Über 80 % der Unternehmen vermelden trotz Milliardeninvestitionen bisher keine Produktivitätssteigerungen durch KI, wie eine Umfrage zeigt – 6.000 Führungskräfte geben außerdem an, dass 1/3 der Führungskräfte KI nutzen, allerdings nur für 90 Minuten pro Woche
https://www.tomshardware.com/tech-industry/artificial-intelligence/over-80-percent-of-companies-report-no-productivity-gains-from-ai-so-far-despite-billions-in-investment-survey-suggests-6-000-executives-also-reveal-1-3-of-leaders-use-ai-but-only-for-90-minutes-a-week
25 Kommentare
The lengths these people go to and the BILLIONS of money they are wasting just so that they can stop paying people to work. I really hope this backfires really soon.
As someone who’s been coding and designing software for 15+ years, I find it insanely and scarily productive
It’s intended purpose is to create a pen, a virtual pen for us all. We are cattle and AI will be the cage. It was never about productivity.
So tired of the AI hype and bubble.
I wonder how much of this is because those who need convincing to proceed (ie the board) have no real clue and therefore slow down potential progress. And perhaps they aren’t wrong. I feel anyone who has benefited has probably taken huge security risks. In summary, i reckon the speed of implementation is directly linked to capacity to accept security risk.
I bought a pet rock, it just sits there like a rock.
We are at such a funny inflection point where AI is so powerful but there are still a bunch of haters and people out of the loop. And the people in the loop know what’s coming, since it’s already here. Going to be a great year
As someone managing a small sales-oriented business unit, I find it tremendously useful, and it has increased productivity dramatically along the margins
Like, the core activities are not impacted much, but the fringe activities that help sustain and drive the core activities benefit greatly from AI
Every hour less spent on admin is another hour that can be spent generating revenue
Is the bubble gonna Pop?
Because they figured out AI could replace them just as easily. Ooops, doesn’t work.
Ok yeah but what if I told you they only do 120 min of work a week? Yeah, bet you feel foolish now
Duh, like outsourcing your IT departments….
It’s not about productivity gains, it’s about cost (i.e. people) reduction. Not sure if people are really getting this yet.
In the end it’ll all be about surveillance and weapons. By then it’ll be too late. We will live in modern times and it’ll be bad then become ok as we get used to it. It’ll be like Black Mirror. Your mom died? Here order this orb with her entire consciousness digitized for a subscription of $299 a month. Oh here’s an in house robot that’ll malfunction every 6 weeks and you’ll have a $900 a month subscription to that and some minimum wage electrical engineer will come fix it when it breaks but hey the robot folds your laundries and grabs your packages from the Amazon drone.
Everything becomes commodified.
Question about leaders using AI is an interesting one. Most leadership positions are held by older individuals, which are often less likely to use modern technology directly, even when they promote the use of it within the organization. I’m not surprised at all with that adoption number, but it does not represent a lack of interest or benefit.
It’s ridiculous to think this will continue to be the case. Deployment, training and trouble-shooting automation systems replacing legacy, labor intensive systems are enormously labor intensive tasks in themselves, but once they are up and running, you see industrial strength attrition and job loss. It’s inevitable. AI is not another cotton gin or mechanical loom. It effect every conceivable area of human endeavor.
This is highly suspicious of being a them issue, honestly. I have been using ai to fix my own problems with code for months now and being able to produce working, reasonable code as a solo developer has never been easier than it is now. Looks like their experience is shaped mostly by their culture allowing only a certain kind of output to count. Held back by processes and red tape is my guess. I „ship“ thousands of lines of features a week. Productivity, as one of the useful metrics, is ridiculous. What are they doing with it?
I have access to 5 models and I have yet to use it for much more than google searches.
The study itself drew parallels with the IT boom of the 80s and how the same trend happened: productivity dropped as companies were adopting personal computing devices:
“dividends of a productivity-enhancing disruptive technology were reaped only slowly, with an initial lag, over the course of decades, due to the time required for the technologies to diffuse into common use, and due to the time required to reorganize around and master efficient use of the new technology.”
The danger of AI is it is confidently wrong all the time. If you aren’t an expert in your field it becomes a liability. We are training entry level people to be prompt engineers and not experts in their industry. So bad data, wrong info gets passed as facts and real decisions are being made based on it. It’s pretty scary what it is already happening in many sectors where Sr level is being laid off across the board and no brain trust is left in an org, but AI prompters.
I currently have pull requests for 11 finished features pending code review, that’s why.
The workflow is still October 2025, the tools are February 2026. The chief architect doesn’t trust it, so a human has to look at every line. Then they send it back, „it’s too much, split it up“, and then I spend a day splitting, but then that turns a finished feature into a series of a dozen pull request that each depend on each other, so they can only pass through the work flow sequentially. I spend half the day keeping all my branches up to date with main.
I’m literally going slower than I was before, and it’s not because of the tools, it’s because our work flow hasn’t adapted yet. Right now, we’re not driving on the highway, we’re going 90 in a 45 zone, and getting to the next red light much faster.
A few months ago these tools were helpful but not transformative. Now they are super human. It’s way too soon to expect huge productivity gains in legacy organizations, we haven’t figured out how to adapt the work flow to the new reality yet. It’s obvious what needs to happen, humans need to stop looking at the code and concern themselves with validating the specifications and product, and for that trust needs to develop. That trust isn’t there yet, including with me. I’ve worked on code so long I can’t let go of that being my product yet. Give it time, it’s been just a few months.
I don’t believe that
You mean a piece of tech fueled by marketing hype and investors searching for returns isn’t all that? Surprise, surprise.
Folks made it sound like “Buy our AI, fire everyone, count your money.”
I gotta say it’s really annoying when I see a message from the VP or CTO at my company that is clearly AI assisted / generated. It makes me not want to read it like “you didn’t put any effort in so why should I.” Anyone else feel that way?
lol I use it all the time for giving it a screenshot and it ocr’ing and reformatting.
Not helpful my ass lol