Artificial intelligence, or AI, is a relatively new frontier with great power and potential. It’s the “it” thing that drives so many aspects of each of our lives, whether we know it or not. It’s baked within the storylines of popular movies and TV shows that gather huge followings, like Ex Machina or Black Mirror. AI is the cool, lucrative, tech advancement that has everyone talking, for better or for worse. It’s exciting to contemplate where the technology will be three, five, or ten years from now. Still, as both a self-described data nerd who’s passionate about all things data and an analyst at GovEx, I can definitively say: AI is far from perfect.
With all of its power and potential, AI, like so many other technologies, is capable of being problematic. Because AI is so multi-faceted and layered, it calls for constant, consistently thoughtful reflection–which unfortunately hasn’t happened enough until more recently. One effort to inject more of this thoughtfulness into common practice is the Ethics & Algorithms Toolkit. I am part of the team that created this tool, in which we dissect AI in an honest and action-oriented way. We wanted our tool to bridge the gap between data scientists and government practitioners, because we believe in AI.
My colleagues and I believe in AI’s power for social good, but we recognize that especially in government, AI can be an ethically-challenging and complex amalgamation of technologies, applications, and functions. Yes, the truth is that AI can be quite tenuous. For many logical reasons, people are often intimidated by AI. Perhaps we fear what we do not understand, perhaps we fear that we will be replaced by computers in the end, or perhaps we fear losing control of the reins (due to lack of regulation). Regardless of whether you are intimidated by AI or not, it is important that you remain informed about this field.
So whether you love all things data science, are a curious, cautious observer, or just happened to stumble across this post, here are 5 important things you should know about AI.
- AI is not everything for all things. Because AI is not the end all, be all solution to everything, governments should know when to say yes, and when to say no to AI technologies. In fact, my advice is to be wary of anyone who champions AI always. Appreciate a healthy skepticism. Just because AI is seemingly cost effective or simple or fast (sometimes) does not mean that it’s always the answer.
- Remember, AI is a reflection of us (meaning our systems, our biases, and our issues), and governments should know how to investigate its own biases and what do do when they come across them. Self-reflection is key before jumping straight in, since it’s likely that you will have some internal, introspective work to do first.
- Not everyone understands AI. AI is complex, nascent, and ever-developing, thus difficult for anyone to understand. It’s important for governments to learn how to translate relevant AI concepts to residents. Is that important to you? Do you care if everyone can be included in the conversation? Why or why not?
- AI is an investment in your future. Investing in learning, budgeting, or requiring practices surrounding AI would be wise. Don’t know where to begin? Check out our Ethics & Algorithms Toolkit to jump start the conversation. Your government should have a plan for investing in AI, whether that be hiring new personnel, engaging its employees in innovative, self-guided or advanced trainings (such as those offered by GovEx), or planning to have a budget discussion. Investing is important because there are a growing number of systems and applications that governments use that have AI as a large part of their decision making structure. In criminal justice, AI has been used to decide who and who not to release on bail.
- “AI” is not a buzzword. It is a billion dollar industry that employs hundreds of thousands of people, and its implications are hugely important for millions. It has the potential to drastically change the course of people’s lives, which is why it’s important for governments to have meaningful conversations about AI that are honest and actionable. It reflects centuries of work in mathematics to develop functions, concepts, and practices. It has and will continue to change our world.