By Benjamin Gilad
You bet. And for the entire forseeable future.
Here is my Thanks Giving present to my competition analysts out there.
AI and strategy- a comedy of errors
An HBR article, “How CEO use AI” provided two examples of companies using AI for strategic planning- one a commercial agricultural research organization and one a government-owned cemetery admin in charge of one cemetery of 145 acres. I didn’t know if I should laugh or cry reading that.
These are the businesses that require AI? What next, a lemonade stand?
What the HBR examples showed, moreover, was exactly what the authors probably didn’t want to show: Gen AI is badly hyped.
What the examples showed, moreover, was exactly what the authors probably didn’t want to show: Gen AI is so badly hyped it comes close to being a joke. Anyone joining the joke is trying to sell something or is an academic who never ran a business.
In asking “what strategic issues may impact the business”, the Agri-research executive team forgot to include four areas that Gen AI brought up: “technological advancements,” “regulatory changes,” “client demand and expectations,” and “funding and investment.”
Oh, well. Who would have thought these were important? (Only each and every CIP™.) But let’s move on. The areas Gen AI forgot to flag were puny little things like the basic elements of competition (price, branch locations, and ultimate profitability).
AI “forgot” to flag the basic elements of competing
So far, CI 1, GenAI 0.
Worse: The authors admit that “while Gen AI is a sophisticated technology, it cannot predict the future. This is because gen AI tools are loaded and trained on historical data.”
Bingo. So how good is this for strategic thinking? Should strategy be based on history alone?
If you use AI to guide your strategy, your strategy will forever be looking backward..
The saga with the cemetery admin was even worse. When Gen AI was asked to be a strategic consultant looking at the list of future trends which weren’t considered by the exec team, it came up with “aging population”.
I was floored. Such an insight. Who’d expected that to affect cemetery operations??
On the other hand, in offering alternative solutions to scarcity of land for burials, AI suggested sending cremated remains into deep space.
Oh, well. Someone should be sent to outer space but it may be the AI consultants..
CI 2, AI 0.
AI and the misguided campaign against “biases”
Among the most vocal advocates of Gen AI’s role in strategic decisions are those claiming it will reduce biases.
When I was a young naïve assistant professor of economics at Rutgers University in the early 80s, I came across a little know journal called Journal of Behavioral Economics out of some local university in Illinois (I think). It was edited by the great Richard Hattwick, who foresaw the (then) nonexistent field as the future of economics.
It changed the course of my career from a mediocre researcher into a mediocre researcher with tenure. I delved into behavioral economics headfirst. The pool that was empty until then started filling up quickly later on with the works of Kahneman and Tversky and their students. Notably, this breakthrough body of work described a few prominent cognitive biases used extensively in making wrong judgements (in experimental setting- one of the major criticisms of the theory).
Alas, like every system perpetuating itself, academic tenure decisions which are based on the number of publications encouraged every PhD candidate in the next decade to discover more biases. The number of biases then started to exceed the number of available tenured positions. But the damage has been done. To sum it up in a simple, short, succinct sentence:
We are biased. Get over it.
The great reveal: We are all biased. Get over it.
The “get over it” is a problem though since those telling us to get over it are biased as hell themselves. Worse, since there is no real way to rid yourself of biases, they are selling you shaman oil. Nowadays, this oil is called AI. It’s the same oil just way more expensive.
Since there is no real way to get rid of biases, those selling AI as a cure are selling you snake oil. It’s the same as a shaman’s cure, only way more expensive. Did libraries make us less biased?
Never mind that AI is biased itself. Never mind that its data come from biased humans. The real, deep, insurmountable problem is that biases are good. Yes, good. Actually, essential to progress.
Here is my undeniably most provocative statement ever: In business, biases are essential to progress.
I know, this is sacrilege.
AI is a risk-bearing “gift”
@Arthur Weiss, a wise CI practitioner and trainer for decades, with practical, real business world experience said in a recent post on LinkedIn “Depending on AI will result in mediocrity, which for many is probably better than they’d achieve anyway.”
How true.
He adds, “It’s a tool that can make a hard job easier / analysis faster but it won’t give true competitive advantage as it won’t have that element of genius that defines winners.”
Just like automation made backoffice operations more efficient, but it didn’t produce winners, did it?
AI makes routine work easier, no more.
Anyone involved in entrepreneurship or even just studying entrepreneurship knows it is exactly the bias of entrepreneurs (and by implication, CEOs) that underlies the skill of competing. If you think like everyone else, use AI to generate “objective data” which everyone else agrees on, you are not being “reality-based”. You are being a fool. It is exactly the ability to see things subjectively and differently that underlies each and every great CI analysis and each and every strategic vision that generates competitive advantage.
You can use AI to tell you what everyone else is thinking; You should use CI to tell you what everyone else is missing.
And be real: To know what “everyone” else thinks you don’t need AI. Just look at your market, talk to a few people (my rule of thumb- less than 5), do a quick 5-force analysis, and it is quickly clear that it’s not really “everyone” that counts.
The final straw in AI’s data-driven bollocks
Amazon Web Services CEO Matt Garman’s claimed in an internal meeting a few weeks ago that nine out of 10 Amazon employees told him they were “excited” about the new mandate of back to the office 5 days.
What followed was a letter signed by 500 employees saying, “Amazon claims to be a data-driven company, and statements which rely on ‘we believe’ and ‘I feel,’ or exclaim how many people are excited about something to justify themselves fail to achieve our standard of decision making for critical issues that impact hundreds of thousands of the company’s employees and its customers.”
We can look at this in two ways.
1. Data-drivel by Mr. Garman who only “surveyed” people who would be afraid to contradict him (sample bias, similar to how a pollster “discovered” Iowans will vote for Harris), or,
2. The reality of how important decisions are made. They are based on management assumptions about the future, and the future is immune to data. Is it bias? Maybe. Is it a good decision? Probably, time will tell. Can AI help?
In your dreams…
In short, do NOT listen to those saying AI will help your strategy tools be better. It will not simply becasue there is no “better” in mediocrity. It’s your subjective perspective that is important and that has at least a chance of impacting management. Management already knows, or don’t care about, the mediocre, routine “competitor news.”
Don’t believe me? Think “sophistication” is the name of the comepting game, not “biased” intuition? Do yourself a favor and read this article (reference provided by @Glen Lyons. Thanks, Glen!)
Alternative perspective
CI is the bedrock of strategy but only if the CI analyst understands Mr. Weiss’ point: AI can make the routine tasks easier, but it doesn’t do competitive intelligence.
If you ask AI, are you biased? I wonder what it will say.
Develop your skill of unique insights- our next intensive therapy weaning you of corproate mediocrity is in March/April. Register here. Registration – Enroll Now – Academy of Competitive Intelligence