Research suggests we should reset expectations on AI’s impact.
We have all been inundated with the constant reports that “AI is coming for your job” and that CxOs intend to downsize their workforce as AI automates tasks.
Whilst AI in all its forms, whether analytical (what I used to call “predictive”) or generative, is an amazing stream of technological innovation, I have always struggled with the idea that it will reduce jobs at scale.
The idea of “automating repetitive tasks” is not new, and it is not AI.
This is the role of “old school”, classic, deterministic, rules-based procedural logic in software.
Over decades we have, to an extent, created productivity improvements by eliminating manual tasks. We have done this through this “traditional” software in ERP and other enterprise applications and, of course, through RPA.
AI is a largely probabilistic approach and is thus tuned for aiding and augmenting analysis and decision making.
So, whilst there is much debate and even emotion around this topic, it was interesting to read a new Gartner survey suggesting that AI’s impact on workforce productivity is less transformative than hoped.
Randeep Rathindran, Distinguished Vice President in Gartner’s Finance practice, reported that “despite the excitement surrounding AI, its impact on productivity has been inconsistent, leading to what some describe as the AI productivity paradox.”
As the saying goes, “No Sh*t Sherlock!”
Gartner’s survey results reveal that just 37% of teams using traditional (analytical) AI and 34% using generative AI reported strong productivity gains.
Of course, that’s not bad, but equally could be a correlation not a causation! Since AI is unlikely to be the sole contributor.
They go on to say that, in essence, AI’s impact has so far been comparable to that of other emerging technologies, rather than an outsized game-changer.
But it is still early days.
Rathindran asserts that “CFOs should recalibrate expectations on how AI will truly impact worker productivity and headcount.” Rather than viewing AI as an immediate path to cost or labor savings, leaders should focus on the internal conditions required for AI to generate value.
Now we are getting somewhere!
This has always been the case. We need to;
- Eliminate bottlenecks
- Remove duplication
- Simplify processes
But this is far from easy.
The factors that truly drive productivity gains are in the essence of the “end to end processes” of our organizations.
This comes to the core principle that underpins all automation.
“The first rule of any technology used in a business is that automation applied to an efficient operation will magnify the efficiency.”
“The second is that automation applied to an inefficient operation will magnify the inefficiency.”
A Bill Gates quote that is burned into my memory.
An efficient operation requires us to think “Outside-In” and across an “end to end” process, before we get into the weeds or machinery (choose your metaphor 😉) of individual tasks.
Before we get immersed in the “HOW”, we have to keep asking the critical questions related to streamlining processes and to automation of all types . .
- WHY?
- What outcome are we trying to achieve?
- What does “GOOD” look like?
The goal is never automation. Automation is the mechanism. Whichever approach we take.
Once we have designed “an efficient operation”, we can apply all the digitization technologies at our disposal, including (but not exclusively) AI, for the relevant tasks – “Horses for Courses” . . .
Gartner’s guidance for CFOs is to challenge assumptions baked into AI-related business cases, particularly around short-term productivity gains or workforce reductions.
You can read the article “Why CFOs should reset their expectations on AI’s impact” here . . .
Research suggests we should reset expectations on AI’s impact.
Thanks for reading . . .
