Instagram story viewer> @keepithuman.consulting> Posts

Maura Charles | AI on Human Terms

@keepithuman.consulting

420

followers

1711

following

Designing how humans & AI work together for good.
🧭 Long Island & NYC

POSTS STORIES REELS

Silos don’t start because people won’t work together.

They start because every team is being scored on a different game.

Marketing chases one number. Tech, another. Procurement, a third. Everyone nods at the same slide and defines “success” three different ways.

Dozens of priorities on paper. People can really organize around two or three.

From the outside, it looks like momentum. Up close, it’s chaos with a calendar.

I help leadership teams find the few objectives that matter, and actually operate toward them together.

#leadership #organizationalculture #teamalignment #futureofwork #keepithuman by @keepithuman.consulting

AI Won’t Fix Your Sh*tty Operations.

No amount of AI is going to fix broken business operations. It’ll just help you break them faster.

That’s one of the biggest lessons of the past year. Plenty of companies bolted AI onto a messy system, assuming it would sort itself out. It didn’t.

Layer AI onto something that’s already broken, and you get more confusion and more frustration for the people trying to do their jobs.

So rebuilding the foundation can’t be a separate project you’ll get to someday. It has to be part of your AI plan. Decide what you’re actually trying to accomplish. Fix what isn’t working. Then figure out where AI fits.

I know the instinct is to add it now and fix it later. This isn’t one of those times.

Don’t automate the mess and call it progress.

That’s the work I do with leadership teams: building the foundation into the AI plan, not after it. by @keepithuman.consulting

Where Were You a Year Ago, Bill? A Nation Asks.

Is AI dangerous? Short answer: yes.

Longer answer: it’s dangerous on more than one level. To individuals. To businesses. And in what companies can now do with it, intentionally or not.

In August, Bill Gates published an essay warning that the world isn’t ready for AI and that tech companies are downplaying the risks. I agree. I also wanted to ask, where have you been? We needed that conversation a year ago, from the people who knew better.

Here’s the everyday version of the danger. We assume these tools are accurate. We assume they’re acting in good faith. But they’re unreliable, inconsistent, and confidently incorrect. They have all the same flaws humans have, plus a few of their own.

Responsible AI isn’t about being afraid of technology. It’s about understanding the risks before we hand over more responsibility.

That’s the work I do. by @keepithuman.consulting

One Human. An Army of Agents. What Could Go Wrong?

Everyone’s chasing the same idea right now: a whole company run by agents.

Technically, it’s possible. One sharp, entrepreneurial person who knows what’s out there, open source or paid, can deploy an army of agents and do work that would have been impossible for a small business just a few years ago.

Even the SBA calls AI “the great equalizer.” It’s close. Companies with direct partnerships with OpenAI or Anthropic still get early access and a head start. The playing field is more level than it was. It isn’t flat.

But here’s my problem with the one-person, all-agent company: it’s incredibly risky.

One human in the loop means one set of eyes on everything. No one to say “that doesn’t look right.” No one to catch what the agents got confidently wrong. When that person is tired, distracted, or out sick, the whole operation runs on autopilot.

Possible isn’t the same as wise.

The advantage isn’t size anymore. It’s knowing where humans need to stay in charge, and having enough of them there to matter. by @keepithuman.consulting

Three Ways Keep it Human Works With You.

AI won’t improve the way your company works just because you’ve added it.

At Keep It Human, I work with organizations on three parts of the problem.

First, Human + AI strategy: deciding where human judgment needs to stay at the center, and where the technology can actually help.

Second, executive alignment: getting leaders in the same room and on the same page about what’s next.

And third, workshops and town halls: helping teams think through responsible AI, ethical technology, and what all of this means for the work they do.

The common thread is simple: don’t start with the tool. Start with the people and the decisions that matter.

If your organization is trying to figure out what responsible AI actually looks like in practice, let’s talk. by @keepithuman.consulting

One of the more irritating things AI does is try to tell me how I’m feeling.

“You seem upset.”

Really? Based on what?

These tools are looking at patterns in our language and making assumptions about what those patterns mean. And sometimes those assumptions are completely wrong.

Humans do this too, of course. But there’s something particularly frustrating about a machine confidently interpreting your emotions when it doesn’t actually know you.

And that confidence extends beyond emotions.

AI can still make up information and present it as fact. Which means we’re trusting systems that can be wrong about both what we’re asking and why we’re asking it.

That’s worth remembering before you take every response at face value. by @keepithuman.consulting

I caught myself calling a group of AI tools “they” the other day. Then immediately worried I’d just promoted them to people.

Then I got into an argument with an AI agent. Which is a strange sentence to write.
The agent told me that when it apologizes twice in a row, it’s because it feels shame.
Shame.

It doesn’t feel shame. You can program a system to react a certain way to a certain trigger. That’s very different from actually feeling something.

The language these tools use matters. Especially when it starts making us forget what we’re actually talking to.

Now picture that same tool talking to your customers. Or your employees. That’s not a tech quirk. It’s a trust problem, and it lands on whoever deployed the tool.

What your AI says about itself is a decision. Somebody at your company should be making it on purpose. by @keepithuman.consulting

We say “agent” constantly. I’m not sure we all mean the same thing.

Here’s what it actually means: a tool with a job.

I can talk to Claude and ask it questions. Fine. That’s a chat. But you can also build an agent, give it a job or a set of jobs, and let it go do them. Answering your email. Updating your CRM. Booking the meeting. Browsing the web. Less assistant, more Tasmanian Devil: it takes off without you and doesn’t stop until you make it.

You’ve probably heard agents are “taking over the internet.” Less a hostile takeover, more an infestation. They’re just… everywhere. Out on the web, and increasingly inside the systems your business runs on.

So the definition is easy. The harder question is who’s responsible for what they do once they’re loose.

Spoiler: still us. by @keepithuman.consulting

When’s the last time you overruled AI? 🤔

#Tomorrow I’m leading a Human-Centered AI Nook, a virtual business education session for small business owners, hosted by @nooksandsocials a vetted business community based on Long Island.

We’ll talk about where AI quietly makes calls in your business, what to watch for if you sell AI to clients, and how to stay the one who decides.
No tech tutorial, just a real conversation and a checklist you can use right away.

🗓 Wed, Oct 21 · 2–4 PM ET
💻 Virtual · Guests welcome
📩 DM me to register

Humans decide. AI complies.

#SmallBusiness #LongIslandBusiness #AIForBusiness #HumanCenteredAI by @keepithuman.consulting

Trusting the first answer. Believing something because it sounded polished. Fixing the workflow you should’ve scrapped after spending weeks on it.

Oops. 🙃

7 ways high achievers get played by AI, and how to stop doing it again.

Which one’s your oops?

Humans decide. AI complies.

#KeepItHuman #AI #ArtificialIntelligence #Leadership #FutureOfWork CognitiveBias AIatWork WomenInTech DecisionMaking CriticalThinking by @keepithuman.consulting

Did you know that raisins and grapes are extremely toxic to dogs?

Luckily I did, so when I found Lulu the dog with a ripped-apart bag of oatmeal raisin granola, I knew I had to get her to the vet asap - and that my day was about to get more expensive.
She’s resting comfortably and I’m running a couple of hours behind on work now but it’s only Manic Monday so I’ll catch up… #labpuppy #labsofinstagram by @keepithuman.consulting

A few months ago I decided to head up the NYC chapter of an organization called @aipoweredwomen — I partnered up with the amazing Tanaz Parmar Pilla and this week history was made! More than 100 women came in from all over the metropolitan area to connect and devise solutions to making sure women get a voice in the future of AI — tools, algorithms, policies, education, allllll of the aspects.

Upskill. Lead. Transform.

@anduslabs graciously sponsored this launch event at @betaworks in Meatpacking. Catering by @breadsbakery and Whole Foods kept us going.

I’m so grateful for this newly forming community and for all of the people who made this kickoff possible.

And everyone who listened to me stress about whether or not we’d have enough food and beverage. (We did.)

Ping me if you’d like to get involved. We’re building a community that takes action and would love to have you along for the ride. by @keepithuman.consulting