Nowadays, AI takes a lot of centre stage when we talked about what our future holds. Elon Musk & Stephen Hawking (to name a few) fears AI might become our master & overlords. Vladimir Putin believes any nation that leads AI will rule world. China puts a lot of effort to be a dominant future in AI and technology in general, by having “Made In China 2025” strategic plan. Few Googlers revolted towards their management over Project Maven with DoD. Amazon took a backlash over its Rekognition. IBM’s Ginni Rometty believes AI will actually enhance or augment or intelligence.
Lot of hope. Plenty of fears. And several suggestions as well, like AI regulatory oversight, proposed by none other than Elon Musk.
Also, an oath. An AI oath.
An Oath For All Type of Professions?
The atomic bombings of Hiroshima & Nagasaki in August 1945 has killed 100,000 to 250,000 people, directly or indirectly, either through the blast itself or its complications afterward. No need to go into its gory details, as I think it has been well documented for the last 73 years.
US Government Accountability Office has estimated that the 2007-2009 financial & economic losses of around $22 trillion. This is based on losses on economic output & homeowner’s wealth. And that’s in US alone. I still consider its aftermath still ongoing, so it’s a bit difficult and too early to estimate its total losses at the global level. But I think I can safely assume it’s not going to be $22.000001 trillion. I stand to be corrected of course.
Deepwater Horizon oil spill fiasco is relatively well-documented (there’s even a movie based on it). The estimated total cleanup bill currently is around $65 billion. Let’s not even try to quantify or calculate things that we can’t, like, I don’t know, the lost of 11 lives (may they rest in peace)? Or the damage to the marine & wildlife habitats? What about the tourism industry?
And yet, if we just take these three examples, I wonder where’s an oath for professions like nuclear scientists and engineers, chemical engineers, civil engineers, accountants, economists, lawyers, bankers, financial analysts, fund managers and CEO’s (to name a few). Even if we have the oath of fiduciary duty for financial professionals, why do we still have to end up with this? Or this? Why do we need such oath for a technology that may hypothetically impacted us (in a negative way) but remained silent on the things that actually affects us? Do we fear so much of our future that we forgot to safeguard and taking care of our present? Let’s be honest, even with all these oath of fiduciary duty, can we actually say that we feel safe, everything has been taken care of and there’s nothing for us to worry about?
Face it, we have not actually succeed in at least 2 areas. We haven’t succeed to think of and come up with a meaningful oath for professions that really need them. And we haven’t succeed to ensure that we can and will actually implement that oath.
Don’t believe it? We have USA gymnastics team doctor case as our Exhibit A.
No need Exhibit B. With effort & Google, results are aplenty.
Yes, it’s unfair to generalise all medical professionals are like Larry Nassar or all financial professionals are like Fabrice Tourre. But is it fair to assume and generalise that AI is the only thing that can be turned into a weapon of mass destructions, hence the need of some kind of oath? What’s next? Do we need an oath for all Internet (especially social media) users not to become a ‘keyboard warrior’ or sowing conflict & mistrust? Or an oath for all of us only to use sustainable resources & prevent destruction of the planet? If we’re going to go that path, I wonder how many oaths that we need to take in our lifetime (and how many we actually take into heart & translate it into action).
Intentionally Bad vs Unknowingly Evil
Let’s take Hippocratic Oath for an example. If we look into such oath, we can see that such oath mentioned the dos and the don’ts by medical practitioner. While I’m not saying the original version of the oath is the ultimate guidance since 5th century BC, you can’t deny that it serves us well for the last few millennia. Few scandals aside, we managed to have a very respectful profession, eradicate plagues like polio & malaria significantly and also a breakthrough in medical technology like ingestible chip.
I think I can safely assume that a big majority of the reasons are because medical practitioners are mostly a person whom really care about health service & human being, and financial return (I don’t think the term ROI is appropriate here) is not on the top of their list (important, but not at the top). I mean, it’s not really a secret that to become i.e specialist, it takes years just to get MBBS, a year of internship, another few years of residency and all kind of specialty courses. Not to mention the amount of money & work-related stress involved. So if he or she in this field for money, I’d say he or she already pick a wrong future. Maybe a law school or MBA is the better path for them if money or fame is the objective.
Such oath is more like a rite of passage to them than a dealbreaker. Hippocratic oath (or its modern version of it) already spelled out (albeit in general description) about what to do & what not to do. In other words, it tries to prevent an intention to commit the wrong stuff & encouragement to do right stuff.
But, moving on from medical professionals to other kind of professionals, especially the one that involves a good thing (unintentionally) turned bad, how should we come up with such oath?
Take for example of subprime mortgage lending during early-to-mid 2000s. Once a great idea to fulfil the American Dream of home ownership, the ease to get the loans, even for subprime borrower, has become one of the main contributing factor leading to the crisis. Some still pay for the mistake (or some called it ‘greed’) of the bankers and lawyers. The alphabet soup of financial instruments (anyone still remember CLO or CDS?) has made home lending process sounds more like a doctoral thesis from an Ivy League. Wall Street and Canary Wharf had become Las Vegas and Atlantic City, where the financial instruments are traded to an extent that almost nobody knows who owes what to whom and who’s their real counterparties and liabilities.
10 years ago (plus minus) I was asked why has this financial crisis reverberated worldwide? What makes it so bad that we almost lost the world capitalist & free market. Here’s my answer :
Imagine you live in a small town where everybody knows everybody else. And everybody had sex with everybody else, within that small town (call it ‘intrasex’ if you will). And one day, somebody just announced that he or she has been infected with HIV. In your opinion, what might be the reaction of the people in that small town?’
Damage of trust to each other, right? Nobody knows if they’re next, or is their partner (which, let’s be honest, quite aplenty) ‘clean’? Similar to this f financial crisis. Nobody really knows who holds the actual shit and whom can they trust? For all you know, your closest business partner might actually be your biggest risk. Even if your partner is evidently clean, what about their partners? Or partners’ partners?
And now it’s already 2018, and 2019 in few months time. How many has been prosecuted and sentenced for this? Only a handful.
And where’s the oath for them not to do these crappy things again? Where’s the fiduciary duty? Did I just hear a cricket chirp?
Back to our main topic. Do we need Hippocratic oath in AI? Or for science & technology for that matter? Will such oath helps us in ensuring that technology will be used only for the benefits of humanities?
In my opinion, it helps, but not much. First of all, I’m not sure if it really helps in actually putting humanities before the technological advancement. After all, we have cases like Dr Larry Nassar in breaching such oath. Surely it not really helps (although in all honesty, I’m sure there are a lot of medical professionals that putting patients & humans first and the number of breaching such oath is relatively small).
But such oath may not work much for our startup founders & engineers nowadays. After all, Facebook wasn’t really founded to actually bringing divides among us. Or Twitter wasn’t started to actually sparked the movement of Arab Spring. I’m sure as hell the founders of WhatsApp & Telegram were not actually looking forward for their invention to be used for lynching some innocent people in India or helping ISIS promoting their ideology.
Yes, having an oath helps. But we need something better than an oath.
We need a Devil’s Advocate. Or something close to it.
Working With The Devil (Advocates) That You Know
By definition, Devil’s Advocate is the one who argues against a cause or position, not as a committed opponent but simply for the sake of argument or to determine the validity of the cause or position. Not really someone who thinks negatively about all things on this planet, but rather, a brave one who’s willing to stand and speak up about what might go wrong with their latest invention or how it may be negatively manipulated by someone else. Someone who can think in a manipulative way of how things can go south and suggest on how to mitigate or at least minimise such risks.
OK maybe we can also call it as ‘white hacker’ instead of ‘devil’s advocate’. Only, it doesn’t involve any kind coding or hacking. More towards thinking beyond group consensus. Someone who’s not afraid of not kissing other people’s backside. Someone who’s willing to stand up and said ‘we need to be careful with our next breakthrough, for it can be misused to create a weapon to persecute the innocents, the minorities, the brave and the poor’.
Jennifer Winter wrote a good article on this, so I strongly recommend all of us to read it.
But where can we find such person? How can we develop such person? Or better yet, how can we develop a system or process that anyone can give their devil’s advocate POV without being reprimanded by their working peer? After all, if such breakthrough worth billions of dollars, who will actually stand up against it? I bet not the founders or management. Will Facebookers stand up against Mark Zuckerberg or Sheryl Sandberg if the next breakthrough will brings in billions at the cost of human rights say, in China? Will Googlers do the same to Larry Page, Sergey Brin or Sundar Pichai? Will Goldmanites do it to Lloyd or David Solomon? Or will they just resigned silently from their position as a sign of protest? Or worse, play along?
So in the end, I believe it’s up to us as the global citizen and stakeholder to stand up & voice out our concerns. We can’t just rely on corporate insiders to do our bidding. We need to become the Devil’s Advocate, rather than relying to them or our political representatives. After all, for our political figures, quite many of them just listens to their donors and lobbyists.
To Oath or Not To Oath
Oath for an AI professionals, or any professionals, is a good start, but it should not be the end and assume everything will go fine. We need to be brave to sound off to our corporate behemoths, their founders and their managements on how their products can go wrong or manipulated to do wrong. Just like the knives in our kitchen, every single thing on this planet can be used for evil things. And we, as a responsible citizen, should alert and be alerted on how an innocent breakthrough can turn into a deadly weapons. The founders and the management may have a good intention, but naivety can unravel what they have successfully done. After all, just like what the proverb says ‘the road to hell is paved with good intentions’.
Let’s just hope their (and our) good intentions doesn’t bring us closer to Hell. Ironically, we may need the devil to accomplish it. And for better or worse, that devil might just be… all of us.