Insights

I get this question a lot. Admittedly, I don't love the most obvious answer, which is, "it depends…" No matter how true that response is, nobody likes it.

So the idea of figuring out where and how AI should fit into our lives can feel like going in circles. On top of all that, there's so much AI-related content in our feeds, emails, meetings, and news, it can be overwhelming. Many feel paralyzed and on the outside looking in.

A few ways I personally frame my thoughts on AI:

• Data + Algorithms: AI is only as good as its data and algorithms. Top-tier algorithms with poor data will still lead to poor results.
• Own Your Data: Whether personal or professional, don't just give data away. Whoever controls it will build the most valuable solutions.
• Learn with a Purpose: When taking AI-related courses, pick ones that match your industry or career goals. Learning AI without context is frustrating.
• Recognize the Hype: Vendors will push their solutions hard, but not all are ready for daily use. Test carefully and adopt new tools in small steps.

I have been relatively quiet online for a couple of weeks, trying to learn the ins and outs of agents and where I want to play a role in creation of new opportunities, not just in the reduction of old...

I had the good fortune of coming up for air and spending time with friends. We debated the pros and cons of local AI agents vs. diving all in with AWS or GCP. The conversation was informative, the food delicious, but what I will really remember was the gentle reminder to fight the temptation to do it all alone.

The way AI solutions are being rolled out can seduce you into a solitary existence. Invite others to be a part of your conversations, inventions, and everything in between that you do with AI. There is still value in the shared journey, even when it is more "efficient" with this new tech.

Much of the power of A.I. for you and I is in our ability to build stand-alone applications that help us realize all the wild ideas we've been unable to get started for whatever reason. The Large Language Models get all the headlines, but your ability to create standalone solutions is potentially way more powerful and meaningful.

A friend shared a challenge: "Notes taking over my life." Taking notes while having meaningful conversations with patients, transcribing, evaluating, entering EMRs, navigating multiple languages, scheduling followups. The time being spent after work and weekends was taking a toll.

In less than 15 minutes, I put her problem statement into an LLM to create a requirements document, then took that document and put it into Replit, and boom—a whole application was built.

To my entrepreneurs: don't just rinse and repeat generic apps, but think about creating truly fit-for-purpose solutions quickly. We have access to tools we thought unimaginable just a few years ago.

As a mathematician, I view history through the lens of patterns—like the ebb and flow of a sine wave, where extremes eventually balance out. This perspective always offered comfort, suggesting a fundamental equilibrium in our collective journey.

But now, I see something that challenges the very "magic of the human function."

As AI tools rapidly integrate into every facet of life, we are inadvertently creating intellectual asymptotes. We're being presented with "answers" or streamlined paths by algorithms, and in doing so, we risk underestimating the vast, often unpredictable variability that has historically led to our greatest breakthroughs.

How many brilliant discoveries were accidental? Penicillin was a fortunate oversight. Post-it Notes arose from a failed attempt at a stronger adhesive.

While I deeply appreciate AI's efficiency, I advocate for intentionally embracing the "hard way" from time to time. Seek intellectual collaboration with fellow air-breathers, not just LLMs. There's a richness of learning and human connection that cannot be algorithmically quantified.