Pascal’s Wager
Back in the roaring 1600's, Blaise Pascal, a french philosopher who wore a lot of ruffles and had way too much time to think, made a bold bet: believe in God, and you win big if He exists; if He doesn’t, no harm, no foul. Not believing however, puts you in major jeopardy (eternal damnation and no access to iced mochas) with little upside. This bit of inspired odds-making, now known as Pascal’s Wager, wasn’t about proving anything—it was about assessing risk. Pascal's genius was in admitting that he couldn't logically answer the question of God's existence and realizing that answering the question was actually less important than understanding its consequences.
Fast-forward a few hundred years, and Pascal’s thought experiment looks a lot like modern game theory—a way to make smart bets when you don’t have all the answers. It’s all about making decisions under uncertainty, whether you’re betting on the afterlife or the next big tech wave.
Making Decisions Under Uncertainty
Game theory, in simple terms, is about how rational players make moves in situations where the future is a giant question mark. Here’s the breakdown:
- Players: The ones making the calls—like companies figuring out their AI game plan.
- Strategies: The choices on the table—go all-in on AI now, or sit back and wait?
- Uncertainty: The big unknowns—will AI actually deliver game-changing results? Will your competitors outmaneuver you?
Businesses today are basically sitting at a poker table with AI. The cards haven’t all been dealt yet, but everyone has to decide whether to push their chips in. And to make things even more fun, how you invest matters—are you training your team in AI? Buying AI-powered products? Developing your own AI from scratch? Each path has different risks and rewards.
Deciding When to Invest in AI
The Challenge of Investing in AI Early
Jumping on the AI bandwagon now is a gamble. Doing it blindly is even riskier. The tech is still evolving, and while early adoption could put you ahead, it could also mean wasted money, buggy software, and frustrated teams.
But here’s the kicker: it’s not just about AI’s potential, it’s about what your competitors do. If they jump in first and figure it out before you do, they could leave you eating dust. Competition from well-heeled incumbents and fast-moving start-ups provides the existential threat of not investing in AI if it happens to become the transformational technology that many believe it will be.
The Prisoner’s Dilemma
In game theory, this is classic Prisoner’s Dilemma stuff. If a group of bank robbers gets arrested on suspicion, their smartest move collectively is to all stay quiet. But individually, they will be motivated to rat each other out in exchange for leniency. Acting collectively they have the potential to go free, but not knowing each others' plan makes the risk of staying silent too high.
In a perfect world, all companies would agree to chill out and wait until AI matures. But that’s not how business works. If one company invests while the others wait, the early mover gets the edge. And because everyone is afraid of missing out, they all dive in—even if it’s not the best collective choice.
The Dominant Strategy
A dominant strategy in game theory is the move that works best no matter what everyone else does, even if it's sub-optimal. And in the AI game:
- If competitors invest, you invest too—or risk falling behind.
- If competitors wait, you invest and grab the advantage before they wake up.
So, investing now wins either way. That’s why we’re seeing a gold rush into AI, even when no one’s 100% sure of the payoff. If AI turns out to be the transformational era defining technology that many people say, then you have potentially everything to gain or everything to lose. Companies that learn how to harness the technology early will leap ahead and extinguish those that fail to do the same. Think Netflix vs Blockbuster, Google vs AOL, Spotify vs record stores, Amazon vs bookstores, and Uber vs taxis.
The Pressure to Invest Despite Uncertainty
Even if AI’s magic isn’t fully unlocked yet, companies feel the heat because:
- First-Mover Advantages: Early adopters get the best data, the best talent, and a head start.
- FOMO is Real: No one wants to be the last company without an AI strategy.
- Investors Love AI: Even if it’s still a work in progress, markets reward innovation (or at least the appearance of it).
I'm not saying these are great reasons, but if you're leading a company or large team, you can't say that these pressures aren't real. So game theory tells us that AI investment is the right strategic move, but that only leaves us with an even trickier question: How do you invest?
De-Risking AI Investment
Getting in the game now doesn’t mean going all-in recklessly. My advice: Spend some real effort to identify high return-on-investment opportunities and start there. These are probably processes connected to the core of your business where automation can 10x throughput or cost-efficiency. I'm going to tell you right now that it's probably not a chat bot.
- Start Narrow: Do some diligence—pick projects with high-value potential and concentrate on building your confidence that you can actually pull them off rather than rushing to a deliverable. The key here is value—don't choose trivial projects.
- Stay Nimble: Be ready to pivot—double down on successes and ditch what’s not working. You will learn as you go and the areas where AI can deliver value will likely not be obvious at the outset.
- Go Native: Train your teams—AI isn’t just a tool, it’s a skillset your people need to master. It takes time for them to get a feel for the technology, so the earlier you start the higher the compounding value will be later on.
Your AI investment decision isn’t a one-and-done type thing—it’s an iterated game. We’ve seen a lot of big-tent projects fail from an overabundance of wishful thinking and a dearth of thoughtful analysis. You just need to start investing smartly so you’re positioned to scale when the time is right.
To build confidence, focus narrowly and keep investments manageable. At Machine & Partners, this is part of our philosophy. Recently, we wrote about how we design AI projects to deliver measurable ROI (spoiler: one big part of how we do it is by “going narrow”).
Around 30% of AI projects fail to get past the proof of concept stage. Big AI projects are almost certainly doomed to failure at this point in the AI cycle. The technology is developing so rapidly that you could spend a year on a big project only to discover that the technology you built it on has been superseded by new AI innovation. In my article “Moonshots are for rich kids”, I wrote about the danger of going big.
A strategic bet
AI investment is a strategic bet, not a coin flip. The cost of being too early is limited, but the cost of being too late could be catastrophic.
No one is saying you need to spend cash on AI just for the sake of it. But waiting for “the right moment” isn’t a strategy; it’s a slow road to irrelevance. Getting acquainted with AI doesn’t have to cost you the Earth but can profoundly and positively impact your business in ways unimaginable only a few years ago.
If you think now might be the time to take that bet, consider taking it with us. We’re privileged to be involved in incredible transformations with our customer partners. If you want to know more, contact me, and I’ll happily tell you all about it.
