Ich frage mich echt, wann die #Blase platzt:
- Studie vom Media Lab am MIT:
95 Prozent der untersuchten KI-Pilotprojekte in US-amerikanischen Unternehmen führten zu keinerlei messbaren positiven Auswirkungen.
- Untersuchung von McKinsey:
80 Prozent der Firmen, die Generative #KI einsetzten, konnten keine signifikanten Verbesserungen erzielen. Die Hälfte davon gab ihre KI-Projekte daraufhin auf.
- Befragung der Harvard Business Review:
über 40 Prozent der befragten US-Arbeitnehmer berichteten, dass KI-Ergebnisse sich als gute Arbeit „tarnen“, aber wenig Substanz hätten. Wer mit KI-Texten weiterarbeiten müsse, verwendete jede Menge Zeit auf die Korrektur dieses „Workslop“ („Arbeitsabfall“). Autoren warnen: Das zerstöre den erhofften Produktivitätszuwachs. #aibubble
https://taz.de/Studien-belegen-dass-der-Einsatz-von-KI-keine-Effizienzgewinne-in-Unternehmen-bringt/!6125562/

Curious that whenever someone shows me “the cool #AI flow” they built that’s supposed to be impressive, the conversation goes the same way:

Stage 1: “But you don’t understand. You don’t like AI because you haven’t used it right. Let me show you how much you can do it with.”

Stage 2: “Here are the steps in the flow and the instructions I feed to this agent / custom GPT / Claude project. I tell it to do X, reference document Y, and aim for Z.”

Stage 3: “Now, let me show you the results it gives.”
*Writes task, presses to run the prompt.*

Stage 4: “Umm sorry it’s taking a while. It’s fast but not instant. And by the way, the prompt isn’t perfect, you can definitely make it better. I just threw this together real quick the other day. It makes some mistakes, but it’s really good.”

Stage 5: “Uuuuuuh actually don’t look at the output.” *scrolls or stops screen share or pulls device away.*
“You know it’s already doing so well, if I do more prompt engineering it will get really good but I need to give it better instructions. And it ran just fine last night, I don’t know what’s up with it. And this is a cheap model, if we use another model it will be better.”

Stage 6: “You know, you really shouldn’t judge this so much. The technology will improve, it will get there sooner than you know and then you’ll regret not trying it sooner.”

So curious that this keeps happening 🤷‍♀️

#LLMs #work #tech #AIBubble

One other pattern that I kept experiencing last year when I was actually trying these tools in my work (because I had so many people pushing them on me I started to think I was crazy):

Them: “oh you are struggling with X? I have been using AI to solve it and it gives me the solution in less than 10 minutes! You HAVE TO try it!”

Me: “X is really annoying, I’d love to get help with it. I’ll try that, thanks!”

* I go and try to do X, spend hours prompting AI back-and-forth as it keeps messing up and doing things wrong. I give up and do the thing manually, spending 2-3x the time I would have spent if I did it myself from the start.*

*days / weeks later*

Me: “hey, person, I tried to do X with AI and it didn’t work. I kept getting issues like Y and Z. Did I prompt it wrong? This is what I tried.”

Them: *nods, maybe chuckles* “oh yeah, it does that. I still haven’t been able to figure out how to reliably do X. I tend to redo the output completely too. But sometimes it helps me do like this tiny part. Anyway, these tools will improve soon and we won’t have to do so many manual revisions. “

I kid you not, this stuff happened every other week.

#AI #LLMs #work #tech #AIBubble

Curious that whenever someone shows me “the cool #AI flow” they built that’s supposed to be impressive, the conversation goes the same way:

Stage 1: “But you don’t understand. You don’t like AI because you haven’t used it right. Let me show you how much you can do it with.”

Stage 2: “Here are the steps in the flow and the instructions I feed to this agent / custom GPT / Claude project. I tell it to do X, reference document Y, and aim for Z.”

Stage 3: “Now, let me show you the results it gives.”
*Writes task, presses to run the prompt.*

Stage 4: “Umm sorry it’s taking a while. It’s fast but not instant. And by the way, the prompt isn’t perfect, you can definitely make it better. I just threw this together real quick the other day. It makes some mistakes, but it’s really good.”

Stage 5: “Uuuuuuh actually don’t look at the output.” *scrolls or stops screen share or pulls device away.*
“You know it’s already doing so well, if I do more prompt engineering it will get really good but I need to give it better instructions. And it ran just fine last night, I don’t know what’s up with it. And this is a cheap model, if we use another model it will be better.”

Stage 6: “You know, you really shouldn’t judge this so much. The technology will improve, it will get there sooner than you know and then you’ll regret not trying it sooner.”

So curious that this keeps happening 🤷‍♀️

#LLMs #work #tech #AIBubble

2️⃣ Inzwischen sind die ersten grossen Wetten eingegangen, dass die KI-Blase demnächst platzt. So hat Michael Burry – der bereits auf das Platzen der US-Immobilienblase 2008 gewettet und gewonnen hatte – jetzt eine Milliarde US-Dollar gewettet, dass die KI-Blase demnächst platzt. Seine Wetten gehen gegen #Nvidia, das in seltsame Kreislaufinvestitionen verwickelt ist, und gegen die umstrittene Big-Data-/Ki-Plattform #Palantir.
#MichaelBurry #DNIP #AIBubble
https://dnip.ch/2025/11/11/dnip-briefing-48-betruegen-fuer-die-ki/#Wetten-dass-%E2%80%A6-die-Blase-platzt

Ich frage mich echt, wann die #Blase platzt:
- Studie vom Media Lab am MIT:
95 Prozent der untersuchten KI-Pilotprojekte in US-amerikanischen Unternehmen führten zu keinerlei messbaren positiven Auswirkungen.
- Untersuchung von McKinsey:
80 Prozent der Firmen, die Generative #KI einsetzten, konnten keine signifikanten Verbesserungen erzielen. Die Hälfte davon gab ihre KI-Projekte daraufhin auf.
- Befragung der Harvard Business Review:
über 40 Prozent der befragten US-Arbeitnehmer berichteten, dass KI-Ergebnisse sich als gute Arbeit „tarnen“, aber wenig Substanz hätten. Wer mit KI-Texten weiterarbeiten müsse, verwendete jede Menge Zeit auf die Korrektur dieses „Workslop“ („Arbeitsabfall“). Autoren warnen: Das zerstöre den erhofften Produktivitätszuwachs. #aibubble
https://taz.de/Studien-belegen-dass-der-Einsatz-von-KI-keine-Effizienzgewinne-in-Unternehmen-bringt/!6125562/