Employees say AI isn’t saving them much time in their daily work so far, and many report feeling overwhelmed by how to incorporate it into their jobs. Companies, meanwhile, are spending vast amounts on artificial intelligence, betting that the technology’s power to speed everything from sales to back-office functions will usher in a new era of efficiency and profit growth.
The gulf between senior executives’ and workers’ actual experience with generative AI is vast, according to a new survey from the AI consulting firm Section of 5,000 white-collar workers.
Your comparison doesn’t seem to make sense. Employees that are not benefiting from AI ARE NOT the ones making the jet engines. Employees are the ones flying the planes.
My corporation has rolled out AI in a big way - we have have numerous trainings and testimonials from employees that profess to how beneficial it is. Frankly, I’ve used it twice and in both cases it has made more more lazy than productive. I used it to create slides for a presentation and to create scripts for clients - things I could have done myself (if I wasn’t lazy).
If you are a dish washer, waiter, or assembly line worker I doubt you see much impact from AI.
People who shuffle papers and information see much more benefit. But in time AI should make better decision making. And that will trickle down to more junior levels—but the impact may be less seen as due to AI.
Without reviewing the actual survey, it’s difficult to know exactly what is being measured. I’m guessing that executives’ opinions about productivity gains are directed toward how much time they’re saving, not how much time their workers are saving.
It could be that company executives are using AI more, and benefitting from productivity gains more than their workforce.
It could also be that the tasks executives perform are better suited for AI. I hope this is the case. If so, eventually AI will take over executive leadership positions at major corporations, boosting shareholder value!
The employees numbered 5,000. This is a survey, and there is no “editorial narrative”, it’s a survey done by an independent research company, and reported on by the Wall Street Journal.
I note that you love to shoot the messenger when it’s a message you don’t like (which “doesn’t conform to your personal editorial narrative”.)
And the results of this survey is not a matter of “a few points”, the gulf between upper management perception and workers-in-the-trench reality is VAST.
None of that means AI is no good or won’t get better, but it does indicate a certain lack of evaluation of tools among those in the C-suite.
Exactly! The survey is a bit deficient…it should break down workers into type of work. AI isn’t saving any assembly line workers any time. Same thing for someone like a plumber, most likely. But workers reading, summarizing and preparing presentations or writing code would, potentially, see some AI gains.
Sure - but that seems like an expected result, right? If AI is replacing workers, rather than only making existing workers more productive, then that would track. The ones who are still actually working there would mostly be the ones whose jobs don’t map very well onto AI capabilities.
Fair point, and if it was 5 or 10 years from now maybe I’d buy the logic: the only workers left are the ones who couldn’t be replaced by AI. But we’re in the very beginning throes; it’s unlikely that masses of white collar workers have been let go because of AI. (I acknowledge there are surely some, but hardly enough to move the needle in a survey reaching 5,000 still employed workers - and I would note that some companies, in their enthusiasm have had to walk back their claims and rehire some of those laid off.)
The BLS put the number of white collar workers at about 90 million a few years ago; you would have to lose multiple millions to have a statistically significant effect on a survey of 5,000. (For reference sake, a survey of 750 is usually enough to produce data with a 3% margin of error. In this survey the differences in perception are sometimes 40% or greater.)
Ah - I went back and reread the piece a little more carefully. It’s not saying there’s a disconnect between the efficiency gains (or lack thereof) that non-management white-collar workers are experiencing and what C-suite execs think they would experience. It’s a disconnect between the time savings that each group experiences for themselves.
Which makes a ton of sense, for a couple of reasons. AI is going to map better onto tasks that C-suite management types do than white-collar non-management types - one thing AI is particularly good at is distilling/summarizing a lot of information, and a big part of management is absorbing information about what actual workers are doing it and then using that to guide corporate decision-making. Also, C-suite execs are certainly going to have better access to training on how to use AI than most other folks. To the extent there’s differences between AI products (and everything calls itself AI), a company is more likely to first buy something that will appeal to the C-suite folks than anyone else.
And finally, on the whole C-suite management will (or should) work longer workweeks than the typical non-management worker, and therefore anything expressed in total hours of time saved rather than a percentage of time saved will show more for the C-suite than the regular joes.
I’m all on board with the idea that AI isn’t going to do as much as hardqore enthusiasts think, for various reasons. But I don’t think this is a surprising result. Except maybe to the upside, honestly - a third of workers reported saving more than two hours per week from AI? That’s higher than I would have thought - especially since I would think that white collar non-management workers at large companies would be more likely to underreport efficiency gains, for self-interested reasons.
Primarily, as a Scribe. As I am in back to back meetings, getting accurate notes - and fast without me getting distracted/slowing down the meeting to clarify and align on actions. - savings is about 40-100 minutes a day.
These are activities which I would have to multitask (poorly!) in follow on meetings or stay late working a sundown list if I wasn’t using that functionality.
Secondarily, as a framework generator. Working in an area that is new/different and NOT a historically rich core competence area within my company means that much of my policy/governance/standards are not defined or are so archaic that clean sheet ideas are often my MO. AI tools take internal and external content and get my requests to 70% (the easy part) - almost instantaneously.
Yes, I have to give good prompts. Yes, I have to check the work. No, I don’t let anything stick in those seed files without careful and cautious review.
While that is the EASY part, it is a life saver that enables me to work to 100% and network a solution without the preamble.
This is less daily as a professional activity, but perhaps 2-4 hours per month.
I could give those tasks to the junior staff member on my team, but then they would have to be everywhere I am all the time. I’d much rather be developing them with their own project/programs(and using AI tools similarly).
The C suite has to be more creative. Or it is more common to be business creative in the C suite.
My nephew is an MIT computer programmer. He does not use AI to do his job.
I am academically a bit trained in a few languages. I use AI. My time savings might be one year of work turned into four weeks of work. With visual scripting, I am starting from scratch.
A friend who is a CPP programmer and set me up in my software stack, tells me he finds visual scripting harder than CPP. It is not intuitive.
Back in October on CNBC the BlackRock CEO (Larry Fink) was asked about AI and jobs. He said the company was now able to expand with the same number of employees.
