Am I separating what we KNOW from what we BELIEVE on GenAI?

Success with GenAI demands more than pilots and promises. It takes CEOs who challenge assumptions and move fast.

“I didn’t lie, I made an educated wish” – Deadpool from Deadpool & Wolverine

You just left a board meeting where your investors quoted (again) the Goldman Sachs projection that GenAI could raise global GDP by 7% or $7 trillion in the next decade. Your charge is to come back next quarter with a comprehensive plan of how GenAI can drive revenue and create efficiencies in your company. Because your early pilots, which haven’t made their way into production yet, clearly aren’t producing ROI as quickly as the board expected.

The real question is how your team responds. Are they believers or naysayers on GenAI as a pressing opportunity? What are their reasons why the early pilots failed or aren’t scalable yet? Many are probably thinking about whether priorities embedded in their annual performance reviews are aligned with new GenAI initiatives. Is it possible to deliver meaningful revenue within a one-year timeframe? After all, it is likely to affect their compensation. Others worry if they’ll have a job at all.

We deal with this all the time in business. The only constant is change and yet how prepared are we to make objective assessments and decisions on the best path forward?

We suffer from a litany of cognitive biases

A phrase I heard long ago and repeat often is “People don’t disagree with their own data”. Christopher Dwyer Ph.D. and Psychology Today contributor explains that even when we use research, we tend to promote data that supports our views and deprioritize facts which challenge our assumptions.

Dwyer states it well: These biases are systematic errors in our thinking.

Confirmation bias is when we seek out and are more receptive to sources that confirm beliefs we already have. It’s incumbent upon us to purposefully find data and facts that represent different viewpoints to understand the full picture. As leaders, we can be mindful of this common trend, also referred to as group think.

Onto other biases to watch out for.

Imagine this… you are sitting in a meeting where everyone is talking about the big sales miss from 4 quarters ago. The head of sales was fired last quarter.

You are seeking confidence that the sales projections will show the increase needed year over year (YoY).

The Chief Product Officer (CPO) pipes in that the customer complaints from a previous software release are resolved and customer sat won’t impact sales now.

The CFO weighs in that the main reason last year was off was the previous head of sales overstated the forecast.

The new head of sales adds that this week’s firing of two sales managers, that should’ve happened long ago, will turn around the performance of those sales teams.

The problem, as you and nearly everyone else in the room sees it, is that these are things that weren’t known a year ago. It came out in horrific fashion only 5 months ago when the CFO found out the now former head of sales was padding projections in the bottoms-up forecast. The negative customer feedback comes from a release 2 quarters back. While the tech issues were brewing and the internal teams were focused on it, it’s a big stretch to say the customers were squawking loud enough to impact sales 4 quarters ago. Oh, and those two sales managers were hitting their numbers until their boss was let go.

Welcome to hindsight bias.

This is one of my biggest pet peeves, and it happens a lot. Which just means I get plenty of opportunities to be annoyed. We have a strong tendency to explain past events with information known after the fact, as if we knew everything at the time. It’s a way to avoid admitting we didn’t see a thing coming and believe that we have things handled now.

Let’s layer on another component. In a business setting there’s a tendency to present what we think our leaders or colleagues want to hear versus being willing to go against the grain.

This can lead to either over- or underestimating significant trends and potential outcomes.

Like sales forecasts, the timing of product roadmaps, and big technological breakthroughs.

Missing the mark due to biased thinking has major implications

Let’s peek at blockchain, Blockbuster and Kodak for perspective.

“Everything will be tokenized and connected by a blockchain one day,” said Coinbase co-founder Fred Ehrsam.

While Ehrsam and Coinbase have done quite well, the use case for blockchain is much more limited than the initial hype portended. It does remain promising, but not the panacea for every industry. The financial services use case makes sense as well as other specific applications, but the ubiquitous application of blockchain across industries is unlikely.

"Neither RedBox nor Netflix are even on the radar screen in terms of competition," said Blockbuster CEO Jim Keyes (2008).

The infamous miss of Blockbuster shows what happens when emerging business models are dismissed, and leaders convince themselves they have a defensible moat around their business. Blockbuster filed for bankruptcy in 2010, and Netflix’s market cap as of July 2025 is $535B.

Kodak founder George Eastman made critical pivots that catapulted the company to success. Despite all the data and research they had in front of them, that insight did not endure with the 80’s and 90’s management team and board.

They destroyed one of America’s iconic brands by choosing to ignore the transformation of digital photography, filing for bankruptcy in 2012. In his book, former Kodak executive and head of market intelligence, Vince Barabba, shares the experience from the Kodak engineer who invented the digital camera.

Steve Sasson characterized the initial corporate response to his invention: “But it was filmless photography, so management’s reaction was, ‘that’s cute—but don’t tell anyone about it.’”

So…is your company getting it right on AI?

Following in the example of George Eastman, are you brave and nimble enough to pivot when it’s warranted? Companies need the CEO to be the main champion of bold moves to have a chance at wide adoption.

Is the company considering research and data objectively, and separating fact from opinion? Opinions are valuable and we expect our teams to offer them. The key is having the diligence to separate what is a fact from what is an opinion.

Are you specifically asking your team not to blindly agree with the consensus view and ask tough questions? Are people willing to play devil’s advocate in a constructive way, and is that important role appreciated in group conversations?

Time will tell if the Goldman Sachs forecast is on target, but experts like McKinsey encourage leaders to take a broad view and consider the multitude of use cases for GenAI.

For the next board meeting, you need to know what applications you can build and how quickly, given the current state of AI platforms. It’s important for boards to understand that much of the current investment in AI is building the platform infrastructure, with product applications coming quickly behind that.

Armed with this insight, you can better quantify a timeline and return without the pressure of the recent hype cycle. The good news is that a lot of companies are in this boat, and no doubt reforecasting the use cases, revenue and EBIDTA with AI.

The value creation of GenAI will hold beyond the hype and benefit those who sharpen their understanding and take action.