Hi readers,
I’m back for another instalment of my inflation series. If you missed it, check out Part 1 — Creative Inflation.
I want to preface by reminding you that when I discuss inflation in these blog posts, it’s not always a direct 1-1 comparison with the concept of economic inflation.
The mechanism - increasing the supply of something (money, or content, or in this case, busy work) - devalues the existing supply. But while the impact of economic inflation is on a population’s wallets, these other types of inflation impact quality of life.
Now, to discuss the topic at hand: busy work.
How many of us, especially those into coding or a technical field, before the age of AI, would spend hours trying to reach the magical status of “perfect” when you already achieved “good”, ages ago? Quite frankly, you don’t even need to be in a scientific or technology-based field. I worked at a non profit years ago where sending an email to a member sometimes took an hour of debate between two members of staff on exactly how to word the message.
Not exactly the best use of our time when the response was usually something like ”👍”.
However, nowadays, I could imagine receiving a 2000 word response from that same person that was clearly ChatGPT generated, asking us to provide more detail, more legwork, and more effort on a project that just needed to be “okayed”.
The problem with AI is two-fold: it has allowed those who don’t like putting effort into a response to generate huge amounts of “work” that doesn’t push the needle forward. On the other hand, it encourages the more neurotic of us - the perfectionists, the over achievers, and the people pleasers - to spend hours getting our reports, our code, or our projects just right.
I’ve even had Claude push back on me to tell me to stop fiddling! That I had done enough and it was time to move on to the next step.
So, why is this a problem?
More work means more burnout
We are not seeing the gains from AI across the board the way the tech giants predicted.
In many large organizations, where negative impacts scale quickly, you’re seeing hours upon hours of extra work necessary to review long technical documents, or new features being pushed by a sales lead who just learned to vibe code, and often, the ones picking up the pieces are the perfectionists.
There has always been an imbalance between these two sides - the “good enough” Greg’s and the “perfectionist” Patricia’s. But AI has multiplied this difference exponentially. There is too much supply of “scrappy get it done” attitude and not enough demand by the “making sure it’s good enough to ship” labour that usually provides a bottleneck. The bottle is breaking under pressure and enshittified garbage is pouring out. Companies like Microsoft are shipping flagship products that break.
Now, what does this mean for small business?
Luckily, small businesses have a few things going for them.
- Our projects are not as large and complex. There is only so much documentation you can bog a manager down with - they aren’t managing a team of 200 people, so they can get a handle on the workload easier, and let employees know when it is getting to be too much. A 5% increase in output of reporting by 200 people is a lot more than a 20% increase by 5 people.
- Our workers have more power in decision making. At large companies, rolling out AI has been a very wild-west experience. The “smartest” companies are doing it this way: Employees being told to choose a tool, work with it, write a report on its impact, and then the manager can decide if they keep it. They don’t have time to come together as a group and figure out where the cross-functional impacts lay. A smaller organization can get more of the company in a single room and figure out a more precise strategy.
- Fewer seat-filler employees. Let’s be real. At large companies, there are often more employees who don’t really contribute, anyways. At smaller companies, save a few instances of nepotism, that’s harder to get away with. Your likelihood of balancing the scales between good enough Greg’s and perfectionist Patricia’s is higher.
Now, I write this all to say that of course AI is causing huge problems with the amount of output we can produce. But, I see that problem laying more with large businesses who already suffered from some of those issues before AI. Small businesses struggle with output - not enough team members to run a social media account, or remember to order supplies, or chase unqualified leads.
And solopreneurs?
You need to be wary of how this may impact your customer feedback (getting higher levels of “scrutiny” that are based on an LLM’s perspective) and getting too “in the weeds” when you rely on AI as your extra “assistant”.
Here are some strategies to help you avoid these issues:
- Pass your work for a client through an LLM to challenge it before you submit it.
It’s not foolproof, but you’ll get some ideas of weaker spots that you can tighten. I believe if you’re going to be tested by a computer, the best way to trick it is with another computer. You can even run it a couple times iteratively - though don’t get carried away.
-
Get comfortable with “good” and not “perfect”. You are your own boss, and you are the subject matter expert in your domain. You know when the work is good. You also know that building the work to perfection includes a lot of scope-creep.
-
Build some workflows to handle some of your most time-consuming tasks so you can deal with more direct client issues. If you’d like, I can train you on using a low-code platform like n8n, or you can take a quick look at my AI Cheat Sheet for some ideas on what kind of setup you might need.
Where does the real magic happen?
The less visible AI is, the better. I know people like to focus on the “Generative” part of GenAI and make long documents, or huge amounts of content, but I’d like you to think of AI like the mythical figure of the “brownie” - they come out at night and perform the chores and tasks you don’t want to do. Ok, maybe writing a report is something you don’t want to do - but equally, updating a spreadsheet on a regular basis, or your competitive benchmark, or monitoring your services for errors (even fixes them and only alerting when there’s an issue that needs human attention) are all great ways to get time back in your day without creating extra work. As long as you make sure to use the necessary guardrails to prevent an AI Agent or Workflow from deleting your entire filesystem, or sending out the wrong information to customers, there are all sorts of ways to use it in the background.
In conclusion
On a macro scale, higher productivity has always resulted in higher levels of work - but that’s a large company issue. You know, some of them would chain their employees to a desk if they could get away with it (and some of them already basically do).
You’re a small or medium business owner, or a solopreneur. You’re already doing things “differently” than the big guys - so keep it up. Maybe you have handshake deals, maybe you have a friends and family discount, maybe you have flexible working hours, or other benefits the large companies can’t provide at scale. Make sure you use AI to get more work done where it makes sense, not just to make yourself or your employees “look” busy.
These are exciting times we live in when it comes to technology, and the actual revenue impact of AI is here to be felt by those companies who can use their strengths and pain points to harness it effectively.
If you’re interested in using AI to help your business grow, please get in touch.