Years ago, I read the Harvard Business Review article titled The Simple Economics of Machine Intelligence and I wrote a Customer Success-oriented post called Where is The Future Going, Anyway?
(For your sake, read the HBR article)
The HBR article is dated 2016 — well before AI has become as mainstream in advertisement, daily adoption, and investing. However, the article is still predictive (bad pun, I know) for what and why we’re experiencing today and where we’re headed.
A bit of history: back in the 1980s and 1990s, data storage costs were immense. They inhibited predictive tools. Costs dropped precipitously in the decades since. But, Jeff, why does this matter? Because in order for predictive services, the underlying input (i.e., data) must be cost-effective. Once the inputs (data) become affordable, then prediction becomes a realistic opportunity — or, more precisely, affordable and doable.
The five components are: data, prediction, judgment, action, outcomes.
This leads us back to AI: Because we’ve had predictive AI and it has now become affordable, companies can deliver AI services to deliver judgments. Also, with the significant drop in costs, we will use predictive services more for our every day. But not only that, but we will create new ways to use these predictive technologies that we hadn’t even considered! An example is autonomous vehicles: the mass amount of data begat the ability for predictive tooling which begat the ability for a car to make a judgment on what to do. Does it stop? Speed up? Turn right? What happens in there is a previously unknown obstacle on the road? Not just autonomous cars for us to ride in, but packages to be delivered, semi-trucks delivering products, and much more. Those autonomous cars are taking action described above: slowing down and taking a detour when a road is closed and signaling to other cars to benefit other riders.
Finally, outcomes. Where AI has, in the past, been frustrating because it gives us collated, organized information but does nothing about it, we are entering the phase where not only does it give us that, but take action and communicates the outcome to us. “What is the ROI on that?” “Did that product deliver the value we wanted?” What if AI could help us resolve these very trick, analytically-heavy, and difficult-to-precisely-prove questions? What if it could not only tell us what to do, but then go do it?
Customer Success sidenote: Much of this next phase of AI for CS will be characterized by prediction and judgment (who is at risk and what do I do about it?). Few companies are yet to the stage of AI taking action and then providing outcomes. When we get there, the early successes for action will include: In-app AI guidances, in-app tool adoption/assistance, and agentic solutions for the user(s) to accomplish their goal in the product in a much faster manner. For the last item, think of version 1 as ChatGPT embedded into your application and not only providing information, but using your datasets to build out the structure, workflows, and automatically handling the tedious tasks you formerly did. Getting to action and outcomes will likely follow in the coming years, sadly (though I hope sooner).
Conclusion
Without intending to sound like the hypeman, the next phase of AI is a key driver of success for the economy and technological advancements: after experiencing strides in prediction and judgment, now we will move into actions and outcomes. These have dramatic implications for society, interpersonal relationships and conflict, team collaboration and management, how we measure ROI, and how we do business with one another. If Clorox, the bleach company, can adopt generative AI in their work streams, what could possibly hold up the rest of us?
One final comment: While we should be excited about this era of AI, we should not be “surprised” as if all this is suddenly, unexpectedly happening. AI is following the expected course of data and action — don’t believe me, go read some of Ajay’s early work. One of the biggest risks I see for our industry is not how quickly AI is changing our world and workflows, but how much of a surprise it is as we’ve had the resources and books to study it for years. We are reacting and not strategizing, we are unsure and that is holding us back from decision-making — the very things AI has promised help advance.
Jeff Beaumont 