In an October 2019 interview with Jasper Kuria, Brad Smith; Microsoft President and co-author of a New York Times bestseller explained why the world urgently needs a Digital Geneva Convention and a Hippocratic Oath for software engineers.
Today, what Smith had to say on the ethics of technological progress at last year’s Capital & Growth + Hacker News meetup is more relevant and urgent than ever. The Need for a New Field of Ethics for AI
Tools and Weapons: The Promise and the Peril of the Digital Age by Brad Smith and Carol Ann Browne came out in September 2019 from Penguin Random House with a foreword by Bill Gates.
In their book, Brad Smith and Carol Ann Browne wrote that if something is new and not yet illegal (because the law has not caught up with technology), but confers a decisive competitive advantage, people tend to do it.
For example, Facebook decided early on to use data however they wanted to attain a dominant position in the market. Paying a $500 fine per transgression wasn’t an issue as long as it helped Facebook achieve this goal.
But Smith is a strong believer that as the first generation of humans to endow computers with the ability to make decisions, we must decide how we want machines to make these decisions consulting an ethical code.
“Almost every person in this country has seen a movie that has the following basic plot structure: Humanity invents the machine that can think. The machine thinks. The machine decides to enslave or destroy humanity. Humanity spends the rest of the movie trying to turn off the machine.” A Hippocratic Oath for Software Developers
About two years ago, Microsoft publicly asked a question for the first time: “Should there be a Hippocratic Oath for software developers creating AI systems the way there is for doctors?”
You can’t graduate from the Air Force Academy without taking an ethics course. In the military, there is a code of justice and there are people whose sole job is to make sure this code is followed, even though this doesn’t mean that the military makes no mistakes.
There are no common ethics codes to determine how lethal autonomous weapons and systems that are developed for the military should be used once they end up in the hands of civilians.
“In the top 10 Computer Science departments in the nation, there is only one that requires taking an ethics course to graduate. Ethics is a field that will have to get infused into Computer Science education. There should be a stand-alone course called Ethics for AI that every computer science major must take.”
Brad Smith and Harry Shum (an Executive Vice President at Microsoft) developed the idea for a Hippocratic Oath for Software Engineers and devised six ethical principles in Microsoft.
Shortly after, Oren Etzioni, a professor at the University of Washington got his students to draft an oath, which you can read here.
“You need to know what principles you hold fundamental.”
Shortly before this interview, Apple had removed the Taiwan flag from their Emoji set, at the behest of the Chinese government (China is Apple’s second most important market after the US, accounting for almost $10 billion in revenue).
We asked Smith, what would you do if the Chinese government asked you to remove the Taiwanese character set from Windows or be barred from doing business there, and forego billions of dollars in revenue?
His response: While China is also an important market for Microsoft, Smith explained that they had to step back from setting up a data center there because of the legal and security issues that it could create for their Chinese users. “You need to know what principles you hold fundamental”, he added. “Keep your humanity.”
Smith’s visit to the Vatican, particularly the Sistine Library, has been the highlight of his long and exceptional career. He was invited to the Vatican in early 2019 to talk about artificial intelligence and ethics with Pope Francis, who looked him in the eye and told him to keep his humanity, no matter what.
In the Sistine Library, Smith got to view the Gutenberg Bibles, which were produced using Gutenberg’s original mechanical printing press in 1450.
“We were turning these big, colorful, and artistic pages. But what we realized is that the real message was how technology and the Gutenberg printing press advanced and changed religion. All of a sudden, you could produce more Bibles. No longer was the reading of the Bible something that was exclusively done by a priest.”
What left an even greater impact on Smith were Galileo’s letters to the pope to proclaim his innocence in the 1600s. The case of Galileo, who was imprisoned by the pope for heresy for saying that the Earth rotated around the Sun, is an example of a very different collision between technology and religion.
“In the next three decades, the world will be dominated by AI as probably the single most important technological force. I don't think it's an exaggeration to think about the role AI will play between now and 2050. The world didn’t think that the combustion engine would play a part in the first half of the 20th century and yet reshaped every part of the world. AI is going to impact everybody.”
According to Brad Smith, just like it is the Pope’s job to bring religion closer to today’s technology, it is the software developer’s job to bring technology closer to the humanities.
We Need a Digital Geneva Convention
Smith has proposed a Digital Geneva Convention to protect civilians against state-sponsored cyber-attacks during times of peace (much like the original Geneva Convention protects civilians during war).
In November 2018, his team worked with the French government to create the Paris Call for Trust and Security in Cyberspace. It has now brought together more than sixty governments, a hundred and thirty NGOs, and three hundred companies, all of which pledged their support for the principles Smith and his team were advocating, including the protection of elections.
Does Smith think the rogue nations behind state-sponsored cyber-attacks would abide by such a convention?
“All the nations of the world may sign up for the kinds of principles and rules we've described now, but will everybody follow them? Will everybody even sign them? Well, we don't yet know. But if you get these kinds of laws or principles adopted, it makes it easier to galvanize everyone to stand up and defend them. It makes it easier to hold the violators accountable.” Who will Set the Rules?
Smith told us that he was hopeful the States would become an active participant in setting such rules, but Europe would certainly be a key player.
“It's easier for Europe to move to reason. They don't have a dominant AI Center the way the US or China do, so I can tell it’s easier for them to go back this way. But I also have hope for the United States, not so much focused on Washington DC but some of the states, such as California.” Edward Snowden: Hero or Traitor?
In his book, Smith wrote that for many people, Snowden was both a hero and a traitor. Smith and his co-author didn’t take sides in the debate, but they framed the arguments. “I think what he did made the world smarter,” Smith told us. “He shared information that the world needed to have.”
“I’m not saying he shouldn't have any accountability. I recognize that there is accountability in what he did and I think that is in the essence of civil disobedience and it has been since Martin Luther King in the 1960s. It wasn’t people who were willing to take risks and said it should have no consequences. It was people who said that there would be consequences but the consequences in their minds were worth what they were trying to accomplish.” Using Data to Advance Science
In the interview, Brad Smith underlined the urgent need for stronger privacy laws for the United States. These laws need to evolve as technology advances, and make sure private information about individuals stays off-limits.
"I never met a data scientist who didn't want more data,” Smith said. “We need privacy laws that will protect people's rights while enabling data to be used in multiple ways, especially to advance. “
Will AI Replace Human Jobs?
Smith’s book encouraged those who worry about AI replacing human jobs to look at history. In the year 1950, the automobile industry created six times more jobs than it destroyed, even though halfway between 1932 and then unemployment was at its peak. If we gain new skills to keep up with technology, AI can only help us create more jobs. Does Technology Connect or Divide Us?
In his book, Smith pointed out how every technology that has been able to bring people together also had the potential to divide. Automobiles that let you drive to your relatives in other cities or telephones that let you call them are two examples of such devices. Ultimately, it is up to us to decide how to use them.
“When I started at Microsoft in the 1990s, what was so odd to me was that when I walked down the hallway I never heard the telephone ring because even in the 1990s people were just emailing each other. Sometimes, I would see an employee sending an email to another employee when their offices were right next door to each other and I’d ask, Have you two considered going to lunch?” Racial, Economic, and Gender Divide
Smith concluded by pointing out how the real divide we need to watch out for when it comes to technology is the racial, economic, and gender gap. It can widen if we do not take the right steps today.
“If you look today at who is studying an AP course for computer science in an American high school, what you will see is a group of people that are more male, white, urban, and more affluent than the country as a whole. That is what we need to address, the real challenge is how to open the doors of these fields of the future so that they're more accessible to everyone. If they're not, then I fear that technology will exacerbate and rather than close the racial, economic, and gender divide.”
Liked this? Check out more here.