The Case for Discernment in AI

Completely, 100% written by a human.

Short version: Discernment in AI is a critical skill to be practiced on 3 levels: personally, professionally, and as an organization. Jump ahead to:
– The Case for Discernment in AI Personally
– The Case for Discernment in AI Professionally
– The Case for Discernment in AI Organizationally

As I sit here writing this, it feels like the entire world is sprinting towards AI. Individuals and organizations of every size seem sold on this promise, that artificial intelligence is going to drastically change the ways we work and live.

And to some extent that’s real. For better, and for worse.

But regardless of where you sit on the zoomer/boomer(?) -to-doomer scale, the AI “revolution” is here. And while I don’t see a world where we all come out of this unscathed, I do believe there’s one, specific attribute that’s guaranteed to separate the technology victors from the rest.

Well, aside from access to resources & capital I mean.

That thing? Discernment. Our ability to perceive, judge and distinguish the role AI plays in this world is the boldest, most impactful skill we can all possibly exercise right now.

Hang on. Discernment sounds like a basic, archaic skill.

As it should. Discernment, on a basic human level, is part of our survival toolkit. It’s just slower, patient, and requires more reflection and processing than true human instinct.

And it’s worth calling out that discernment involves making sense of the information you have. But in today’s world, we’re doing this with *way more information* than we’ve ever had before. We can thank the internet and all of its adjacent innovations – like social media – for that one.

So in that sense, discernment is not new. The way we’ve had to call on it has evolved, as the proliferation of certain technologies have forced us to stretch the muscle. But now AI is here, and the calculus is very different.

The Case for Discernment with AI Personally

We’ve already seen extreme cases of this play out on the main stage. Think those who have harmed themselves, or others, at the implicit (and sometimes explicit) guidance of their AI chats.

Or think of the voices who have come out publicly to share their emotional, sometimes romantic involvements with that same underlying tech. Who else was ready for that in 2026?

But there are less extreme, less visible instances of everyday people using AI as a tool to aid their own discernment. And while that’s not dangerous on its own, it can be detrimental when people not versed in technological understanding of AI fail to recognize that:

1) AI tools may be intellectually smarter than us, but they are not discerning enough.

Even if you were to hand ChatGPT all the background info about that conflict you’re having with your partner, or the issues you’re having at work: it will never have the context that you do.

It won’t know your current motives and fears, how those have changed over time, your history or upbringing – all examples of things that you as a human bring to your table.

2) AI “brains“ have a crazy entourage right now, and our brains are painfully human.

Our hunter-gatherer ancestors had very different survival problems back in the day, but arguably more clarity on what to solve for.

We have far more stimulation and information overload, even on just an hourly basis, that our brains try to contend with. Think social media ads, commercials, billboards, all your app notifications, etc.

And decades of psychology research has proven that our brains carry very human limits. To varying degrees, we are all susceptible to influence, peer pressure, FOMO, advertising, addiction, validation, brain rot and the like. We just don’t like to think we are.

On the opposite end, AI tools are touching our lives in personal ways – but with the backing of trillions of data points that no single person can possibly have. And when you consider how much record-breaking money is flowing into these tools to make them smarter (and more profitable), what you’re left with is this.

AI is super smart and rich. A powerful rival in the discernment camp.

3) Like social media, AI tools are built to keep you coming back. There’s a reason why pretty much all of your chat outputs end with a follow-up question. Much like social media, our usage of these tools determines success for these companies. On its face, this makes sense. But if if you find yourself feeling starting to feel even slightly addicted, know that it’s more than a simple willpower issue.

The Case for AI Discernment Professionally

At the enterprise/org level, there’s a lot I can say on this subject.

But this is about individual, professional use of AI. And I must say, depending on what your tech policies are & what you need to accomplish in your day-to-day, there’s real space for AI to make waves.

How workers discern where AI makes them productive and effective is essential. If you can make ethical, compliant use of AI to aid your individual work, and save yourself time or stress (without seeking to actively displace humans or overwork yourself), that’s worth it. High-fives.

Especially for one-person teams or first time solo-preneurs trying to get off the ground. You can see how an AI strategy, even if not robust, could produce real impacts in helping lean functions work smarter/stop working harder.

But it’s also true that adoption without discernment puts us at risk of working harder and “dumber”. And sometimes, without our realizing it.

1) On the more mundane end, consider workers who spend more time checking AI outputs for accuracy – or re-prompting – than they would if they’d performed the task themselves.

2) On the more controversial end – consider the increase of non-technical pros “vibecoding” whole websites and platforms with AI. Sure they look stellar, but many lack the knowledge to verify that their projects actually use the best coding and security frameworks available.

That last point sounds dry, I know. But it presents real risk to the rest of us. And to quote whoever said it first, you don’t know what you don’t know.

AI is positioned to bring new tech capability to non-technical professionals. This is something to celebrate. And, it is also the exact reason why discernment matters.

The Case for Discernment with AI Organizationally

Fun fact, I intended for today’s post to start here. This one really got away from me.

AI applications need to look different depending on the organization: whether you’re in it for the social good or the money, whether you’re a 3-person team or 30,000, whether you’re serving people in places that have AI and data protections, or it’s a free-for-all.

But the common discernment thread across all of them is this. In order to see true ROI, organizations implementing AI need to be discerning about which problems they’re solving for.

Now to be fair, this has been true for almost any tech transformation ever. Knowing your ideal state, requirements, use case, constraints and capacity has always been part of the responsible implementation playbook. In theory, AI would be no different.

And yet AI is having a distinct hype moment right now, that’s creating a unique sense of urgency among AI-curious organizations. This idea that you’ll be left behind if you don’t jump in, because everyone else is already doing it.

And by the way, those that are doing it aren’t all doing this well. That’s a whole other Ted Talk for another blog post.

But as it often goes, the hype campaign trumps reality. The orgs not doing it well aren’t likely to disclose who they are. For the bigger corporates, there’s also real incentive to publicly embrace AI – and its human replacement potential – which has translated to positive stock impacts.

Even recent drama around the major AI founders has put the tech on the main stage – and in regular conversation – in a way that previous technological advancements just haven’t seen.

And if we’re being fair…there is a valid “shiny object” syndrome happening with AI tools right now. After all, we’re witnessing AI do things the public would’ve thought impossible only a few years prior. It’s an understandable trap then, to think that simply having AI is bound to make your organization better.

But discernment as a starting point – even if you find that AI has other applications you didn’t consider – helps ensure that:

1) Your org’s resources are aligned with what AI actually costs – especially as the major suppliers move to token-based usage cost models that make financial predictability a challenge. But also in terms of the implementation costs: an effective org-wide Claude or ChatGPT rollout isn’t going to be as simple as flipping a switch, or purchasing licenses. It never is.

2) Your org’s project execution function stays focused – whether that’s led by a true PMP (project mgmt professional), a designated staff member, or a multi-person working group. Scope creep is a very real, frequent occurrence that can silently derail even the most boring, straightforward-seeming projects. You need that focus dialed up to 1000 for a project as flashy, uncertain and controversial as an AI rollout.

3) Your org smartly integrates AI and human workflows – since you know what your AI experiment intends to accomplish, you can identify where it needs human oversight. You’re also clear on which tasks should live with humans vs AI, and why that’s the case. You also don’t make the mistake of building a dynamic where people are overly reliant on AI to make your org function.

4) Your org knows what results to look for – because seeing true ROI with AI is the best case scenario. And in the event it’s actually not there, you’ve carved out the space to be able to pivot with certainty.

The alternative? An org that finally has AI and, at best, isn’t using it to its full potential. Or worst case, an org whose use of AI is creating hidden bottlenecks and silently amplifying organizational dysfunction.

Much, much more to come on this topic.

To wrap this up

Get out there and question everything – whether you lean on the optimistic or pessimistic sides of the AI debate. Exercise that discernment muscle. Encourage your teams to do the same. And don’t make life-altering decisions all because a chat bot – even the smartest one in existence – told you so.