We’re living in an exciting time for technology: new developments are changing the way we live, work and play. But when innovative technology is applied to concepts such as war, the conversation must turn away from excitement and towards an altogether more difficult concept: ethics.
In recent years, we’ve been lucky enough to see the transition of AI from its science-fiction origins to working technology. Now, artificial intelligence can help businesses to do more for their customers, help us to manage our lives, and has the potential to expand even further beyond these applications.
Unfortunately, there’s also a downside. As with all technology, AI has become more accessible and advanced – inevitably leading us to find ways to apply it to areas such as war and other unscrupulous activities.
With problems arising around how AI could make us vulnerable and move further away from life-enhancing possibilities and closer to danger, is it finally time for innovators in Silicon Valley and their imitators to face up to their ethical duty?
Google’s AI Army
“AI and cyber and some of these other newer realms offer possibilities to our adversaries to [target others successfully], and we must see to it that we cannot be surprised.”
Mike Griffin, Undersecretary of Defence for Research and Engineering
Via The Verge
Although there have been instances of Silicon Valley’s residents acting unethically for years now – such as Facebook’s Cambridge Analytica scandal – the conversation around ethics has mainly been sparked by a few key events in recent months. Chief among them is the nightmarish scenario that you’d usually find in a Hollywood blockbuster: AI being used to create weapons.
The topic of using AI drones to take out enemy targets is one the Pentagon is all too happy to have. Mike Griffin – former space programme director and now Undersecretary of Defence for Research and Engineering – has been a vocal supporter of joining the so-called “AI arms race”. According to Griffin, such technology offers America’s enemies an opportunity to hit them in surprising new ways, and they need to keep up as a result.
From this desire to compete, Project Maven was born – a venture which saw the Pentagon turn to Silicon Valley’s AI experts at Google. Employees under Project Maven have been directed to train AI to better distinguish between individuals and objects in order to make them more effective.
If this sounds like a terrifying prospect, it is.
Thankfully, there is some hope as many Google employees have quit the company in protest of its work on Project Maven, as well as a growing lack of transparency from executives who have opted to take on other controversial projects.
This is indicative of a conscience in Silicon Valley – to the relief of many – and demonstrates that not even employees wooed by Google’s care packages and headquarters can be swayed into unethical practices when it comes to war.
Failing Public Support
“One danger is that the image of artificial intelligence in the public will degrade because of bad uses of AI.”
Yann LeCun, Facebook Director of AI Research
Although anyone who has read an account of Google’s goodwill towards its employees will know that they like to look after their people, the Silicon Valley giant is surely confident that staff unwilling to work on projects such as Maven can still be replaced. But what happens when unethical activity affects public support?
At the time of writing, there hasn’t yet been a major outcry from the public around Project Maven – perhaps in part because it seems a million miles away from everyday lives and the excitement of AI. That doesn’t mean, however, that public support isn’t important. In October 2017, two of the modern day pioneers of artificial intelligence issued a warning: governments must address fears around unethical applications of AI or risk losing public support.
Said pioneers – Yoshua Bengio and Facebook’s Yann LeCun – have both made compelling cases for a greater partnership between the government and tech firms for AI’s sake. Bengio, hailed as the ‘godfather’ of deep learning, told the audience at a deep learning summit in Montreal that his greatest concern was the “misuse of autonomous weapons”, whilst LeCun has urged governments to address threats such as the manipulation of democracy using AI.
Without these recommended steps being taken, the image of AI as an innovative and useful technology could be undermined, leading the public to turn their backs on artificial intelligence and its applications. Of course, there’s also the chance that it will lead to widespread protests and abandonment, sending the development of AI back by several years.
The Difficult Conversations
Unsurprisingly, Project Maven and other AI-centric developments are not the only Silicon Valley export worthy of being labelled unethical. The Valley’s residents have been hit with lawsuits over personal data, sexual harassment cases, accusations of unwittingly supporting terrorist activity, and even startup fraud. All of these storms have, however, passed.
Could the splashing of Project Maven across headlines fare any differently and shine a spotlight on Silicon Valley’s less-than-clean practices? Possibly. Though in reality, the effect would be minimal: if the US’ defence department wants to arm themselves with AI, it’s more a matter of “when”, rather than “if”. After all, they can always find engineers willing to work on the project, and even if the public disagree, the battlefield is far away from home, and the effects can’t be felt.
In that case, we could be waiting a long time for Silicon Valley to collectively answer to regulation and ethics committees. Until then, it’s important that people hold them to task over the controversial applications of technology being suggested – and Google’s recently-departed employee protesters have already started paving the way for us to follow.
Kaleida is a bespoke development house in Manchester, made up of a diverse and experienced team. We thrive on providing bespoke software solutions for clients in a wide range of industries. To find out more about the services we offer, feel free to explore our website, or get in touch directly.
