AI FOMO: My Robots need Jobs
AI was supposed to save us from productivity anxiety. Instead, I’ve somehow reached the point where I feel guilty because my robots aren’t working hard enough.
I have developed a new and extremely stupid form of FOMO.
Some (...most) of the AI tools I use give me a finite amount of serious agent time each day. These agents can research things, write code, build websites, analyse information, create content and generally scurry around doing useful things while I get on with something else.
Which creates a problem. What if I don’t use all of it?
If I get to the end of the day with three hours of agent capacity left, I increasingly have the uncomfortable feeling that I have left a small factory full of highly educated digital workers standing around doing nothing. Surely they could have been researching a business idea? Building me an app? Finding a cheaper mortgage? Creating something to sell? At the very least they could be doing something.
AI FOMO. ....IKR? FML.
It wouldn't be so bad if my AI workforce was free - but they're not. I pay $$$ for these robots and it sometimes feels like the human equivalent of having a team that you're giving no work to - who can blame them when they put their feet up.
From tool to workforce
The first phase of GenAI felt like getting access to a very clever tool. You asked ChatGPT a question, it answered, and you carried on with your day.
Agents change the psychology because they introduce the idea of delegated machine labour. I can now hand over a reasonably substantial task and have a machine spend time working on it. Better still, I can potentially have several doing different things simultaneously...
It's a surprisingly profound capability for an ordinary person to suddenly acquire. Five years ago, if I wanted to build a functioning website, produce professional-looking graphics, analyse a large dataset and create a marketing campaign, I either needed to possess a slightly implausible collection of skills or pay several different people. Likely the latter. And it wouldn't be cheap.
Now I can at least attempt all four before lunch.
What should all these robots actually do?
If AI labour becomes abundant and cheap, we move to a strange world where really anything can be built at mach-speed. You can immediately see this by the sheer quantity of 'AI slop' meandering into every corner of our lives. So when it's easy to make the thing, the onus has to shift much more onto "is it actually worth building?".
The cold reality is that nowadays, anyone can become (in the AI-enabled sense) a:
- Graphic designer (I mean, check out my own sub-par attempt a few weeks back...)
- Developer
- Copywriter
- Researcher
- Project Manager... Game Designer... Data Analyst... Video Editor... AI Consultant... (yikes..)
Don't worry, I get that real humans doing these jobs properly are generally still far more capable than an average joe asking an AI to do it - but how much longer will that last I wonder.
I wrote about this democratisation of capability now that AI has come along in an earlier blog. It's wonderful but also scary. If I could be all of these things then... I better get busy shouldn't I?
And that leads me back to my AI FOMO. Perhaps the correct response to unused agent capacity isn't to desperately invent work for my robots. It is to recognise that something rather extraordinary has happened: ordinary individuals are beginning to have access to something resembling a tiny workforce on demand.
But either way, I've got at least 3 hours of agent execution time left today...
...what should I get my agents doing?