AI isn’t evil. It’s just a tool.
Search for ‘artificial intelligence’ and ‘automation’ and it seems like many see them as harbingers of the next apocalypse.
That’s plain foolish.
The truth is that artificial intelligence and humankind are good bedfellows. Those who say otherwise are just peddling fear.
Let’s bust some myths with facts.
Fact 1: It’s not AI, it’s us
As a species, we are obsessed about the end of days. This preoccupation with apocalyptic scenarios is deep-rooted in our psyche. After all, fear is natural and from an evolutionary psychology perspective, fear of potential threats is precisely what has protected mankind.
It’s little wonder that apocalyptic literature can be found across various cultures and religions, from the Bible, to Nostradamus’s predictions, right down to modern pop culture. ‘The Walking Dead’ is an archetype of zombie apocalypse, movies from ‘Independence Day’ to ‘A Quiet Place’ deal with alien invasion, and Doomsday Preppers isn’t just a reality TV show, they’re an actual movement.
Also Read: Forget founders, here are 12 tips to being a successful startup employee
Naturally, artificial intelligence (AI) fits into the stereotype of an apocalyptic villain – it is new territory, seemingly unfamiliar, and potentially limitless.
That’s probably why even the technology community and its leaders can’t come to an agreement on what AI means for humanity. The late Stephen Hawking said that “The development of full artificial intelligence could spell the end of the human race.” Elon Musk is a prominent doomsayer. On the flip side, Mark Zuckerberg called Musk’s apocalyptic warnings “hysterical”.
As someone who works in the field, here’s the truth. AI isn’t the villain, in fact, AI and people are good bedfellows. Ultimately, what will doom humanity is not the machines, but humans ourselves.
Fact 2: We’ve made progress, but not as much as you think
When Mark Zuckerberg built Jarvis (an AI butler that helps with chores like making toast and turning off lights), Musk said: “It’s really not AI to turn the lights on, set the temperature.”
While this does sound like an argument best left to tech titans, it’s not, because to understand what AI is capable of doing, we need to understand what exactly AI is.
Just like how there are different types of intelligence – IQ, EQ – for human beings, there are different categories of AI.
The first category of AI is Narrow AI when machines or bots can complete narrowly or well-defined tasks. Things like turning lights on and off require following a prescribed set of steps. The machine still needs human beings to ‘feed’ possible scenarios and “what-ifs” to them, in order to function.
The next category is Strong AI, the one that our apocalyptic nightmares are about; when machines and bots become self-aware and self-directing (like human beings but faster).
“Human decisions are removed from strategic defense. Skynet begins to learn at a geometric rate. It becomes self-aware at 2:14 AM, Eastern time, August 29th (1997)…”
That’s how the story goes in Terminator, but in reality, it’s 2019 and Skynet is not here yet nor is it coming any time soon.
Most of the AI progress we’ve made strides in is in the first category – using machines or bots to complete tasks.
That’s what Osome’s bot does. Rather than having highly experienced human specialists endure mundane tasks like checking whether all required documents for registering a company has been submitted and is all filled up correctly, the Osome bot does it. Because it’s a bot, it does it faster, more consistently, and more accurately than a human specialist can (because of human error).
It’ll do it any time – weekends, public holidays, even at three am in the morning. It can also get better at doing it because it can use past cases and experiences to influence its future decision-making. Thanks to progress in natural language processing (NLP), bots today can also better understand and interact with the humans it serves.
But as anyone who has used Siri, Alexa or Google Home can attest to, we’re still not quite there yet. Especially if you have an accent.
Needless to say, when it comes to the kind of self-aware, self-directing, AI that our nightmares are made of, we’re nowhere near that.
Fact 3: AI is just a tool and can be used for good
Like fire, knives, and guns, AI is a tool and it can be used for good, or for evil.
Osome use AI in professional services to free employees from mundane, routine tasks that are not just bad for productivity but also their own wellbeing. The results are happier and more engaged employees and a more consistent, convenient and affordable service for our customers.
In one year, we’ve saved our employees and over 1,000 clients from tens of thousands of hours of dull paperwork.
Healthcare is also ripe for improvement through AI. A study by Accenture estimated that AI can help the U.S. healthcare industry alone save up to $105 billion in the near-term.
The study estimates that US$40 billion worth of value and savings can come from the use of AI in robot-assisted surgery. A study of 379 orthopaedic patients has already found that robot-assisted surgery resulted in five times fewer complications than surgery helmed solely by human surgeons.
The study also estimates another US$20 billion of savings can be made through the use of virtual nursing assistants. The nursing profession is famously short-staffed and overburdened by administrative paperwork that leaves nurses distracted from their core care duties. AI nursing assistants can help cut time spent on paperwork, and help nurses monitor patients, at all hours and even remotely.
Woebot and Replika are both AI-powered chatbots developed to help people with their mental wellness. Woebot is built by clinical psychologists from Stanford and delivers bite-sized, on-demand cognitive behavioral therapy to anyone with a smartphone. It’s affordable and an initial trial has shown encouraging results; a study showed that Woebot users reported a reduction in depression symptoms over a two week period, compared to a control group.
Both Woebot and Replika work precisely because they’re not human; with no fear of judgment from another human being, users can be completely honest, and the bots can gather numerous cues from text and emojis at super speed to figure out appropriate responses and follow-up questions.
The University of Southern California has also launched the Center for Artificial Intelligence in Society, which is dedicated to using AI for social good, solving problems like climate change, homelessness, and more.
Fact 4: We already know how the story goes
Part of why people are so terrified of AI is because they think it’s an unprecedented event.
That’s not true, AI is the fourth industrial revolution. The first industrial revolution was powered by steam, then science made mass production possible through engineering and machinery, and next came a digital revolution through the creation of semiconductors, computing, and the Internet.
All three revolutions triggered a similar sequence of events – a shift in the industry driving the economy, job creation in some sectors, job loss in others, and a change in living and working conditions for people.
In earlier industrial revolutions, our economy shifted from a largely agricultural one to a capitalist one.
Now, mass production has long been possible, powered by machinery and factories. Transportation has improved. New jobs were created in engineering, factories, and urban areas. The standard of living has improved for some. But, poor policies and a lack of protections for the poor and working class caused many to suffer from awful work conditions, low pay, and child labour.
In the digital revolution, the economy shifted again. Manufacturing was no longer the star. New jobs were created in computing, programming, and more. Job loss in manufacturing occurred, going to automated machinery or less developed economies. Standards of living improved for some. Poor policies and protection for the poor and working class caused many to suffer from unemployment, poor working and living conditions, and low pay.
In the AI revolution, the economy will shift again. A new industry or segment will be the star. Job loss in manufacturing and services will occur, going to AI bots, who will assist human specialists in their work. Product and service standards will improve as a result. Cost of products and services come down and become more affordable. Standards of living rise, and work-life quality will improve for human specialists.
But what about those who lose their jobs? If we go by history, society fails to protect them and they will likely fall through the cracks. But the story for that can still be re-written if the government and individuals take critical steps in the next 2-3 years to make the distribution of benefits more equitable.
We need better policies to address income inequality, distribution
Global income inequality is staggering — the richest one per cent own 45 per cent of the world’s wealth.
It is clear that we need better policies, social protection systems, education systems to address this.
Also Read: Why remote working is the future for startups
Failing to do so means that the AI revolution will only add on to the injustices of previous industrial revolutions; benefits will go to the rich and educated, the poor and working class will continue to reap little.
Education systems need to be improved, not just for the young, but for those whose jobs are most susceptible to automation. Retraining programmes are not enough, since jobs may be displaced before human beings have the time to complete retraining and find new employment. Safety nets for workers need to be put in place to support them through the whole process.
AI isn’t as advanced as you think it is. AI is not evil. We know what’s going to happen. Yet the true horror comes from our inaction — only 25 out of 100 countries assessed by the World Economic Forum are well-positioned to benefit from the AI revolution.
—
Victor most recently founded Osome in Singapore in May 2017. Osome has helped over 800 clients in Singapore, Malaysia, and India save over 3,000 hours on paperwork, and they have recently closed a US$2M seed round led by Berlin-based VC firm Target Global. Target Global has also invested in food delivery unicorn Delivery Hero (they had acquired foodpanda in 2016).
Victor has a background in physics and technology – he has taught at his alma mater Moscow Institute of Physics and Technology as an assistant professor, where he had also earned a Master of Science in Applied Mathematics and Physics.
e27 publishes relevant guest contributions from the community. Share your honest opinions and expert knowledge by submitting your content here.
The post AI and people are good bedfellows appeared first on e27.