can work and learn in real-time, and perceive the environment is a big challenge.

Thus far, there has been minimal progress in that regard. Our biological perception is really faster than any AI that has ever developed”.

Potential dangers

Regarding potential dangers, he said AI should be controlled to stop it from generating fake news, forgery, hate speech, conspiracies, pornographic content and harmful content. “Reinforcement Learning from Human Feedback (RLHF) and content moderation were somewhat successful. However, the subfield, responsible AI, has a long way to go”.

He also spoke about quantum computing technology. “The issue is nobody has solved the problem of scaling quantum computing up effectively. We know a lot of things to do with a quantum computer. There is plenty of research. But we don’t know how to build a large quantum computer successfully.

Tip of the iceberg

The general public will see a sudden advancement of quantum technologies if we’ve ever managed to create one. The truth is a lot of background research happened decades before those viral moments or the tip of the iceberg”.

Nuwan Shilpa is an AI artist who started out using Google deep dream a few years ago. “I’m quite an ardent follower of new technologies, the moment I saw that AI art generation was available at a ready to use consumer level, it was a no-brainer for me to start using it,” he said.

Shilpa believes AI will have a massive impact on society and play an integral part in all aspects of life. “It could be a double edged sword of sorts, perhaps only time will tell of its impact, but I envision that AI could be an integral part of the evolution of the human condition”.

We asked him how AI could be used as a tool. As an artist, I feel that it is one of the most versatile of tools, out there unrivalled by any others, for the exploration of the imagination”.

Shilpa says that AI can be understood as an evolving network of machine intelligence, based on human activity built upon electronics that interfaces back to humans. “After all its created in our image.

It’s almost surreal. I think AI will be one of the greatest tools of introspection that will be laid before civilisation”.

We asked him what humans would use their free time in a future where AI replaces most jobs. “With AI my mind is cast into exploring a more idealistic scenario, perhaps it could be the very push that is needed for a collective transcendental experience. A quite literal freedom from time itself”.

Meanwhile, Piyadigama also allayed the fears concerning the economy in terms of automation lessening the value of commodities.

“We don’t have to worry about the so-called tendency of the rate of profit to fall (or the famous TRPF crisis theory). Marx, Ricardo and Smith believed in the labour theory of value, which is now discredited. When we ignore substantial value theories like LTV, the crisis doesn’t exist.

Japanese Marxian Economist Nobuo Okishio and John E. Roemer corrected equations of Marx to prove there is no such theoretical necessity even in the framework of LTV.

Two chapters

A lot of evidence suggests Marx found himself that TRPF is wrong later in his calculations. LTV fails to explain the profitability of capital-intensive industries, and the proposed solution by Marx also fails. I have included two chapters on those issues in the book I am preparing on the failure of Marxian economics. The collapse of the capitalist system is contingent. Technological advancement and excessive automation do not necessitate it”.

Although AI technology has grown in leaps and bounds, it is still in its infancy. And as Piyadigama and Shilpa aptly put, the potential here is enormous.

The work hours are already being reduced in developed economies like how British firms cut to a four-day work week. We asked ChatGPT for its opinion on how the nature of work can change with AI.

“Here are a few ways AI technology is likely to change the nature of work in the coming years.

AI can automate many routine and repetitive tasks. This will free up time for humans to focus on higher-level tasks that require creativity, critical thinking, and problem-solving.

AI can augment human capabilities by providing insights and recommendations that humans may not have considered. For example, AI-powered analytics tools can help businesses identify patterns in large amounts of data that humans may not be able to see on their own. This can help businesses make more informed decisions and improve their performance.

AI will require a shift in the skills and capabilities of the workforce. As AI becomes more prevalent, workers will need to adapt and learn new skills to remain competitive in the job market. This will require investment in training and education programs to prepare workers for the new world of work”.

[email protected]

AI artworks by Nuwan Shilpa