The Dangers of Giving Tech an Ethical Blank Check

Technology has the potential to bring out the very best of us, as well as the very worst of us. This power becomes increasingly more important as tech becomes omnipresent: cell phones are working their way into rural rice paddies in Asia; drones are delivering emergency supplies to parts of Africa that are impassible by roads; and everything from our lights to our security cameras and the locks on our doors are part of the Internet. Computers are not yet omniscient but they store, process and learn such vast quantities of data that we can barely comprehend their magnitude. They are already beating our best players at our own games (see Google DeepMind’s AI crush the world champion GO player).

As we design and build programs that think for themselves, we can only hope we are able to instill in computers a sense of right and wrong before they become omnipotent. Of course, even the notion of what is right and what is wrong is subjective and varies greatly culture by culture. It is time — in fact, it is a bit late — for us to start thinking and debating about what powers we endow these intelligent machines with.

Over the past 10 years, technology has brought about seismic changes in many different areas, including robotics, autonomous vehicles, genomics, synthetic biology and more. Many of these categories didn’t even exist a decade ago.

The rate of change continues to accelerate exponentially — so much so that it is getting harder to keep up. Moore’s Law has kicked into warp speed as everyone now has access to a supercomputer in their pockets; costs of storage, processing and distribution are close to zero; and the world is much more connected. This globalization and democratization of innovation leads to impacts that are bigger, and distributed faster and more widely. Tech development has gone bottoms-up.

These are uncharted waters. In his book Thank you for being late Thomas Friedman references Brynjolfsson and McAfee, who sum it perfectly: “when you keep doubling something for 50 years, you start to get to some very big numbers and eventually you start to see some very funky things that you have never seen before […] This is what happens when the rate of change and the acceleration of the rate of change both increase at the same time — and we haven’t seen anything yet.”

With reality evolving faster than ever before, unintended consequences are more likely and potentially more perverse. If we are not thoughtful and do not ask hard questions ahead of time, we run a real risk of negative events that can dampen the positive change that technology can bring about.

Such tension is ubiquitous these days:

· Airbnb, Uber and Lyft, while seeking to democratize travel, also enable discrimination. An HBS study found that requests from Airbnb guests with distinctively African-American names were 16 percent less likely to be accepted than those with white-sounding names; an NBER study found that one is more likely to wait 35 minutes longer or have 2x a cancellation rate if you are African-American or have an African-American sounding name.

· Twitter and Facebook make it easier for us to share our thoughts and experiences but their algorithms allow for powerful broadcasts of fake news. For instance, NYTimes reported that a fake story claiming that Pope Francis (a vocal refugee advocate) endorsed Donald Trump, was shared almost a million times. Given the reach and undeniable influence of such networks, they need to rethink their algorithms to have a more nuanced and accurate notion of “relevance.”

· Automation and robotics will make more of us more efficient but they also threaten to put millions of people out of work. A McKinsey analysis showed that existing technologies could automate 45% of the activities people are paid to perform and that about 60% of all occupations could see 30% of more of their constituent activities automated. These numbers will increase (and fast) as new technology is developed.

· Facial recognition software have the potential to keep us safe by identifying terrorists and criminals, but they, too, can be biased and inaccurate. An MIT Tech Review story pointed to a Michigan State University study of several commercially available software (including from vendors used by law enforcement) that showed less accuracy on women and African-American and younger people.

· IoT products, commerce engines and communication tools often violate the privacy and security of their customers. As 3.4 billion Internet users and as many as 15 billion IoT devices globally capture and share data about users and their behaviors, more vulnerability is introduced to the system. An FTC Report estimated that 10 thousand households can generate 150 million datapoints per day. There is no transparency about what data is captured, what is done with such data and users have no effective choice about what they share. User privacy is at is most vulnerable point as more sensitive information is being created and there are increasingly more points of attack (see Dyn DDos attack as an example where hackers used DVRs and IP cameras to disable part of the Internet).

· Scientists have started conducting human genetic engineering. According to Nature, two different teams in China have reported editing genes of non-viable embryos (one modifying a gene linked to blood disease and the other trying to make them resistant to HIV). Even though there are still many technical difficulties to be overcome, they raise a number of ethical, regulatory and scientific questions that are far from being answered.

AI, drones, space, self-driving cars, bioinformatics and genomics, new computing platforms (such as human computing), food technology, cleantech and many others will continue to rattle our notions of right and wrong, good and evil.

I am excited about what the future holds and the wonders that tech can enable — especially its role in making people live better, healthier and longer. But we can’t give tech a blank check. We need to pause and be deliberate about how to reprogram the world around technology such that it achieves its full potential in moving society forward.

I plan to continue to explore this theme from two angles: as an investor in new tech startups and someone who thinks out loud. If you are also thinking about these topics or working on these types of companies, I hope you will reach out to [email protected].