Home ›› Tech

A lukewarm defence of machine takeover

Sairas Rahman
28 Apr 2023 17:06:57 | Update: 29 Apr 2023 01:00:59
A lukewarm defence of machine takeover
A Cyberdyne System Model 101 Terminator, the iconic, fully automated killing machine featured in The Terminator series — Courtesy/StudioCanal

Those who grew up in the 90s, do not need any refresher on what a machine takeover is. Plenty of movies explained the phenomenon quite well, with The Terminator and The Matrix doing a stellar job teaching average Joes the dangers of unchecked artificial intelligence (AI).

Humans, flawed and prejudiced to the core, project their fears and paranoia on the world around them. When the likes of Bill Gates and Elon Musk warn that AI could very well end humanity in the years ahead, even the skeptics cannot help but feel a tinge of horror.

Existential crisis aside, is AI all that bad, and can they become any worse than humans? Many who suffer from a delusion that the past was a golden era, should study history a bit to learn how exactly humans have been treating each other in the last 10,000 years.

Even before computers got inside everything, humans were spending the lion’s share of global GDP on devising clever ways to kill or enslave each other more efficiently, with bullets, bombs and engines of war.

If machine overlords do take over the world in a few generations, would it be too outrageous to hope things might actually get better for humanity? Nothing on this earth has ever been able to surpass human brutality, why should AI be any different?

AI tools such as the Dall-E 2, and ChatGPT can not only understand human communication, but can already write engaging news articles, essays, transcriptions, and marketing pitches that outperforms anything humans can write within the same timeframe.

Machines surpassed us decades ago

The most famous battle between Deep Blue and the world chess champion Garry Kasparov occurred in 1997, although some experts believe the computer had an unfair advantage.

And while the computer system used to defeat Kasparov is not technically "smart," we cannot deny that it has surpassed the human ability to analyse chess moves and choose the right one.

Fast forward to the present times, and data now drives much of what we do, from analysing election results to developing healthcare solutions. People who build models that explain and predict patterns in the ocean of "big data" are in short supply.

But that may not matter due to the development of computer software that proves to be better than humans. MIT's data science machine can automatically generate predictive data models from raw datasets in as little as 2-12 hours.

It can take several months for a team of data scientists to complete the same task.

AI is smarter and faster in visual recognition, which is used in applications as diverse as facial recognition in photo storage applications and the analysis of shared social media images to improve marketing.

Any object recognition task is likely to be completed faster and more efficiently by a robot than by a human.

AI workers never need respite

Robots are undeniably better at repetitive tasks than humans. Amazon was the first major logistics retailer to adopt robotics and now owns 30,000 so-called Amazon robots, which it uses in its US warehouses.

Humans make mistakes, but not computers. They receive instructions and execute them exactly as written in the code. It is extremely important for such tasks as data entry, where a typo can lead to chaos.

AI can perfectly perform work that requires copying, pasting, transcribing, and typing. Lack of rest, boredom from repetitive tasks, even "hanging out" are purely human problems. For example, staying up all night or being stressed affects a worker’s performance the next day.

On the contrary, computers never need to sleep, rest, or have fun. No matter what, their operational capabilities are the same. This is what the machines were made for.

Jobs such as mining, factory work, and machine assembly expose workers to danger. Whether it is extreme temperatures at work, hazardous fumes, or falling objects, there will always be circumstances and situations in which people can be seriously injured or killed.

People can use AI in manufacturing to improve the efficiency of processes and protect people from industrial harm. Opportunities for using AI and machine learning in manufacturing include logistics optimization, product development, predictive maintenance, and, of course, robotics.

While machines can also be injured or crushed while performing hazardous work, they are not as fragile and are built to withstand enormous pressure, heat, airborne toxins, and other threats.

While the initial costs of building and training an AI machine are high, the total operating price is much lower than paying a human for the same job. Only electricity and periodic maintenance are required to use the machine.

Will AI take away jobs?

Introducing AI and advanced technologies suggest that, in the long term, machines will become even more refined and processes even smarter. Therefore, experts agree that AI will fully automate many professions in 5-10 years.

However, the problem is that this can lead to a reduction in the number of jobs.

This is already happening as technology continues to rapidly replace humans in the workplace in manufacturing, service delivery, recruitment, and the financial industry, causing human workers to move into low-wage jobs or become unemployed.

According to the World Economic Forum report, AI will replace 85 million jobs globally by 2025. Large-scale and inevitable changes are coming. But many of them are not catastrophic.

The same report also states that it will create 97 million new jobs in the same field that will require more complex and less routine skills.

Billion-dollar prospects

Former Google CEO Eric Schmidt believes we’re in “the golden age of investing in AI.” In the third quarter of 2021 alone, $18 billion was invested in artificial intelligence companies, an absolute record.

The global artificial intelligence (AI) software market will reach approximately $126 billion by 2025.

Take a look at the statistics based on forecasts for the period from 2018 to 2030, by segment – Logistics, packaging, and materials segment will hit $31 billion, AI analytics $70 billion, autonomous cars $87 billion, and robo advisors $255 billion.

This capital influx is a sign that investors believe in the potential of AI, and they bet that it will eventually automate many jobs instead of creating value with machines. If someone doesn't use AI, they are at a disadvantage.