You Probably Don’t Need the Newest and Shiniest Thing… and Shouldn’t be Distracted Chasing It
While everyone’s chasing tomorrow’s AI breakthroughs, there’s gold in the tools we already have. Here’s how to mine it effectively and start creating value today.
Heard any “game-changing” news about the AI future this week? If you’re not entirely sure… that’s probably the right answer – especially if the game you’re in doesn’t involve fabricating chips or building massive data centers.
But we’re not here with any of that today. Instead, we’re qchecking in with another of our semi-regular reminders to step back and assess the real-world, in-context application value of AI tools and capabilities that are already on the market, in evidence, reasonably well understood and documented, and potentially capable of changing — if not “the game” entirely — quite a bit about how you and your company play it.
Put simply: Don’t let the biggest, newest, shiniest thing that *might* be shimmering on the horizon distract you from all of the shiny and still largely underutilized things that have already piled up around us. There should be more than enough to keep you occupied, friends.
As always, it’s tremendously useful to be grounded in what types of things the tools (specifically, LLMs here) can/can’t and should/shouldn’t be able to do well. There’s plenty in each of those buckets. Getting grounding will help you restrict the scope of consideration to things that might actually be feasible and reliable and thus, worth your while, and it can help you resist the temptation to think we’ll be solving all of our problems and automating away our least favorite colleagues with just a liberal sprinkling of Gen AI fairy dust in a few months’ time.
From there, you can move to a set of practical questions that will help you identify likely domains of application within your organization, team, or personal workflows:
How might you evolve customer interactions through personalized, real-time support, predictive insights, or enhanced engagement across multiple channels?
How might you upgrade the employee experience by automating mundane tasks, enabling personalized learning and development, or fostering a higher-agency / more personalized work environment?
Where can you potentially streamline operations by automating routine tasks, optimizing processes, or enabling faster synthesizing of information and decision-making?
How might you accelerate the development of new products by enabling more rapid prototyping, personalized offerings, or enhanced research and development capabilities?
Where can you foster a competitive edge by learning faster, improving data capture, or enabling more data-driven decisions?
Don’t be surprised if you come up with a few ideas in each of these domains – particularly if you’re thinking through these questions collaboratively, which I’d recommend. And when you’ve identified some potentially actionable opportunities that should map reasonably well onto what today’s AI tools can already do, then you’re ready to design an experiment, choose some appropriate tools, and get to work.
As we’ve argued before, the AI-enabled future of your company may have already played out in a lab. Don’t waste time catching up to it, and don’t get distracted. Start where you are; use what you have; and do what you can. It’s almost certainly more than you think.
@Jeffrey