Doing Good Without Being Left Behind
Many organizations are telling their people to use AI without any guardrails or pre-planned outcomes. More and more people are using it to increase their output without improving impact. It’s become a box-checking exercise.
At the same time, others are avoiding it because they aren’t sure what to do with it. They have analysis paralysis. And they care about preserving the natural environment, critical thinking, and continuing to align their organization’s work with their environmental, social, and governance (ESG) goals.
In a time when AI—generative artificial intelligence, really—is everywhere and on everyone’s mind, is there a way to use it and keep the nice things we have? I think so. Like so many other things, moderation is key!
Dimensions of responsible AI use
If we can’t put the AI cat back in the bag, how do we use it responsibly?
There are multiple dimensions of responsible use:
Environmental – AI has impacts on our environment through data center electricity use (requiring alarming amounts of energy production), water consumption to keep servers cool, and land use affecting habitats and ecosystems. One study estimated it costs about 30 times as much energy to generate text versus simply extracting it from a source. [Scientific American]
Workforce – Taking humans out of the loop at the foundation results in a loss of knowledge and wisdom. Using AI for junior and mid-level jobs means depleting the pipeline of people who become senior, where decisions are best made by humans.
Societal – Using AI to outsource thinking is making us dumber. Studies show that students who use AI for a first draft of an essay cannot answer basic questions about the topic, while those who write their own first draft and edit with AI exhibit a sound understanding of the topic. [Range Widely] What happens when we no longer have to think critically or form our own thoughts? A dark age. Those who continue to explore the humanities will be our future saviors.
Ethical – Bias, security, amplification, privacy, and inclusiveness are all ways we need to think about what we feed the model and use it for. Every day, AI exploits data about people. Well-meaning people share personal information with unsecure systems unknowingly. When we use it to filter out job applicants, we are imposing unconscious bias.
It’s easy to barrel down the path of AI at any cost, with predictable outcomes, most of which are detrimental to humanity and business. It’s easy to spend a lot of money, reduce the workforce, and not be better off in a year than you are now.
Responsible use of AI
Responsible use, then, slows us down so we can use AI to achieve outcomes that improve human quality of life—or skip it when it’s unnecessary.
Responsible use of AI includes:
Use AI for select and discrete tasks and activities — Knowing what types of things AI is good at and exploring those for your organization helps you find efficiencies that support your team and contribute to the bottom line. Think repeatable tasks, pattern finding, and computation.
Identify a problem for AI to solve – Once you know what AI can actually do, pick the problems that match its strengths. When you know what problem it is solving, you can define the outcome it should achieve. For example, when you cannot fix all the systems at the bottom of your tech stack but need to change how people work with data and knowledge to serve their customers. AI can be the orchestration layer that makes the impossible possible—now, not in 5 years.
Use non-AI-powered methods or tools – Many of the things people want to use AI for can be achieved with an existing product or methodology without AI. We’ve been using algorithms for decades to sort and filter information, no AI necessary if you structure and classify it appropriately. Putting in place governance and operations could avoid a technology investment in the first place.
Pick the right tool – When you have clear goals and objectives, you can find the best tool for the purpose. Whether that’s an all-in-one solution or a targeted one that does one thing well, choosing the right product will save money and time. No sunk costs or investing in the wrong thing.
Choose ethical vendors – Live your values and work only with companies who have documented and transparent standards and policies that match your organization’s principles. There is a lot of risk in choosing black boxes.
Create a continuous feedback loop – Try something and compare what happens to your stated goals and objectives. Decide what to improve, expand, or phase-out. Continuing to do something that isn’t working for you creates risk and cost that multiples the longer you do it.
Reclaim time for humans – What will you do with the time savings that AI brings you? Why not give people that time back? Giving people the time they save using AI to leave early, start later, or take an extra day off will motivate people to use AI effectively. If all it does is put more on their plate, why bother?
I see and hear about a lot of people finding ways to use AI because they are told to. Not because it makes sense or improves their lives. It’s costing us dearly.
It is possible to use AI and still be good. It’s not a tradeoff, but it is a deliberate decision.
Can I help?
For over 20 years, I’ve been helping small and mid-sized organizations build foundations that support them now and in the future. From inaugural websites to mobile-friendliness to social media to designing connected content, organizations that start with why get better outcomes and have more resilient teams. The best time to have made a transformation was last year. The second best time is now.