Douglas Flora, Executive Medical Director of Yung Family Cancer Center at St. Elizabeth Healthcare, President-Elect of the Association of Cancer Care Centers, and Editor in Chief of AI in Precision Oncology, shared a post on LinkedIn:
“Although no one can go back and make a brand new start, anyone can start from now and make a brand new ending.’— Carl Bard.
In the summer of 1921, Frederick Banting was failing at almost everything.
His medical practice in London, Ontario, couldn’t cover rent. He’d taken a part-time teaching job to survive. He had no research experience. No particular expertise in diabetes. No reason to believe he’d succeed where better-trained scientists had failed.
When he approached the renowned physiologist J.J.R. Macleod at the University of Toronto with his theory about treating diabetes, Macleod’s response was polite but dismissive. Fine, he said. Use the lab while I’m away this summer. We’ll see what happens.
What happened next defies every modern assumption about how breakthroughs occur.
Six months after his first crude experiment, Banting’s extract saved the life of a fourteen-year-old boy named Leonard Thompson—a child who’d been given weeks to live. Eighteen months after that sweltering Toronto summer, Banting stood in Stockholm accepting the Nobel Prize.
He didn’t do this through perfect preparation or years of careful study. He did it by starting immediately with whatever tools he could access, learning relentlessly from every failure, and iterating toward something that worked.
The lesson seems almost insultingly simple: stop preparing, start doing.
Yet it’s the lesson most of us will ignore this January, just as we’ve ignored it every January before.
The Ritual We Know By Heart
January 1st arrives. We wake transformed by possibility.
This year will be different, we tell ourselves. This is the year of change. Gym memberships are purchased with genuine conviction. Ambitious plans sketched in fresh notebooks. Resolutions declared with the fervor of true belief.
By mid-February, it’s over.
Parking lots are half-empty. Notebooks gathering dust. The resolution that felt so urgent six weeks ago now feels like someone else’s dream.
We console ourselves with familiar stories. Too busy. The timing wasn’t right. We needed more preparation—next year, when conditions are better.
But the autopsy of failed resolutions reveals a more straightforward, harsher truth: we confused planning with action. We mistook preparation for progress. We waited to feel ready instead of becoming prepared through practice.
The gym membership fails not from lack of desire. It fails because we set massive goals without building sustainable systems. We imagine transformation as a dramatic event rather than a gradual process.
Healthcare leaders approaching AI are making precisely the same mistake.
But the stakes here are considerably higher than an unused gym membership.
The Storm That’s Already Here
The numbers tell a story we can’t ignore anymore.
Over 50,000 layoffs in 2025 cited AI as a primary factor. Amazon cut 14,000 corporate roles. Microsoft eliminated 15,000 positions. Intel plans to slash 24,000 jobs by the end of 2026—reducing its workforce from 99,500 to 75,000.
In healthcare specifically, 28% of employees reported layoffs in their workplace this year. Among healthcare managers who made cuts, 18% directly attributed those decisions to AI advancement. CVS Health let go of 2,900 employees. Verily shut down entire medical device divisions to pivot toward AI-driven solutions.
The MIT numbers are even more sobering. AI can already perform 11.7% of all U.S. jobs, potentially replacing $1.2 trillion in wages across finance, healthcare, and professional services. By 2030, researchers estimate that 30% of current jobs could be fully automated, while 60% will see significant AI-driven transformation.
But here’s what should really get your attention.
The gap between ‘AI-capable’ and ‘AI-resistant’ workers is widening every single month. Entry-level positions in software development dropped from 43% of all postings to just 28%. In data analysis, from 35% to 22%. In consulting, from 41% to 26%.
Companies aren’t hiring fewer people overall. They’re just skipping workers who can’t demonstrate AI competence and hiring only those who already can.
The paradox is almost cruel: 96% of companies say AI skills are beneficial for candidates. 83% say demonstrating AI capabilities improves job security. Yet only 23% of organizations offered any AI training in 2025.
Translation: They want you to have these skills. But they’re not going to teach them. You’re expected to figure it out yourself.
The question is whether you actually will.
What Banting Had (Which Wasn’t Much)
Let’s be honest about what Frederick Banting brought to that Toronto laboratory in May 1921.
He had a borrowed workspace. An inexperienced graduate student assistant named Charles Best. Access to stray dogs. Basic surgical tools. A crude understanding of pancreatic anatomy. A theory that most experts thought was probably wrong.
What he didn’t have might matter more.
No research experience. No sophisticated equipment. No comprehensive knowledge of diabetes. No clear methodology. No respect from his peers. No certainty whatsoever that his approach would work.
By every reasonable standard, Frederick Banting was not ready.
He started anyway.
That Toronto summer was brutally hot. Banting and Best sweated through endless days of complex surgeries on dogs, extracting pancreatic tissue, creating crude preparations, injecting diabetic animals, and measuring blood sugar levels. Most experiments failed outright. Many results were inconsistent and couldn’t be replicated. The work was agonizingly slow.
When Macleod returned from Scotland in September and reviewed their summer’s work, his assessment was diplomatically devastating. The results lacked proper experimental controls. The methodology needed serious refinement. The conclusions were premature.
Any reasonable person would have recognized this as a failure and quit.
Banting kept going.
Not because he was sure he was right. His notebooks from this period reveal profound doubt. But he’d committed to the problem itself: children were dying from diabetes, and someone needed to do something about it. So he documented every failure, learned from each setback, and adjusted his approach.
On July 30th, 1921, something shifted.
A dog named Marjorie—dying from induced diabetes—received an injection of their crude pancreatic extract. Her blood sugar dropped. Her symptoms improved. She lived.
The preparation was still imperfect. The extract was still crude and contaminated. But it worked.
And that changed everything.
Six Months That Rewrote Medicine
Banting didn’t declare victory and stop experimenting. He brought in a biochemist named James Collip to help purify the extract. They tested it repeatedly. They failed. They learned. They improved the process. They moved from dog pancreases to cattle pancreases from slaughterhouses.
By January 1922—just six months after that first successful experiment on Marjorie—they were ready to try their preparation on a human being.
Leonard Thompson was fourteen years old and weighed sixty-five pounds. Diabetes had reduced him to a skeletal shadow. The doctors had told his parents to prepare for the inevitable. In the early twentieth century, there was no real treatment for Type 1 diabetes—only brutal starvation diets that might buy a few extra months of suffering before death.
On January 11th, 1922, Banting and Best gave Leonard his first injection.
It caused an allergic reaction. The preparation was still too crude, still too contaminated with other pancreatic substances.
They could have stopped there. Could have declared the experiment premature. Could have returned to the laboratory for another year of careful refinement before trying again.
Instead, Collip worked frantically for twelve days to further purify the extract, removing the impurities that had triggered Leonard’s reaction.
On January 23rd, they tried again.
This time, Leonard’s blood sugar dropped. His strength began to return. His wasted body started to recover. The boy who had been given weeks to live would survive another thirteen years—not cured, but given back his life.
Within months, insulin was in mass production. Within a year, it was available worldwide. Within eighteen months, Banting had won the Nobel Prize.
Six months from the first successful experiment to saving Leonard Thompson’s life.
Not six years. Six months.
Your Own Six-Month Window
It’s late December 2025. In six months, it will be summer 2026. In twelve months, we’ll be approaching 2027.
In that time, you could develop genuine competence using AI to solve real problems in your practice. Or you could still be preparing to begin, still reading articles, still attending webinars, still waiting to feel ready.
Organizations restructuring around AI aren’t waiting for people to feel ready. They’re identifying who can deliver results with new tools and who can’t. That gap widens every month, and it’s not widening based on who attended the most conferences. It’s based on who has actual hands-on experience solving real problems.
This is your inflection point.
Not to master AI comprehensively—that’s gym membership thinking again, setting impossible goals that guarantee failure. But to become demonstrably more valuable by developing practical AI competence in your specific domain.
While headlines scream about AI destroying jobs, there’s a quieter truth underneath: 91% of companies using or planning to use AI will hire new employees in the coming year. They’re desperate for people who can actually work with these tools. Yet they’re simultaneously eliminating entry-level positions and refusing to provide training.
The opportunity is massive for anyone willing to close the gap themselves. To invest in their own capabilities when their organizations won’t. To become the person desperately needed but inadequately trained.
A Resolution Built For How Change Actually Happens
Forget massive transformation. Forget perfect plans.
Here’s what Banting actually did: he used the tools available to him immediately and learned by solving real problems. You can do the same thing starting today.
Week 1: Make a Twenty-Dollar Investment in Yourself
Stop waiting for your organization to provide AI training. They’ve already told you they’re not going to—remember, only 23% offered any training last year.
Subscribe to ChatGPT Plus, Claude Pro, or Gemini Advanced. Pick one. They’re all excellent, and the specific choice matters far less than actually starting. Twenty dollars per month is less than most gym memberships, and unlike that gym membership, you’ll actually use this.
This is your Frederick Banting moment. He didn’t wait for someone to give him a state-of-the-art laboratory. He used what was available. You now have access to tools more powerful than what most Fortune 500 companies had three years ago, for the cost of two coffees.
Week 2: Build Your Personal AI Partner
Here’s where it gets interesting. These AI tools can become your personal executive coach, your McKinsey consultant, your data analyst—whatever you need them to be.
Start with a conversation that establishes context:
‘I’m a [your role] with [X years] experience in [your field]. My background includes [key experiences]. Over the next 12 weeks, please help me develop practical AI skills that will make me indispensable in my organization. My biggest professional challenges right now are [list 2-3 specific problems]. I learn best by [your learning style]. Can you create a personalized 12-week learning plan for me?’
Be specific. The AI can’t read your mind, but it can build remarkably well on what you tell it.
Think of this as your intake session with a personal trainer—except this trainer costs $20 a month instead of $200 an hour and is available whenever you need it.
Weeks 3-12: Your Practical Apprenticeship
Now you stop learning about AI and start using AI to solve real problems. Each week, tackle one practical skill that directly impacts your work.
Week 3: Create presentations that don’t waste your life.
Ask your AI partner: ‘I need to create a compelling PowerPoint presentation about [your topic] for senior leadership. Please show me how to structure a consultant-quality deck, then help me build it in the next 10 minutes. Use the pyramid principle and make it executive-ready.’
Watch how it thinks. Ask it to explain its structural choices. Have it critique your drafts. By week’s end, you’ll be creating in 10 minutes what used to consume hours of your time.
Week 4: Analyze numbers like someone who actually understands them.
‘Here’s our department’s P&L [paste sanitized numbers]. Please teach me how a CFO would analyze this. What questions should I be asking? What patterns should I look for? What insights am I missing?’
You’re not just getting answers. You’re learning how experts think.
Week 5: Write business cases that actually persuade.
‘I need to write a business case for [your initiative]. Please show me how to structure it, what data to include, and how to make the ROI argument compelling. Then help me write it.’
Week 6: Design your career like you’re advising someone you care about.
‘Act as my executive coach. Based on what you know about my background and goals, help me create a three-year career development plan. What skills should I develop? What experiences should I seek? What blind spots might I have?’
Week 7: Make sense of data that’s been sitting there taunting you.
‘I have this dataset [describe your data]. Please teach me how to analyze it for meaningful insights. What visualizations would tell the story best? What patterns should I look for?’
Week 8: Have difficult conversations without losing sleep the night before.
‘I need to deliver challenging feedback to [role]. Help me structure this conversation. What should I say? How should I handle potential objections? Role-play this with me.’
Week 9: Think strategically instead of just reacting.
‘Our industry is facing [challenge]. Help me think through strategic options. What frameworks should I use? What questions should I be asking? What am I not considering?’
Week 10: Fix workflows that have been broken for years.
‘Walk me through how to analyze this process [describe workflow]. Where are the bottlenecks? What could be automated or eliminated? How would a process improvement consultant approach this?’
Week 11: Become conversant on topics faster than you thought possible.
‘I need to become knowledgeable about [topic] quickly. Create a learning plan for me. What are the key concepts? Who are the essential voices? What should I read first?’
Week 12: Build a system so you’re not starting from scratch next month.
‘Help me create a system for capturing and organizing what I’m learning. How should I structure my notes? What’s worth saving? How do I make this knowledge easily accessible later?’
The pattern is straightforward: Every week, bring a real problem from your actual work. Don’t use hypotheticals. Use real projects, real challenges, real documents—sanitized appropriately for confidentiality. Let the AI teach you while helping you solve genuine problems.
A Critical Note for Healthcare Professionals
If you work in healthcare, you need to understand something important right now: these frontier AI tools—ChatGPT, Claude, Gemini—are not yet HIPAA compliant.
Please don’t put patient information into them. Do not upload clinical data. Do not share anything that could identify a patient or violate privacy regulations. This is not negotiable.
But here’s what you absolutely can do: Use them for everything else.
Analyze de-identified, sanitized data: draft policies and protocols. Create educational materials for staff. Build presentations for leadership. Write grant proposals—design quality improvement initiatives. Develop communication strategies for difficult conversations. Plan your career trajectory. Structure stakeholder meetings. Optimize administrative workflows.
Think about all the work that consumes your time and steals hours from patient care—that’s where these tools shine right now. Use them to reclaim time. Use them to think more clearly. Use them to communicate more effectively. Use them to become more strategic.
As these tools become HIPAA compliant—and they will, the business case is too compelling—you’ll already know how to use them. You’ll be ready to apply these capabilities to clinical work the moment regulations catch up. But you’ll have spent twelve weeks solving real problems and building practical competence with everything that doesn’t require regulatory compliance.
Start where you can start. Learn what you can learn. Build the capabilities that will matter when the regulatory landscape catches up with the technology.
What You’ll Actually Have After Twelve Weeks
Not comprehensive AI mastery. That’s not the goal and never was.
Something more valuable: twelve weeks of solving real problems with AI while building capabilities that distinguish you from every colleague who’s still ‘preparing to learn about AI someday.’
Think about what you’ll have accomplished.
Presentations created in minutes that used to take hours. Financial data analyzed with insight you didn’t know you could access. Business cases drafted at a level that would have cost thousands from consultants. A career development plan is one that most people never create. Workflows that have frustrated you for years. Difficult conversations navigated with new confidence.
But more importantly, you’ll have developed something irreplaceable: practical fluency.
You don’t just know about AI. You know how to use it to solve actual problems. You’ve built muscle memory. You’ve learned to prompt effectively, iterate quickly, and extract value efficiently. You’ve developed judgment about what works and what doesn’t, when to use AI and when not to.
When your organization asks, ‘Who here can actually work with AI?’ you won’t be hoping someone else raises their hand. You’ll be the person with twelve weeks of daily practice, real results, and concrete examples.
That twenty-dollar monthly investment? That’s roughly $240 over twelve weeks.
Compare that to a McKinsey consultant at $500- $1,000 per hour. An executive coach at $300-$500 per hour. A business school course costs $3,000-$10,000. An online certification program costs $500-$2,000.
You’ve essentially built your own personalized training program for the cost of a nice dinner.
And unlike those expensive alternatives, this isn’t theoretical knowledge you’ll struggle to apply. It’s practical capability developed through solving your actual problems, refined through daily use, and proven through tangible results.
The Gift That Compounds
When the University of Toronto applied for patents on insulin and the methods for producing it, Banting, Best, and Collip sold their rights to the university for one dollar each.
One dollar.
For a discovery that would eventually generate billions in revenue and save hundreds of millions of lives across the next century.
Banting’s explanation was characteristically direct: ‘Insulin does not belong to me, it belongs to the world.’
He understood something we sometimes forget in our age of intellectual property and competitive advantage: the greatest value isn’t in what you can hoard but in what you can become. The capabilities you build belong to you in a way that no organization can take away.
Organizations can restructure. Jobs can be eliminated. Markets can shift. Entire industries can transform overnight. But the knowledge that lives in your mind, the skills that become second nature, the judgment you develop through practice—those travel with you wherever you go. They compound. They open doors. They create opportunities.
This isn’t about getting rich from your AI skills, though you might. It’s about making yourself valuable, relevant, and indispensable in an economy that’s rewriting every job description. Investing six months could determine the next six years of your career, about becoming the kind of professional who doesn’t wait for permission or perfect conditions.
Who invests in themselves when their organizations won’t? Who starts before they’re ready and becomes ready through practice. Who closes the gap themselves instead of waiting for someone else to bridge it.
The Resolution That Actually Sticks
The gym membership will fail by February. It always does.
Massive goals, elaborate plans, perfect conditions that never arrive—then quiet abandonment when reality intrudes. The pattern is so reliable we could set our watches by it.
But this resolution is built differently.
Twenty dollars a month for twelve weeks. One fundamental problem per week. Fifteen minutes of daily practice. Specific enough to be actionable. Modest enough to be sustainable. Grounded in solving actual problems rather than abstract learning. Built around tools you can access today, not capabilities you need to develop first.
Most importantly, it creates demonstrable value within the six-month window that might determine whether you become indispensable or expendable in the AI era.
Frederick Banting wasn’t ready to discover insulin. Leonard Thompson didn’t have time to wait for him to feel prepared. So Banting started anyway, with crude tools, imperfect knowledge, and relentless determination.
He gave himself six months to solve an impossible problem. He didn’t solve it perfectly—that first injection on January 11th failed. But he learned from that failure, refined his approach, and twelve days later tried again. And that time, it worked well enough to save a dying boy’s life.
Then he kept iterating until that crude solution became one of medicine’s most significant breakthroughs.
You have six months until summer 2026. Twelve months until we’re having this same conversation again next January.
The question is what you’ll have to show for that time.
Will you still be preparing? Would you still be reading articles? Would you still be attending webinars? Still waiting for your organization to provide training that isn’t coming? Still telling yourself you’ll start next month, next quarter, next year?
Or will you be the person who started before they were ready and became ready through practice?
The person who invested twenty dollars a month when it mattered most. Who brought real problems every week? Who learned by doing instead of reading about doing. Who built capabilities that can’t be replaced, outsourced, or eliminated? Who closed the gap themselves when no one else would.
The person who looked at the six-month window between now and summer 2026 and decided to fill it with action instead of preparation.
Stop preparing for AI. Start using it.
This is your resolution for 2026.
Start this week. Start with one problem. Start with twenty dollars. Start before you’re ready.
Your six months begin now.”
More posts featuring Douglas Flora.