The Harvard Business Review wrote an article a couple years back titled, The Simple Economics of Machine Intelligence. It’s a fascinating piece. Digest it slowly and over several cups of coffee/beer/wine/La Croix/whatever.
Changes afoot caused by machine intelligence:
Cost of goods and services that rely on prediction will continue to decline.
Cheap is good, right? Well, yes, but it also means it will substitute other things. For example, the American corn subsidy. Because of the super cheap price of corn, we now have it everywhere: cattle feed, high fructose corn syrup, ethanol, etc. (For a more extensive list: What is corn used for?). Effectively, when a good becomes cheap, we’ll find many uses for it — including outside of the norm.
Back to prediction. Where before we couldn’t do _______ because we didn’t have the capability or it was simply too expensive, we now have the data. This improves prediction. True, more data creates more problems, but it also gives us new opportunities and ways to interpret our surroundings. Cheaper data means more access to and lower prices for prediction.
Value changes: substitutes lose value and complementary goods gain value.
With Uber/Lyft, the value of taxis and other public transportation dropped, right? They haven’t been eliminated (yet?), but they are definitely struggling. However, complementary goods (Ubereats and breweries) will gain value as more people place greater reliance on Uber/Lyft (no one wants a DUI, after all). This, interestingly enough, greatly affects breweries, wineries, pubs, and other establishments. The arrival of cheaper transportation positively affects complementary items just like this. Think of what this means in your industry.
Cheaper data leads to easier access, affordable prediction, and higher reliance on prediction. That, in term, requires more judgment. Perhaps AI will someday will surpass humans in this area, though for now you (a human) can see everyday examples where systems with greater data provide better predictions so we can make better judgments.
For example, in Customer Success software we’ve aggregated data, so we can now build customer health scores. That, in turn, now requires human judgment to explore, assess, and act. This means our (human) judgment become much more valuable — and needed! — than before.
Another example is moving from on premise to cloud — data became accessible and significantly cheaper to track usage. This led to predictability. That, in turn, birthed new industries, roles (Chief Data Officer, Data Architects, and CS Ops, anyone?), and debates. This is an incredible opportune time for us.
From HBR, Here are the five in order:
For Customer Success, this could look like:
- Data: aggregate customer events, attributes, and data points
- Prediction: enhance our understanding of our customers through customer health/maturity/engagement scores
- Judgment: better our predictions will lead to know when, why, and how to help the customer
- Action: perform focused, purposed touch points that truly matter to the customer
- Outcomes: instill better decision making and actionable steps leads to better outcomes for our customers and for us
Do you see the causal relationships? Do you see how when data becomes cheap it has a cascading effect on many other areas. It starts as a trickle, upgrades to a stream, and eventually a roaring torrent. What does this mean for you? Where is your company, department, team, or role going? As automation eliminates data entry and AI takes on prediction, we humans have an insane growth potential for judgment. What does this do for you?
Yes, it could mean the elimination of your role, but do you really want to be doing that for the next 15, 20, 40 years?
Likely not. We can advance. We can move. Reflect, contemplate, and think on what it would look like if you worked less with data, less with prediction, and possibly even less with judgment. What would it mean for action and driving outcomes?