Ethical AI will only come from Ethical AI Companies.

By Dr. Phillip Alvelda

June 5, 2019

Is humanity already losing a war with Artificial Intelligence?

This may sound like a bit of an alarmist question, specially coming from the CEO of an AI company, but bear with me.

Last year, Tristan Harris captured my imagination with a startling insight. In his words, “Every time you touch that blue F icon on your phone, you fire up the world’s largest artificial intelligence that has a sole mission of contending to capture and keep your attention for as long as possible to maximize ad revenues. And like Gary Kasparov versus Deep Blue in chess, and Lee Sedol versus AlphaGo in Go, you will lose.”

And he was right, of course, and not just about losing control of our attention to social media in service of advertising revenues. These AI systems are already EVERYWHERE, from how online goods, airfare and train tickets are priced, to what media you find and pay for, to, as I recently discovered, which BMW features are bundled together to extract an extra 20% fee for features you didn’t actually want in the first place…all individually customized and tuned by sophisticated AI systems in order to extract maximum revenues from unwitting consumers.

The result is a digital economy that is increasingly weaponized to drain individual savings into AI-enabled corporate shareholder accounts. Individual consumers with merely human sensibilities are economically defenseless against the growing onslaught of automated AIs arrayed against them.

Need some evidence that this transformation is gathering steam? Simply note how corporate productivity is increasingly decoupled from real wages. Or notice how last century’s stock exchange leaders have all been superseded by the AI-enabled information companies. Or just realize that much of the global wave of populist nationalism is directly attributable to the increasing rate of economic disenfranchisement as middle class and rural jobs vanish in the face of accelerating automation in the information industry.

Jared Diamond has always maintained that throughout history, societies typically collapsed when their ruling class managed to accrue enough political power to become so extractive from the economy that they financially undermined the general social contract with the middle class. Our new AI tools have enabled a whole new level of economic extraction with unprecedented efficiency, and the power of this extraction is accelerating at a ferocious rate. It is already quite literally tearing at the economic fabric that binds our society together, a situation looking to dramatically worsen in the coming years.

So is there a solution in sight, or are we doomed to a dystopian AI-driven economic deconstruction of our society? The good news is that I fundamentally believe there are plenty of opportunities to build AI-empowered industries dedicated to support and empower consumers and better individual lives, rather than simply to extract increasing financial value from the economy at scale. The bad news is that it will take purposeful efforts driven by companies able to maintain strongly-held and unusually high, ethical standards. The steadfast few must finance growth through the monetization of contribution rather than extraction.

I can tell you first-hand that there are mighty capitalist forces arrayed to seduce the greedy or even just the unwary entrepreneur. At every turn financiers ask questions like “Why don’t you concentrate on where the money is?” A couple of weeks ago, I actually had a venture capitalist suggest, “…forget all this healthcare delivery and access enablement stuff. Can you apply your technology to help insurance companies extract more revenues from premium customers or better reject customers who can’t pay?” I remember thinking in that very moment, “Well, so much for that meeting. Hope he enjoys counting his money when the pitchforks and torches are coming.”

But even saying something like that to myself was difficult. Any even moderately experienced startup founder can instantly tell you their company’s cash burn rate and the expected drop-dead date when the bank account runs dry and the doors are shuttered. The temptation to compromise values for precious capital or to advance the automation of extractive rent-taking industries that are already tapped-in to large cash flows is mighty. But it can be done. And while rare, there are highly ethical angel and venture investors who see and support staggering opportunities offering both financial and social benefit in parallel to benefit an abundant, ethical society.

This is a mission we take very seriously at Brainworks because we are keenly aware that the very same AI technologies that promise to transform healthcare, transportation, energy, education and more, can also be used to disenfranchise, bankrupt, and even oppress the citizens of the world. The world changing technologies we shape are very much dual-edged swords whose application will reflect the very ethics of our designs.

I’ve come to see society balanced on an ethical knife’s edge almost entirely unaware. The governments and companies we choose to trust, and patronize, and empower, even with seemingly innocuous things like which friends and posts appear in our social media feeds, or which programs populate our TV channel guide, are the very companies who are using your payments to develop ever more powerful and consequential services. Google had in their initial public listing documents a company assertion that they would “not be evil,” a mantra that has subsequently been erased from company literature. Facebooks recent ethical and privacy lapses still resonate with distrust over our recent Presidential election that was corrupted in large part by their systems. Would you trust Facebook, Google, Amazon, or Apple to make life and death medical treatment decisions for you? It is critically important for society to understand that by patronizing one AI superpower over another today, even for what seem mere petty conveniences, we are setting that company on the path to manage and direct, even to control, more and more of our lives.

It’s true that privacy has become more of a national focus following the Cambridge Analytical abuses. But so far, most of these conversations look backwards, seeking to prevent repeated abuse, and are largely conducted by the general public and politicians lacking even the faintest technical clue of difficulty, opportunity, or consequence regarding topics they are simply untrained to understand. By default then, these decisions are being made by the companies developing and holding these powerful AI technologies, and the ethical standards each of them maintain respectively, are by similar default, becoming the core ethics of our society.

So please realize that ethical AI systems will only be created by deeply ethical companies. Choose wisely which companies you patronize today, for they will be the creators of tomorrow’s Artificial Intelligence overlords.