AI's Productivity Boost? Just 16 Minutes Per Week, Claims Study (nerds.xyz)
- Reference: 0180983980
- News link: https://it.slashdot.org/story/26/03/14/0539247/ais-productivity-boost-just-16-minutes-per-week-claims-study
- Source link: https://nerds.xyz/2026/03/foxit-ai-productivity-study/
> According to research cited in Foxit's [3]State of Document Intelligence report , while 89% of executives and 79% of end users say AI tools make them feel more productive, the actual time savings shrink dramatically once people account for reviewing and validating AI-generated output.
>
> The survey of 1,000 desk-based workers and 400 executives in the United States and United Kingdom found executives believe AI saves them about 4.6 hours per week, but they spend roughly 4 hours and 20 minutes verifying those results. End users reported a similar pattern, estimating 3.6 hours saved but 3 hours and 50 minutes spent reviewing AI work. Once that "verification burden" is factored in, executives gain just 16 minutes per week, while end users actually lose about 14 minutes.
[1] https://nerds.xyz/2026/03/foxit-ai-productivity-study/
[2] https://www.slashdot.org/~BrianFagioli
[3] https://www.foxit.com/state-of-document-intelligence/
Results Likely Skew to Optimistic (Score:2)
Given the size of the study, the difference of just a few minutes looks like noise. I'm thinking that because the sample chosen was composed entirely of people already using AI, that many of them are early adopters who would naturally take a more optimistic view of AI, skewing the results more toward "time saving" than is actually warranted.
I remember a time when... (Score:2)
...the same questions were asked about CAD
Salespeople sold CAD to management as reducing the time it took to make drawings
Engineers saw it differently
Back when it was hard to make or change drawings with a pencil, we made less of them. The effort required to redraw a complex drawing was so great that only the most necessary changes were made
With CAD, we could easily make lots of drawings with many different versions of an idea, and then pick the best one
CAD didn't increase "productivity", it increased quali
Re: (Score:2)
It's true that CAD improved the quality of designs. However, all indications so far point to AI degrading the quality of work performed with its assistance. AI is not saving time, but it is helping to produce worse outcomes.
Re: (Score:1)
And the worse your product, the more frequently must your customers replace it ... buying your new *,ai designed "upgrade" .... we know where that goes ...
At least it's a positive number! (Score:2)
No minus sign in front of the productivity change.
They hid the negative using apples and oranges! (Score:2)
Sarcasm sure, but I wonder if they figured this out on their own or if they used AI and spent an hour figuring out how it hid the negative.
estimating 3.6 hours saved but 3 hours and 50 . Converted that is 3.6 and ~3.83, so regular people see a savings of about -0.23 yes that is a minus sign.
Well Duh! (Score:2)
"account for reviewing and validating AI-generated output" but then the fix is easy "don't review or verify anything!"
Doesn't mention how often they had to correct (Score:2)
If you're spending 4 hrs a week on verification but it's correct 90% of the time, maybe you're wasting your time.
Re: (Score:2)
That depends on what you're reviewing. Will this code change expose all of our customer data and get us sued and fined into bankruptcy? Dude, chill! The AI is like 90% sure it's OK!
Re: (Score:2)
"We've achieved one 9 of reliability; is it really worth spending more than 10% of our time trying to improve?"
AI is a tool. As are many of its users (Score:2)
Meanwhile, AI probably saves me a day a week.
I use it a couple different ways, as a well-supervised junior dev who does the crap I don't like.
And then as a high-knowledge tech resource I use as a sounding board for technichal architecture questions.
In neither case do I trust it completely, but by working with it similarly to how I work with some of my on-the-spectrum techs, I get good results.
Ok, but hear me out... (Score:2)
If we increase our investment just 10-fold, we could probably get that to 16 minutes per day . Imagine the possibilities!
Mental load reduction? (Score:2)
Did they expend less mental energy by the end of the day? Did the LMM make tasks less intimidating so they were more readily worked on? 16 minutes saved AND less metal energy expenditure would be significant. Would be an interesting follow up study with this cohort.
Re: (Score:1)
Are savings in 'metal energy expenditures' measured in watts .. or twinkies/break ? Are you implying that use of *.ai leads to lower air-conditioning costs ?
Not even the worst of it. (Score:2)
There is, presumably, an amount of time savings where this could be justified(at least for things that you, ultimately, only do because they pay the bills; not ones of some intrinsic value); but it seems particularly grim to deal with the changed nature of the work for such paltry savings.
Going from 'thinking about things you know about' to 'keeping a close eye on an erratic intern who can bullshit really fast' is a fairly dramatic downgrade in terms of the quality and apparent futility of what you are d
Sounds about right (Score:2)
I've used AI to generate python code. It saves me a little time as it gives me a template of what I wanted. The code however is almost never correct and most often does not run correctly. I still need to study the code and fix it. It's always wrong on tiny details. Like I had one that best described as a shopping cart template. I have a list of things. Compute a metric for each item. Also compute a metric for the total of all items.
Things AI code got wrong: Item = Array(i) is not the correct way to get an
Why i'd never vibe-code: editing isn't any fun. (Score:2)
That's what it comes down to. When you start vibe-coding, you're no longer really coding, and you're not even really creating anymore.
You're just editing. All you're doing is code reviews and quick bug fixes...and those tend to be my least favorite parts of my job.
At least code-reviewing a junior developer, you're teaching, mentoring, instilling some new disciplines or expanding their horizons.
There's no satisfaction in doing that to a bot. Especially because the next time it codes something for you, it is
Most important employee at Renholm Industries (Score:2)
The shit is really going to hit the fan when Stockholders realize the A-suite can be replaced by Jen, Roy, Moss and Richmond. Stockholders should demand all management be replaced by AI. I think the most important employee at every company today is Richmond. Renholm Industries runs much more efficient today because it is run by AI. Richmond monitors the mainframe where the AI lives. Richmond does not know what the lights mean but he will notify someone if any lights stop working.
let them eat cake! (Score:2)
Now go convince the management.
They will sell out their best and brightest for short term gains, collect a golden parachute then do it again at the next company.