Sponsored by Google Cloud
Choosing Your First Generative AI Use Cases
To get started with generative AI, first focus on areas that can improve human experiences with information.
AI is potentially the most democratizing piece of technology ever built but it is creating inequality
When we talk about AI, we often hear about its potential to level the playing field. In theory, everyone has access to these powerful tools. What I'm actually seeing – boots on the ground and meeting thousands of people each month – is the exact opposite. The most democratizing piece of technology ever built is creating more inequality than we've ever seen. Let me explain why.
There is a divide. And it is astronomical.
I recently gave a talk where I demonstrated just how dramatic this divide has become. We brainstormed a new AI program, ran it through Claude to create a brief, built a presentation, live-built an app to gather feedback, and then created a rollout plan. All in half an hour.
The audience was stunned. People stood up and said: "This is what I do in three months if I'm in charge of a program rollout." These aren't one-man bands or fledgling startups. These are technically skilled senior employees from massive multinational organizations and NGOs. Yes, in larger organizations, things already move slowly, but the fact that we condensed months of work into 30 minutes was mind-blowing to them.
While I and others who have adopted the AI-first mindset can do this, many can't. The divide ends up being astronomical. The amount of work that someone who uses these tools can actually achieve, compared to those who don't, is creating a chasm in productivity.
And this divide is already impacting the job market. I'm seeing it already. If we take software engineers today, as just one example, it's harder and harder for them to get a job with the skillset of yesterday (I’m looking squarely at coding here).
To understand this divide better, we need to look at how AI is changing the nature of work itself.
We need to shift our thinking from single-player to multiplayer productivity. In the past, we spent 95% of our time on solo tasks, occasionally getting feedback. Now, it should be the opposite: 95% of the time, you should be working with AI as a co-partner in whatever you're doing.
But incorporating this into your workflow takes time. It changes the way you approach every task. It's like having someone sitting next to you all day - at first, it's awkward, but over time, you develop habits where you're both looking at things in parallel. I'm now at the point where I'm writing a document while the AI is coding an application for a task I have to do tomorrow. Five minutes in, it pops up with a question, I quickly answer, and we both continue our work.
This is a new muscle we need to develop. Anyone who doesn't is going to be left behind dramatically. Again, that's where the real divide comes in.
Compounding is great. If you compound in the right direction. What's really scary is that this divide compounds over time. The combination of rethinking your workflows and new AI technologies coming out just keeps amplifying the results. Those who are using these tools effectively are constantly thinking about how to do things more efficiently. They're primed to take advantage of each new advancement.
On the flip side, non-adopters are heading towards what I call "AI poverty." I hate to say it, but I'm starting to become a bit pessimistic, and people who don't adapt may lose their jobs. At some point, companies are going to look at the productivity gap and make tough decisions. They might not understand exactly how AI works, but they'll know that one person is doing ten times more work than another. And that's going to lead to some harsh realities in the job market.
It doesn’t look good. You’re going to end up in a market where workforces are gradually shrinking. There will be firing, but not much hiring. The powers that be will begin to expect those that leverage AI to fill in the productivity gap of those they let go. This is already happening, and publicly. Just look at Duolingo’s employee cull at the start of this year.
But this isn't just about individual workers or companies. The AI divide is creating rifts on a much larger scale.
Want some more bad news? This divide isn't just between individuals, but also between regions and organizations. As mentioned before, large organizations always move slower, and that’s ubiquitous across all countries.
However, there is a marked difference in attitude towards cost-saving that's going to amplify this divide. You will have some cut-throat big U.S. organizations that are going to capitalize on this like crazy. They can move much faster than European organizations too, given the less stringent regulation around AI.
In the U.S., it's much more common to say: "Hey, we found a way to cut 50% of our costs, let's go and implement it." While in Europe, people are going to first think: "Wait, what is this going to mean for our people?" This difference in approach will likely lead to varying speeds of AI adoption and, consequently, different levels of disruption in different parts of the world.
The implications of this divide extend far beyond individual companies or even national economies. We're looking at the potential for widening global economic disparities on an unprecedented scale.
The irony is painful. We have the most democratizing piece of technology ever created, with the potential to level the playing field like never before. But right now, we're heading into the most divisive result we've ever seen.
We need to bridge this gap. We need to ensure that everyone, not just a select few, can harness the power of AI effectively. If we don't, we risk creating a world where the AI "haves" and "have-nots" are separated by an unbridgeable chasm. And that's a future none of us should want.
You May Also Like