After several groggy decades, artificial intelligence (AI) has finally woken up. Recent advances are bringing us ever closer to an exciting technological frontier, where AI’s far-reaching potential will impact every area of our daily lives, from business productivity to interpersonal relationships, to simply helping us live longer.
Naysayers will always play down its transformative effect, emphasising the dangers of an AI-driven society. But while caution is valid, such scaremongering is unfounded. This article breaks down five of the main machine intelligence myths and explains why we have no reason to fear artificial intelligence.
Robots taking over jobs will lead to mass unemployment
You’ve probably heard about robots taking over jobs. But ask yourself: would making certain jobs redundant really be such a bad thing, especially when it frees up our time for more valuable things?
Amazon’s checkout-free grocery store has become the symbol of the automation economy, where repetitive tasks are entrusted to efficient machines. Because, unlike us, machines can’t get bored and drop off in performance, this will see a rise in productivity. And, as Bill Gates recently commented, with greater collective productivity comes more individual liberties.
In many areas, AI is supplementing rather than replacing human jobs. In finance, for example, there’s evidence that AI—fintech—won’t replace traders, but assist them, eliminating behavioural bias and reigning in another human fallibility: overconfidence.
AI is far more likely to lead to the creation of professions. These will be professions that value human ingenuity––ingenuity that we’ll have more time for while the machines, with their artificial intelligence jobs, do all the heavy lifting.
The rise of the machines will trigger a war we cannot win
Everyone remembers that scene from Terminator 2; Sarah Connor watches on hopelessly as kids playing peacefully in a playground are blown to bits when Skynet becomes self-aware and triggers the nuclear apocalypse known as Judgment Day.
Humanity’s judgment day at the hands of the machines is nothing more than an existential fantasy. But it’s a fantasy that authors and Hollywood screenwriters peddle because we buy into it. Because as the dominant species on earth we enjoy reflecting on our own vulnerability, and imagining what would happen if, by chance, the tables were turned.
Some subscribe to it all too seriously. Tesla’s Elon Musk has warned of the “existential threat” machine intelligence could pose to humanity. Short of his own self-driving cars running amok en masse, however, we should all still sleep soundly.
As a leading computer scientist at the University of Oxford has recently pointed out, 95% of AI research is directed towards human life extension, not obliteration. Machines aren’t going to turn against us unless we specifically program them to.
Doctors could lose out to AI in healthcare
You can understand why people might be cautious about the role of artificial intelligence in medicine, especially when it’s a matter of life and death. But machines aren’t going to substitute trained professionals, they’re going to support them, wading through the bureaucracy and deluge of data so that doctors can focus on what matters most: our longevity.
According to Robert Darnell, director of the New York Genome Center, to avoid going under healthcare professionals need to start harnessing the processing power of AI. “Having doctors cope with the avalanche of data that is here today, and will get bigger tomorrow, is not a viable option.”
IBM Watson has already proven the awesome potential of AI in medicine. By scanning millions of journal articles, it managed to diagnose a brain tumour and recommended a treatment plan in just 10 minutes. To give some context, the same diagnosis would have taken experienced doctors 160 hours.
IBM Watson and Babylon Health are currently the industry’s main heavyweights, but there are many new AI startups looking to challenge them. Such has been the recent surge in interest that we predict an estimated 90% of hospitals in the U.S. will use AI in healthcare by 2025.
Advances in AI will usher in an age of mass surveillance
Recent big-scale data breaches make AI’s role in mass surveillance a touchy subject. It’s even led some to envisage a Big Brother society, where malicious AI permeates every aspect of our daily lives, compromising personal privacy and freedom of thought.
But such a scenario is unlikely. AI is value neutral, incapable of acting maliciously unless explicitly told to. As with any revolutionary new technology, AI is a magnet for scepticism. But as we’re the ones who set the goals and the values, it’s not AI we should be way of but whose hands the AI is in.
The reality is that what AI can do––and in some cases already does––is often very helpful. We’re grateful, for example, for the number-plate reading AI systems the police already use to make drivers accountable. We should also be thankful for a company like Droneforce, which uses AI to hunt down the modern menace of unauthorized drones.
We should embrace surveillance capabilities for its protective rather than persecutive potential. Teachers, for example, have expressed interest in the preventative capabilities of IC Realtime. If it sees a group of students suddenly gathering, it can send staff a notification out that a fight may be about to break out, allowing them to take the necessary preemptive action.
AI will lead to a rise in discrimination
Let’s admit it, AI doesn’t have the best reputation for anti-discrimination. In 2015, Google Photos ran into trouble when its facial recognition software algorithms labeled a black user and his friends as gorillas.
Similarly, in 2016, Richard Lee, a New Zealand citizen of Asian descent, had his passport photo rejected by AI facial recognition software, which tactlessly cited that “the subject’s eyes are closed.”
The problem in both cases came from bias in the AI’s training sets. AI learns from the data we feed it. So if some image recognition software continually places women in kitchens, for example, it does so as a reflection of our own social bias which visually stereotypes women in this situation, not as a projection of its own.
These are just growing pains, not underlying prejudice. The fact of the matter is that AI is poised to hamper discrimination, not enhance it. Want proof? Just look at Spot, the judgment-neutral AI chatbot, which fights workplace harassment by asking victims exactly the questions and storing their responses in anonymously filed case reports.
The future of AI is in our hands
Like any revolutionizing technology (the Internet being another example), AI has become a magnet for our fears, be they about our wealth, our health, our employment or even our very existence.
But as much as we wax lyrical about its power, strip it all down and AI is just a set of mathematical algorithms searching for patterns to solve complex problems. Depending which problems we decide to prioritize, its future is as bright as we want it to be.
