Photo by Elisa Ventur on Unsplash |
I don't know about you, but I'm done listening to the phrase, "Garbage in, garbage out," in discussions about artificial intelligence (AI). It comes up all the time. Why? Simple, the promise of AI is completely tied to the quality and quantity of your data. And we have to be honest with ourselves. In the nonprofit world we have some significant data challenges.
As the wave of hype crested late Spring 2023 after the advent of approachable AI, the world rolled up their sleeves and started to work on how they might apply these new capabilities. But as they did, people started coming to the realization that the early adopters had taken a shortcut.
They climbed the mountain of other people's data and planted the AI flag proudly. But on their victory lap down the mountain, they have faced an avalanche of legislation, law suites, and confusion from the very people that allowed them to climb the mountain in the first place; the people who generated the data needed to power AI.
The early winners in the AI race won't be affected by the fallout of these boulders for a while because there is so much demand for progress with these tools (and they have so many lawyers). But everyone behind them will have to play by a different set of rules.
And that is where most of us come in. To see the benefits of AI you have to invest significant amounts of money and time into your data. This means your data policies, technical infrastructure, security and data collection strategies have to be well developed. You also need to invest in the ethics of privacy, transparency, and equity.
Come to find out, it takes millions of dollars and many people to give the robots what they will need to do the jobs we now expect of them.
Are you feeling discouraged? I wrote the first half of this post to imitate the roller coaster we have been on over the past 11 months. The amazement, excitement, confusion and deer-in-the-headlights look have all been a feature of this incredible year. A year ago, few people would have imagined what we assume to be true today. And as far as the actions of the early adopters, I understand why their explorations led them where they did. Inventors don't always stop to dot the "i's" and cross the "t's." So while those "i's" and "t's" get worked out, don't be discouraged. But know that we have a big job ahead of us.
Now we must ask, "Where do we go from here?" Are you asking that question with your leadership, teams and colleagues? Let me suggest that our most immediate task is all about data. While our software development staff are experimenting and prototyping the integration of AI, most of us should be investing our time on the data side.
Few organizations know 1) what data they have, 2) the quality of that data, and 3) the accessibility of that data for today's purposes. I've spent a lot of time thinking about this. Back in the early 2000's, my job was to help the nonprofits charged with translating, distributing and licensing the Bible make a major jump in format (for more on this read my chapter in a book I edited called Innovation in Mission). Since then, I have invested hundreds of hours in helping nonprofits think through digital transformation and the data that they would need to power their future plans. My main takeaway from all that work, "Working on your data is not exciting, but without investing in your data, none of the exciting things can happen."
If you are reading this post and wondering about the status of your personal or organizational data, I suggest 3 efforts that will begin to prepare you for the demands on your data that AI will make. As Marissa Mayer, former CEO of Yahoo! said, "With data collection, 'the sooner the better' is always the best answer."
- Digital IQ (DQ): Invest in your own intelligence about data and do the same for those in the organization you serve. One of the organizations helping people think through DQ is the DQ Institute. Their framework is a great place to start as you think about how to raise the level of sophistication and understanding about data and digital life. Andy Crouch's book "The Life We're Looking For" is a wonderful resource in thinking through how faith and digital life intersect.
- Ethics: Sure you have a privacy policy on your website, but how much work have you done thinking through the ethics of collecting data from your beneficiaries and the permissions you may need to do what AI will eventually make possible. Is is also important to consider that just because you can do something with AI doesn't mean you should. How will you decide? From a social sector perspective the latest AI Now report is a helpful resource. From a Christian perspective, Jason Thacker's thought leadership provides a good starting place. In this research Gretchen Huizinga explores the way our worldview impacts our approach to AI ethics. Finally, Mark Graves dives into the impact that generative AI will have on scientific and theological research.
- Data Inventory: One of the most helpful exercises is to know what knowledge you are responsible for stewarding (I wrote a book about Knowledge Stewardship that is available by request). You can't even consider how to engage with AI unless you know what you have, where it is and what format it is in. Some people and organizations have done amazing jobs of managing their data and so this is a fairly straightforward activity. For others this will be a huge challenge and will take a significant amount of money and time. This Data Sharing Toolkit walks you through how to do a data inventory.
Comments