AI in HAIL MARY
Forget "AI Best Practices". Try Desperation Instead.
You know that one person who refuses to use GPS until they’re hopelessly lost? The one who’ll drive in circles for an hour before finally muttering “fine” and turning on Google Maps?
I used to laugh at them (I am the opposite, the one who says “navigate to work” to my car every morning)... until I realised we’re all that person when it comes to AI.
This hit home recently while researching how businesses use Gen AI in crisis situations. My colleagues (Graham Kenny and Kim Oosthuizen) and I published our findings in Harvard Business Review. While the HBR article focused on how GenAI is levelling the playing field for smaller businesses, we discovered something unexpected: a pattern of successful Gen AI use that looked suspiciously like that friend who only turns on GPS after getting thoroughly lost. Instead of the carefully planned AI implementations we expected to find, complete with change management strategies and detailed rollout plans, we found something far more interesting. And far more human.
What we found looked less like a strategic playbook and more like a series of Hail Mary passes: those desperate, last-second throws when all other plays have failed. And unlike in football, where Hail Marys rarely connect, these desperate AI plays were scoring touchdowns.
The “Fine, I’ll Ask AI” Club
Lucy, one of the executives we wrote about1, ran a boutique winery in Australia, exporting all her wines to China. Then, overnight, China slapped a 200% tariff on Australian wines. Her market vanished. Traditional strategies failed. Bank accounts drained. And only then – when she had literally nothing left to lose – did she turn to ChatGPT.
Spoiler alert: it worked. But that’s not even the most interesting part of the story.
The Desperation Advantage
The pattern kept showing up. There was Luke, who ran a truck customisation business. His supplier casually dropped a 30% price increase bomb via email. Only after cycling through denial, anger, and an impressive collection of creative swear words did he turn to AI for help.
Then there was Julia, leading a marketing agency where everyone was so swamped with daily fires that strategic planning felt as realistic as a unicorn riding a hoverboard. Her team turned to AI not because it was trendy but because they were desperate for a way out of their mental maze.
"Sometimes, rock bottom is the most solid foundation to build on." - Every business guru ever (probably)
Why Last-Resort AI Often Wins
Here’s what blew my mind: these last-resort AI adoptions often worked better than carefully planned implementations. It’s counterintuitive, but think about it:
No Analysis Paralysis: When you’re desperate, you don’t spend six months evaluating every AI tool on the market. You just try something.
Reality > Best Practices: These folks weren’t following any AI implementation playbook. They were solving real problems with whatever tools they could grab.
Nothing to Lose: Fear of failure? That ship had already sailed. They were free to experiment radically.
The Accidental AI Experts
Every business success story in our research started as a Hail Mary - that moment when you've run out of conventional options and throw the ball as far as you can, praying something good happens. And somehow, Gen AI was catching these desperate throws.
Here’s what happened to our desperate pioneers:
Lucy didn’t just save her wine business—she ended up creating her own custom GPT to help track budgets and expenses. From “What’s a prompt?” to automating her finance function in months, Lucy freed up time to do what she loves most: crafting wines.
Luke took his supplier crisis to ChatGPT and got six solid strategic options in seconds, from negotiating with suppliers to completely rethinking his customer relationships. He admitted that he couldn’t have generated such a comprehensive list on his own—certainly not that quickly.
Julia’s team went from being stuck in daily operational weeds to mapping out vivid future scenarios with Gen AI’s help. They even spotted a critical gap in their IT infrastructure that would have crippled them in a chaotic future.
When you're drowning, you don't take swimming lessons. You just swim. Turns out, that works for AI too.
The Reverse Psychology of Technology Adoption
Maybe we need to flip our thinking about AI adoption.
Instead of the traditional path:
Attend endless AI readiness workshops
Create detailed AI implementation roadmaps
Form an AI steering committee
Finally, let someone try AI in 2026
How about this:
Hit a wall with traditional approaches
Realise you have nothing to lose
Try AI with a specific problem in mind
Actually solve real business problems
Look at our cases: a winery reinvented by AI when tariffs killed its market, a truck company finding strategic options when prices skyrocketed, and a marketing agency breaking free from operational paralysis. None of them started with an AI strategy. They started with a problem that needed solving.
The “Just Do It” Paradox
Look, I’m not suggesting you should wait for a crisis to try AI. That would be like waiting to learn swimming until you’re already in the deep end. (Though, based on our research, you might turn out surprisingly good at it.)
But there’s something powerful about stripping away all the AI anxiety, implementation frameworks, and “best practices”.
Sometimes, you just need to jump in.
A Final Thought
The irony isn’t lost on me (it’s the second post in a row when I write these words). As someone who spends his days teaching organisations how to implement AI “properly,” I’ve discovered that some of the best success stories come from people who have never attended a single AI workshop or read a “Top 10 Tips for AI Implementation” article.
Maybe the best AI strategy is to stop overthinking it. After all, no one ever changed the world by following a best practices guide.
And if you’re sitting there thinking, “I’ll wait until I really need AI”... well, just remember your friend with the GPS. Sometimes, the longest route is the one where you pretend you don’t need help.
Now, if you’ll excuse me, I need to tell my car to navigate to work.
We anonymised our cases, changing industries and names but maintaining the context and challenges.
Excellent post, Marek! It resonates deeply with our community of executives at Amcham. To explain the concept of a Hail Mary here in Brazil, I might say it's like pulling all the defenders off the field and going all-in on attack in the final minutes of a match—a desperate move that sometimes creates a miracle. The idea that innovation often arises from the 'nothing to lose' mindset is something many executives here can relate to. Congrats on the clarity and the insights!