It’s not pleasant when things go to hell in a handbasket.
The type of hell and the parameters of the handbasket may vary, but in general it involves the sudden need to stop doing X because now you need to do Y, which wasn’t foreseeable up until this very moment in time, and no one’s sure why or what’s going on. Bonus points if you’d almost made it through to completion of whatever X was.
Then, tomorrow, you can wake up and reboot all over again. There really ought to be a badge for simply surviving it.
Bad news, campers, in that it looks like this isn’t going to settle down anytime soon. Then again, we’ve all been fooled before about timelines, burst bubbles, and corporate calamities of various sorts.
That, too, demonstrates the handbasket we’re parked in. Hence, Step 4 of the Org AI Literacy framework is getting comfortable with chaos.
Clarification upfront: I’m talking solely about workplace chaos. There’s a lot of other chaos, too, that we arguably shouldn’t be comfortable with. How you manage that is a personal choice I’ve got no business speaking to. But if it all gets to be too much, there’s no shame in seeking help from a health professional.
The 5 circles of chaos hell
Per the dictionary (accessed directly, not through Google’s new “dictionary” replacement), chaos is defined as a state of utter confusion or disorder; a total lack of organization or order.
Since one person’s disorder trash might be another person’s freedom treasure, let’s unpack the workplace dynamics that are often considered chaotic.
Ambiguity and uncertainty
There’s no plan. There’s no plan for a plan. All we’ve got is a hand-wavy vision of something that seems cool, but it’s not clear how it’s possible, who’ll do it, what it’s for, or why it’s the right thing to pursue. We’re sitting in meetings and nodding wisely as if we’re in agreement and know what happens next, but we each have something different in mind.
Free-for-all
Eff the strategy! Eff the plans! Eff the status reports! We’re gonna move quickly and break things until something takes hold. Just do it. Trust your gut. Don’t stop to tell anyone what’s going on or ask for more context. Don’t worry that it’s literally someone else’s job to figure this out or that four other teams are working on the exact same thing. To the victor belong the spoils!
Head-scratcher directives
The big bosses are making decisions that seem to have no bearing in reality, and when we ask them why, the story changes from day to day. “Build some agents to do it” is bandied about as if it’s the solution to everything, and us front-line folks are stuck figuring out how to make it work OR ELSE.
Busy-ness theater
We’ve taken 120 different goals, everyone’s working 80 hours a week, we’re tokenmaxxing like fiends, and shouting from the rooftops How Very Busy We Are! What are we working toward, you ask? Well, lots of things, but we’re so busy we can’t stop and tell you what they are. Here are 50 different status reports if you’d like the proof.
Whiplash and thrash
We’re trying really hard to make things work, but every time we get traction, something changes dramatically and we have to pivot or redo a bunch of stuff. We gave up any semblance of tracking goals, and the user stories have been redefined and reprioritized so many times it’s not clear who the user even is any more.
None of these chaos engines is new in the world of business. But, as with all things, AI accelerates and tangles it up even more. There’s never time to get used to a change before another one comes along and steamrolls it.
Caught in a permanent storm
If you’re surrounded by chaos long enough, you’ll feel like you’ve lost control, lost the plot, and lost your mind all at the same time. Humans can only take so much of this before their better angels fly away. Frustration, resentment, and anger ensue, and even the most mild-mannered employees start to behave in uncharacteristic ways.
Now add a dash of AI-related existential crisis on top of it, and you’ve got all the fixin’s for a work-culture nightmare. No one wants to work anymore. No one wants to be at work. Productivity falls off a cliff, and output quality follows. Which only adds to the vicious chaos circle.
Good times.
The knee-jerk reaction is to try like hell to put everything back in order. But as much as we’d like to do that, we’re smack in the middle of unprecedented disruption. There’s no way to avoid chaos, and often no way to even mitigate it. Someone could announce they’ve achieved AGI six months from now, and it would turn technology on its head again. Or the economy could go badly sideways and make the Great Depression look like child’s play.
Or it could all work out just fine. We just don’t know.
The way out is through
When chaos strikes, the trick is to reframe it: You’re not changing the plan. The plan is change.
That’s uncomfortable for pretty much everyone. Humans like predictability. So your systems will need to establish change as the norm, and a predictable path for handling it quickly.
Faster planning cycles
Leaders still set direction (see framework Step 2), and the front-line folks still identify milestones and implementation steps. The planning cycle itself, though, speeds up. If you’ve been setting annual plans, move them up to quarterly. If you’re spending days and weeks workshopping vision decks and roadmaps, cut that time and the size of the deliverable at least in half. You still want due diligence to confirm you’re going in the right direction, but put more emphasis on reviewing data and outlining the desired outcomes instead of precise actions.
Build iteratively
This applies mainly to things you launch — processes, products, programs, campaigns — and not business-as-usual deliverables. (Half-baked marketing collateral is not a good idea.) Don’t wait for the picture-perfect, one-and-done launch, because by that time, you’ll probably have to walk back or redo some of what you just introduced. Work quickly and get something out there that gets the basic job done or sets the stage for additional changes. Then you can gather information about what is and isn’t working, and refine. Framework Step 3 goes into more detail.
Workstreams
We’re used to organizing work in terms of siloed functions: sales, marketing, finance, IT, etc. This sets up long alignment cycles and dependency risks, though, and so you might have to rethink your approach. There’s a lot of buzz about super-IC’s and tiny teams that do All The Things with help from their trusty agents. However … agentic work needs a lot of human judgment, and there are very few people who have expert judgment in every single function. Smaller, project-based teams might work well for your business, but experiment with a few structures before you do a sweeping reorg.
Decision-making
There’s also a lot of buzz around “flattening” (aka, getting rid of managers) and pushing decisions down to the people closest to the work. If you’re only flattening, though, you’re working around the problem. These structures formed because there was a governance issue to solve, and that issue is likely to return. You do still need to get rid of long approval chains, endless proposal refinements, and information overload, and leaders need to be OK with not having Yay or Nay authority over every little thing (see framework Step 1). But you also can’t go governance-free. Figure out where the key touchpoints are for accountability, and keep those.
Rules of engagement
These are the touchpoints and ground rules that provide stability. For example, make it OK to take action with only 70-80% of the information. Revisit annual goals at least once a quarter, and get in the habit of defining clear success criteria for every level of execution, so you can quickly course-correct without a lot of dithering. If a team identifies something they need to pivot or want to move forward with, put tiers in place to identify whether to just do it, whether to have a chat with a pre-identified leader first, or whether it needs a more formal justification because of other business implications.
One other really important thing: Your “predictable” path for handling change will also likely need to change along the way. Stay flexible, because rigidity will fool you into thinking you’re adding stability, when you’re really hindering progress.
Will there still be hell in your handbasket? It’ll probably seem like it while new ways of working take shape. But, as it turns out, the steps you’re taking to manage chaos are also the systems you’ll need for governing and collaborating with agents.
Why not get out ahead of it early?
All opinions here are my own. All text is my own, too, including the em dashes. I welcome constructive comments and discussion on LinkedIn and Bluesky.