AI: the watershed for skills?
As with all innovations AI has created its own manifest destiny which will either vaporise human skills, or stop the productivity paralysis that afflicts most developed economies. Such dichotomies are comforting, like choosing between good and evil.
Whilst AI stocks have been rampant, the safe bet is that the human workforce is going to be augmented. However, more research and direct experience suggest the clear line of AI dominance is less assured. It may be that even writers have jobs in the future.
Whilst anecdotes are not sampled peer reviewed analysis, let me share one.
Last year, Steph, Lizee and I used AI a lot for a project. We needed to draft functional roles and their associated KPIs. ChatGPT took the strain out of the workload by presenting usable texts that could fit our needs.
But it wasn’t perfect. For a start it was verbose. Ten words filled the place where five would suffice.
This happened while we were putting the sdr.advisory site together. Lizee used the AI function to get a description of what we are and do: it came up with a superb marketing text that was stylistically like a merge between Forbes and Rolling Stone. I was so pumped reading it, I had to use a defibrillator.
That characteristic was the first indication of it being AI: no one in business writes like AI, with all the added syllables, or if they did, it was 150 years ago.
Padded texts meant we had to edit the texts back to something that fit normal reading times and patience, and expressed the central information in a way that anyone could understand quickly.
We had become editors. Now that may seem new, it isn’t. I have been doing the same role with Excel. I put the data into the program and create the arguments, Excel spews out the result and the charts and I determine if it means anything; if it suits my needs.
In both cases - AI and Excel - my productivity jumped, even in net terms after editing long winded sentences. That type of evidence is why Goldman Sachs, amongst others, produced global job profile analyses that showed AI would decimate entire skill sets. On that basis buy more MFST and NVIDIA, hello UBI to protect society.
But then, as with all hype cycles, the more research suggests that hallucinating AI is just a feature and that, well, a skilled person is off the endangered list, they will transition, probably. That is encouraging for all of us, not least for governments who rely on tax.
It is still a challenge for employers, for businesses, which define jobs and the nature of work through the proven past, when the jobs, in IT certainly, are changing rapidly, as are their type and boundaries.
Role classifications are also strained because they are compiled periodically but the market in AI job platforms can assess millions of jobs and determine, de facto, what is current and relevant.
This disruption adds another type of uncertainty to working, even if the AI hallucinations are no longer a feature.
