Thoughts on AI on the Eve of Copilot

The Equity Problem

Recently I had coffee with a new friend I met at a business event. These days, you can’t seem to get to know someone without asking about their opinions on #AI. As someone who spent most of her life consulting on Technology Strategy, I am excited about the possibilities, and a bit wary of the hype. That said, I use AI regularly throughout my day, but I don’t obsess over it. It is a utility for me to make space for meaningful work (in fact, I used it to help me write this). When my new friend asked, “How are organizations going to make access to AI equitable?” it stopped me in my tracks.

Oh, right - can everyone get access to the best AI tools? Is that important? And off my brain went…

As organizations scramble to decide what to do about AI, what to allow and not allow, individual users are embracing it (whether their employers want them to or not). Understandably, companies are concerned about IP and data leakage causing some to create internal products in lieu of free or commercial ones. While these tools are good, I have doubts any homegrown solution is going to beat what vendors like Microsoft and Google are going to offer. The benefits are huge: these tools allow employees to leverage AI inside their existing productivity software to maximize output, without allowing their confidential company info to leak out into open-source ecosystems.?

But while private GAI technology is available… it has a price. Currently, private services range from $25 to $50/user/month. As of this writing, my IT team says Microsoft Copilot will have a minimum license purchase of 300 users. That is NOT an insignificant investment for many companies and certainly has an impact at larger scales.

So how do organizations take advantage of best-in-class GAI services while also justifying its cost? Do they have a responsibility to figure this out? I argue they do.

Democratizing Innovation

If we only empower certain groups within a company to use AI, their voices are the only ones that will shape these technologies as they continue to develop (AI is a reflection of the data it stores and has access to). This will limit innovation if whole segments of a workforce feel disenfranchised from these new capabilities.

The democratization of innovation is one of the greatest promises of AI, as it helps uncover untapped potential. Your worst-performing employee could quickly become your best if they have access to the right support at the right time. Historically, manual efforts have failed at this.

For example, I have ADHD and it is a huge energy drain in order for me to perform like someone with a “typical” brain. With my AI assistant attending meetings with me, I can relax and focus on meaningful connections, instead of worrying I’m going to miss an important detail. I know my robot got it, and I know it can help me unpack what I heard later. This tool allows me to be fully present and keep track of my commitments and tasks, historically challenging for neurotypical brains. With those challenges removed, I have way more energy to commit to areas of passion and innovation, like finding meaningful connections with clients or developing helpful content that empowers everyone to leverage tech.

While robots are cool, we must take a people-centered view of the technology's potential. If only certain demographics or job roles within a company have access to AI, it limits the potential for untapped ideas and talent to emerge. A diverse range of perspectives feeding into AI data and application development will also help ensure the technology works for all.

Rather than see AI as a way to replace underperforming staff, giving all employees access could help uncover new skills and strengths with the right assistance. It could be a company’s new differentiator, not its biggest risk.

Give the Choice to Your Employees

At my company, each employee has a learning stipend that can be used for professional development. There are some principles and guidelines, but overall we are not policed on how we use those funds. I see no reason this could not extend to giving employees technology stipends to choose their own AI tools and other services. This puts ownership over the technology into individuals' hands while still addressing equity concerns. It signals trust in staff to make choices benefiting their roles.

Now before all my Cyber Security friends come for me, I’m not suggesting a free-for-all here. IT could work with Business Operations to create a marketplace of approved tools and offer training and support for those who choose to use them. My hope is that innovative and future-focused companies put effort into securely selecting and configuring approved commercial tools rather than trying to block new technologies or develop homegrown solutions that limit innovation. Based on the speed of releases nowadays, it will be pretty hard to beat the performance promised by Copilot and others.?

As AI becomes more integrated into our lives and workplaces, these ideas around inclusion, empowerment, and people-centered design will be crucial to realizing the full benefits of AI. If not, we will continue to create a rather boring echo chamber of the loudest people in the room.?

Bonus: How I used AI to help me write this article

As someone with ADHD, I expend a LOT of energy just to show up the way my neurotypical coworkers do. I started using a speech-to-text AI service about 6 months ago that transcribes my meetings and then allows me to use a little robot to analyze and craft content from it.

I have a hard time organizing my thoughts in a way that makes for a good article, so I decided to record a 10-minute voice memo of all the thoughts and ideas I had about this topic. I uploaded that memo to my AI service, then asked it to write an outline, identify themes (thanks Miri Rodriguez for that one!), and write some drafts, using my style and tone. I prompted it through a few versions then brought a version here to finalize. The entire process took me about 45 minutes. What do you think of the results?