Hypergiant Industries builds AI systems for enterprises and governments around the world. We were fortunate to have the chance to interview Will Griffin, Hypergiant’s VP of Ethics and Diversity, to talk about the challenges of bringing ethical awareness to the day to day operations of a major AI systems developer.
As the VP of Ethics and Diversity in AI at Hyper Giant, you have joined a small but growing number of professionals with ‘ethics’ in their title. What led you to taking this position with a growing tech startup?
Our Co-Founder and CEO, Ben Lamm’s vision of “delivering on the future we were promised”, inspired me. Ben is an uber-successful serial entrepreneur with multiple exits to his credit. I know Hypergiant will be successful and, combined with a heavy emphasis on ethics; we have the potential to re-shape the way technology companies think about their obligations to society.
Going back even before your current position, what aspect of AI Ethics first caught your attention, and how has that shaped your view of technology?
Over my career I have primarily worked at the intersection of media and technology (including launching the first original content on the PlayStation Network and the first video on demand channel on cable with Comcast). There was very little industry regulation and the goal was always to be first. We followed the Silicon Valley mantra of “go fast and break things.” Unfortunately, we have seen the negative impacts of that approach – especially on privacy. Since AI has the capacity to transform society in ways not previously imagined, I feel an obligation to be a part of making sure AI innovation works for the benefit of humanity.
What has surprised you most about your work as the VP of Ethics?
I have been surprised by two things. 1) The amount of emphasis and demands the public and policymakers are now placing on ethics in technology and 2) the number of technologists willing to admit how little they know about embedding ethics in their workflows and the products they make.
AI Code of Ethics
Hyper Giant has a heavy emphasis on the use of ethics in AI development, and even has an AI Code of Ethics, which is something rare to see in a tech startup. What led to the decision in the first place to create this code, and what was it like developing it?
The decision to create the AI Code of Ethics came from the top, Ben Lamm. I was brought in specifically to ensure that we stay at the forefront of best practices in operationalizing ethics in our AI workflows and educating our team, clients, and partners on the importance of embedding ethical reasoning at every step of the process. Developing our ethical code and framework has been a lot of fun. As part of the process I visit, and vet our approach, with some of the top thought leaders on ethics in AI. Sangeeta Mudnal (Group Project Manager, for Microsoft’s Responsible AI initiative) is on our Board of Advisors and is very helpful. Frank Buytendijk , head of Ethics at Gartner Research is a great sounding board and thought partner. The professors at Stanford’s AI Lab have been gracious in sharing their approach to ethics education with their computer science students.
Can you give an example of a time where you needed to make technical compromises to align with your ethical code?
To date our AI uses cases have not caused us to make technical compromises. Instead, we have to make changes to our AI development workflows. In the old days (pre-2019), AI developers would be presented with a business problem and immediately begin crafting a technical solution and the thought of ethics (or compliance) was an afterthought. Now we incorporate ethical reasoning at the outset and it forces the team to create a range of technical solutions to a given problem and choose the solution with the least ethical challenges and most people-centered beneficial outcomes. In the end, it makes the team more creative because we have to think of more solutions to client problems.
There are many accusations that an ethics code only leads to ethics washing, as there is no proof that a company actually follows such a code. How do you ensure that this is not the case with Hyper Giant?
The AI Code of Ethics has been helpful, but the mindset that created it is the secret sauce. Every company should (and likely will) develop a statement for public relations reasons, but the companies that embed ethical reasoning into their workflows and culture are the ones who will own the future. We have found that becoming well-versed into ethical reasoning has increased our overall creativity and led our product teams to develop more robust AI solutions.
How has having an AI Code of Ethics helped shape the work HyperGiant is doing?
Boards of Directors and C-suite executives are well-aware of the well-publicized cases of ethical failures in the technology industry. The image of Mark Zuckerberg sitting in front of Congress being taken to task for Facebook’s myriad ethical failures, including billions of dollars in fines, has prompted introspection on the part of senior leaders. This introspection has increased the calls for companies with a well-defined set of values and tangible experience embedding ethics into the products they create. This has increased the number of pitches we have invited into, but more importantly demonstrating ethics has allowed us to expand our scope of work and amount of projects with our current clients.
Can you describe an example where it was hard work to help your client’s adjust their expectations so that they would be compatible with your ethics code?
I would not want to identify a specific client, but I can speak to issues that apply to every client. Embedding ethics into AI development workflows is challenging at first. The first step is getting the CEO and other responsible C-suite executives to mandate compliance with the AI Code of Ethics. The second step is traing the relevant stakeholders in our ethical reasoning model and vetting all proposed use cases through the model. Once clients see the creativity unleashed and the number of technical solutions they are able to create, they began to embrace the model as a value-add to their overall workflow.
Ethics in Technology
What do you believe is the biggest challenge we are facing in AI Ethics?
The biggest challenge is the prevailing Silicon Valley conventional wisdom that ethical mishaps are just the price to be paid for being first (“move fast, break things”). With the prediction that AI leadership will be worth trillions of dollars, it is easy to see why many companies will not want to change their approach to product development and deployment. We view our role as educating the industry on the economic value of ethics in AI and the trust it will create within the marketplace. We believe the trust generated is what will be required to compete and gain market leadership in the AI-driven world.
At Ethical Intelligence, we are working to change the narrative around AI Ethics from ethics being a blocker in technology to ethics being an asset to long-term innovation. What inspires you about the field of AI Ethics, and what good do you see if doing in the world of AI?
I agree wholeheartedly with Ethical Intelligence’s approach. I am inspired by all of the discussion (and research) in industry, academia, and the media to raise awareness of the importance of ethics in AI. If privacy had this much attention at the early stages of the internet, I believe social media and other platforms would have evolved with many more protections built-in. The growth of AI gives us another chance to get ethics in technology right, and I am excited to be a part of that process.
Anything further you would like to share?
Ethics equals trust. Trust equals sustained economic value creation.