The Tech Industry Has a Responsibility to Do More Than Just “Move Fast and Break Things”
The tech industry, more specifically the AI space, has been thrust into the spotlight once again. Companies are being held accountable for making short-term decisions that make a long-standing impact… and it’s about time.
The companies that push their “ghost workers” or “crowd” to standardize their annotated data that drives down cost and increases speed of delivery finally being exposed on causing irreversible damage on influencing bias in algorithms in favor of profits. The same companies have recently been exposed for other unethical practices such as asking the same “ghost workers” to describe their skin tone in the job application process, thinking that by weeding out certain applicants because of their skin color will help in ensuring they’re not biased in their annotation.
Why have these practices even existed in the first place? I think it can be explained by the mantras that dominate the tech industry itself. We’re taught to “move fast and break things” and while that is great for innovation, translating this into business decisions doesn’t always equate to doing what’s right for society.
You don’t need to be the CEO to have your company’s values guide you to the right decision in difficult situations. The importance that values have as an organization is not simply to attract talent to work for you or inspire people to work harder but rather to guide employees and businesses themselves when facing difficult decisions. These decisions aren’t just what guide leaders of organizations but ones that individuals are facing every day in their work life.
Short Term Sacrifice, Long Term Impact
While AI is pioneering the future, businesses have long faced difficult decisions, as doing the right thing for society can often come at a sacrifice to profits. There are inspirational examples of businesses that were faced with making difficult business decisions that ended up being good for society. In hindsight these decisions seem like a no-brainer but in the moment, they were certainly anything but simple as they were bad for business.
One example of this came in 1982 when Johnson & Johnson’s worst nightmare became a reality. Seven people died in Chicago after taking Extra Strength Tylenol, their best selling product, that was laced with cyanide through tampering. Most predicted that Tylenol would never recover from the sabotage, yet it became a hero moment for the company. The decision as to what to do was placed on the company’s chairman, James Burke, who quickly decided to pull the product from not just the shelves of the Chicago stores where the tampering occurred but the entire product line of 31 million bottles. Those Tylenol capsules that were pulled were replaced with another product in the safer tablet form free of charge. At this time, recalling products was unheard of and this one cost J&J more than $100M and a few years later another $25M with the relaunch of safer Tylenol non-tampering bottles.
The move turned out to be the right one. The stock was teetering in the short term during the recall but quickly steadied and recovered two months later to record highs. The stock has paid out investors in dividends ever since. The interesting part of this entire story was when faced with this difficult decision, James Burke was calm and steady in his decision making. When asked how Burke seemed so in control of the decision at the time, he simply stated that the J&J Credo provided his guiding light.
The Credo, which is nearly 350 words, lays out the guiding light in the first sentence: “We believe our first responsibility is to the patients, doctors and nurses, to mothers and fathers and all others who use our products and services”.
When faced with difficult moments, the values of the company provide the lighthouse to where to steer the ship.
Saving the Soul of AI
The long road isn’t often celebrated in tech, where you’re constantly pushed for instant ROI without a discussion of downstream consequences. But these are exactly the discussions we should be having, especially in AI. We’re finally beginning to hear the world’s best brands begin to detach themselves from the AI tools that source from the “crowd”, but frankly, it’s long past due and the voices aren’t loud enough to make a significant difference.
From my conversations with Data Scientists, they overwhelmingly believe that ethics, fairness and bias in AI are real problems that can be solved today before we go down an irreversible path. There are brilliant researchers who often root against the success of their own research until they’re confident that businesses will make the right decisions. This harkens back to 1941, where researchers retracted papers they submitted to Physical Review on plutonium, holding them until the end of World War II. Do we have to wait for AI’s Hiroshima moment or can we take pre-emptive steps?
There are many facets of doing the right thing in AI, but regardless of industry, it starts with the data supply chain: from data sourcing, to data annotation to data enrichment and management. AI is not yet fully regulated and doing this rigorous work can be hard and go unnoticed. Still, the regulation of artificial intelligence is coming. The short cuts businesses tmake when deciding how their AI supply chain is sourced will catch up to and cost more in the long run, and rightfully so. To pre-empt businesses from making the wrong decision in these moments we need to continually force the conversation, demand that innovation does not have to come at the detriment to humanity and have consequences for those that are not centered on this.
Why I Work at Sama
One of the many reasons why I chose to work at Sama, one of the only B-Corp certified companies in AI, is the strong foundation built on integrity, humanity and the importance of diversity at all levels. These values have guided us through many difficult times, helping us ensure we do the right thing whether that’s taking care of our data labelers during COVID-19 times, or how we approach building the world’s most trusted human in the loop annotation tool that leverages the diversity of our labelers. We’re constantly tested as a business and while it’s tempting to take the short cut it’s typically true that the right thing to do is to stand by our core values regardless of the cost.
At Sama, we believe the right decision for your data supply chain should start with partnering with companies that have a technology-enabled impact story that increases economic opportunity in underserved communities rather than crunches it through margin. To keep us honest, there was a randomized controlled trial through MIT researchers over 3 years, to measure exactly that: Whether a flourishing business can also deliver significant, measurable impact and stand for ethics in AI.
I work at Sama to continually be a part of something bigger that creates growth the right way, because doing the right thing should always be celebrated over short-term, ruthless business decisions. You can be a part of saving AI’s soul and still run an amazing tech business.
Why do you choose to work where you work?