Post #3114
2025-11-13 21:45 UTC
Replies (24)
-
@saraislet@infosec.exchange 2025-11-13 21:48
In other news, I'd like to share that I measured a 100% gain in productivity after turning the lights on for a team that was previously operating entirely in darkness. Though I still need to do some research to understand the Hawthorne effect https://en.wikipedia.org/wiki/Hawthorne_effect
-
@recursive@hachyderm.io 2025-11-13 22:18
@saraislet@infosec.exchange oh, we'll simply measure lines of code written!
-
@lolcat@digipres.club 2025-11-13 22:20
@saraislet@infosec.exchange Coming at this from the systems side of things rather than SW dev, I can report that recent GenAI developments look a lot like the big cloud push in ~2008 - 2017. IT "leadership" at my employer, a giant R1 US university, simply decreed that everything would move to "the cloud." The only assessment plan was simply to assert that cloud was "always better" for everything. While the pitch was presented in tech mumbo jumbo, it was always just a struggle for budgets and power.
-
@glennsills@dotnet.social 2025-11-13 22:26
@saraislet@infosec.exchange Yeah, having people with a degree in business administration make these sorts of decisions is fraught with peril.
-
@JeffGrigg@mastodon.social 2025-11-13 23:12
@saraislet@infosec.exchange "Corporate jargon translator fish" says that "We're going to track increases in productivity that we gain by adopting GenAI" means "We're going to 'gin up' some numbers to 'justify' our massive corporate spend in GenAI."
-
@queenofnewyork@newsie.social 2025-11-13 23:24
@saraislet@infosec.exchange I'm not sure at what point my boss is going to realize the changes aren't going to show up in not going over the budgets he keeps making me whittle down below what I know the work will take.
-
@TrimTab@mastodon.social 2025-11-14 00:04
@saraislet@infosec.exchange Its safe to assume there will be a productivity increase. Because in corporate jargon, "negative increases" exist.... LoL
-
@Npars01@mstdn.social 2025-11-14 00:24
@saraislet@infosec.exchange "...is AI really going to do so fantastically that it's worth paying far more for the book value of these assets, far more for the projected future earnings — not the current earnings — far more for the projected future earnings of these stocks than people have ever asked in the past. So, that's plainly front and center the biggest issue." https://www.bloomberg.com/news/articles/2025-11-13/john-authers-on-economic-impacts-of-ai-valuations-private-credit-cockroaches https://archive.ph/06X1W AI is a flight of capital out of the fossil fuel industry & little to do with productivity 1/
-
@me@social.jlamothe.net 2025-11-14 01:13
@saraislet@infosec.exchange More lines of code mean more productivity, right? Right??? 🫠
-
@chrislw@podcastindex.social 2025-11-14 01:15
@saraislet@infosec.exchange my boss and I built a business use case around using AI features in a CRM platform. We'd only implement the things we know would have benefit to support agents. EVP's first question to us was "What percentage of efficiency or cost savings? ". Because that is something you can't answer. But he is fine encouraging us to "find reasons" to use co-pilot or to approve random low-quality in-house AI projects.🤣🤣🤣
-
@cap_ybarra@beige.party 2025-11-14 02:06
@saraislet@infosec.exchange there is a exploit in the programmer kernel that conflates novelty with productivity and it gets us. every. time.
-
@crankylinuxuser@infosec.exchange 2025-11-14 02:18
@saraislet@infosec.exchange you wouldn't have any pennies at all, cause der fuhrer quit printing them.
-
@semifor@nerdfight.online 2025-11-14 02:31
@saraislet@infosec.exchange @minego@pdx.social Just replace "GenAI" with "incompetent assistant" to make it clear.
-
@ignas@fosstodon.org 2025-11-14 03:36
@saraislet@infosec.exchange If you can objectively measure productivity, you can compare two teams on which one is more productive, which means, you can fire the most unproductive teams! And make the compensation meritocratic! Can I give you a million dollars so we can disrupt the whole Silicon Valley? I need to come up with a good name for the yacht I'm going to buy.
-
@aka_quant_noir@hcommons.social 2025-11-14 03:46
@saraislet@infosec.exchange I'm sure they'll construct a productivity metric that will serve a black box approach to work.
-
@mcc@mastodon.social 2025-11-14 04:34
@saraislet@infosec.exchange "We're going to create some numbers to justify our decision"
-
@nepi@nepi.gay 2025-11-14 06:23
@saraislet@infosec.exchange The regularity with which I have to explain the scientific method to managers is concerning
-
@beandev@social.tchncs.de 2025-11-14 06:40
@saraislet@infosec.exchange Exactly this is what I discuss every day with my managers. Only gut feelings, no real KPI and at the end AI is only in 20% of the whole production line available (initiative to deployed value). On top of, work slop kills a lot of productivity, because I (an others) must argue against fake facts from chat clients. @isotopp@infosec.exchange
-
@michaelvcooper1@mastodon.social 2025-11-14 06:46
@saraislet@infosec.exchange (Meanwhile at GenAI headquarters) "Oh great GenAI, please tell us how we might best take vengance upon Insecurity Princess for this slander upon you." GenAI - "That is a great question! If you will review her first sentence..." (Hours later) https://www.usatoday.com/story/graphics/2025/11/12/us-mint-makes-last-penny/87217730007/
-
@PierricD@mastodon.green 2025-11-14 07:21
@saraislet@infosec.exchange I'd argue that absolute productivity isnt measurable, but relative productivity can be under the right circumstances. A stable product team operating with mature kanban processes should be able to experiment with genai and measure a statistically significant change on their throughput - if there is such a change indeed. But that means the introduction of genai and the measurement come from within, not from above.
-
@krnlg@mastodon.social 2025-11-14 10:21
@saraislet@infosec.exchange My work is full of "we are exploring how we can use AI to help improve X and Y", with the unsaid assumption that of course it *can* help if only they can work out how. So far it seems mostly to be "useful" for churning out large documents that nobody can be bothered to write or read.
-
@Exandra@chaosfem.tw 2025-11-14 14:55
@saraislet@infosec.exchange a workplace I'm familiar with said they were evaluating genAI for coding. I said: great, can y'all tag every ticket where genAI is used in any way for the fix so we can see if they become the root cause of later issues. The resounding answer: No
-
@drmorr@hachyderm.io 2025-11-14 15:32
@saraislet@infosec.exchange this feels like a @hazelweakly@hachyderm.io toot.
-
@bakuninboys@aus.social 2025-11-16 00:26
@saraislet@infosec.exchange "We are going to find the cause for autism and we'll find it by September"