If you are worried about losing your job to AI – you’re doing it wrong!
The true promise of AI for the workplace is not in replacing human workers with new tech like GPT AI. Sure, some functions will be done via automation which people do now or used to do, but these are high-repetition, low-judgement operations. If your job can be largely described this way, you may want to start thinking about what the workplace will look like moving forward. If you find yourself resenting having to “copy, paste, and click and check” all day – there’s good news!
We all see articles about the new whiz-bang AI features like image generation, fake-looking videos, decent-sounding AI Voices, and *sigh* AI Browsers and think “this is it?”, “we are going to be replaced by this crap”? -- No. Those are hype-generating demos meant to get people thinking about what can be done with these tools. So, what’s going to do the work? Agentic AI and purpose-built AI tools for automation – that’s the future of productive AI.
Imagine “building” an agent which can scan your inbox, and automatically sift through all the junk and reply-all email chains to highlight the items you need to focus on – like meeting requests, reports coming due etc. How about an agent which scans inventory levels, looking for the economic order quantities, and checking supply chain expectations vs pricing comparison (all this is high-repetition) and then, drafting a series of orders for a human to verify (high-judgment) who can just click “go” and get on with growing the business. There, that’s the benefit.
While it’s true that some of what is driving the AI frenzy right now is cost-cutting, it is arguably the smallest part. If companies are focused on that, thinking they can just get rid of people, they are doing it wrong and missing the point. Hear me out . . .
Actual productivity for AI is a series of factors which together contribute to the overall throughput of an organization. I think of it like this:
Productivity = (Amplified Effort + Reduced Costs) x Organizational Intelligence
So, there you see the reduced costs are a piece. And when you factor in what you must pay to use the AI, costs go up the more you use it.
Now, let’s look at the Amplified Effort component. If AI can take the tedium out of everyday operations and free your team up to get more accomplished while simultaneously learning the processes as it goes, possibly even helping to keep corporate rules aligned with those processes, that will amplify the effectiveness of the team. That’s real productivity gain and it will increase as you use it more.
Organizational Intelligence is a fancy term for your company getting its act together around business processes. The more time you take to streamline what you hope to automate the better the result. This is a factor all its own and it’s vital to the success of implementing AI in the workplace. This is where management should be focused right now. Getting the organization ready for Agentic AI automation is where the efforts should be right now. There’s the return on investment not cutting jobs but re-aligning people to function in the new streamlined, automated, Agentic workplace.
I hear the creatives out there saying “But it can create pictures and video – That’s my job!!” Well, I hear you, but it’s not creating anything. AI can’t create anything as creativity requires imagination. AI can imitate creativity, but it’s not the same thing. Anyone who’s listened to AI “created” music knows it’s not creative, it’s generated and very much a derivative of actual human made music. There’s no meaning to the sounds. People who are OK with AI generated image and video slop, get what they deserve. As humans, we know what’s art and what’s slop. We should use our judgment when looking for media to consume.
I was told by my guidance counselor in high school that I might not want to go into Computer Science as a career because (and this was in the 1980’s) “Soon, computers will program themselves”. Well, to be fair, code generators have been around since the mid 70’s. Anyone who has had to maintain a code-base which was “created” by a code generator knows it’s crap. The meaning gets lost in the code. Today, vibe-coding will in fact generate code, and maybe even a half-decent MVP, but someone still has to connect it up and likely debug it – very high-judgement stuff.
Jobs will be here for quite some time – maybe when AGI hits that tune will change, but for now, the notes have meaning.
-John Remmler
October 2025