https://www.linkedin.com/pulse/ai-crush-your-career-turbocharge-russell-brand-jx1cc/
Leaving aside the new and growing career in fixing typos in computer-generated images, what else do you know about the future?
There is work that really shouldn’t ever have existed. Remember well, the “B-Ship” in HitchHikers Guide. And work that should have been automated even without AI. Think of highway toll booths. We can expect most things like that to go away.
We should expect that most of these jobs have already gone away. And to a large extent, they have.
The November 2023 “The Impact of AI on UK jobs and training” from the UK Department of Education basically says (in 33 pages, peppered with equations and citations) that the areas where AI will not most significantly impact jobs are the areas where simpler automation technologies already have. Everything else will be impacted. The big question is who will be augmented (augmented is their word — it sounds so much more scientific than turbo-charged) and who will be replaced.
If those are the only two choices, I know where I want to be.
There is work that humans do where it is hard to replace humans with technology, but where technology (AI or otherwise) can allow humans to be much faster, more effective, or more efficient. Here, some of the humans will get turbocharged, either because they work well with the technology or simply because they are very good. A few of those humans will replace larger numbers of mere mortals. Those mere mortals will lose their jobs, but have they really lost them to AI?
There is some data that suggest that a top programmer with AI help can accomplish about twice as much as they normally would have. Many simple non-critical programming tasks can be done by non-programmers using the current LLMs in much they same way the “no-code” was adopted.
Between these two effects, how many mediocre programmers will no longer be needed?
There has been a trend to move to “self-service,” whether that is for customer-facing tasks (support, sales, etc.) or internal tasks (dashboard and report generators). AI will enable more and more tasks to be pushed out to self-service. Right now, a professional whose time in their field is very valuable can spend five hours with a do-it-yourself graphics tool instead of hiring a graphics person to spend a single lower-paid hour to produce a better result. Will AI make that ratio less unfavorable? Or push us to be even sillier than we are now?
Cal Newport has raised this question in another context previously: Where are we doing things ourselves because the technology makes it possible, rather than letting someone else do it who could do it faster, better, and ultimately cheaper?
For more and more tasks, AI will allow non-experts to do things whether or not they should be doing those things. For those of us who currently perform those tasks, the biggest challenge may not be how to incorporate AI into our work but rather how to help our clients choose to let us do things for them well rather than allowing AI to do even more things badly.
Perhaps being able to explain the value of quality will be the single most important tool to thrive in the coming age of AI.
#ai #aijobs #llm #chatgpt #artificialintelligence
