lol, "if only someone had warned us about this sort of thing?!"
-
@glent@aus.social @MugsysRapSheet @Lazarou No. that would be the "normalized answer" and not necessarily correct at all. Just "average".
-
-
@Lazarou "double checking" and "accident".
EPIC.
seems like today double checking is some kind of failure
-
@Lazarou meanwhile my ceo
@agasramirez @Lazarou all CEOs went to the same "cool kids" CEO school.
-
-
@agasramirez oh wow, I've heard talk of this attitude but to actually see it in action!
@Lazarou As a CEO (which I am not) I would be happy, if there were still persons in my company, who are not totally relying on LLMs.
@agasramirez -
@Lazarou I have my doubts about this post.
I do not doubt that AI can and will produce fake insights. A very easy way to get that is to ask an AI questions whose answers are not in the numbers you’re feeding it. Or your questions are heavily biased towards things you want to hear („Show me how campaign X increased sale“).
But unless you’re completely ignorant of your own business, you will notice it quickly. Especially if it goes on for 3 months. I deal with a lot of managers and every single one of them looks through raw data. Every system (with or without AI) is faulty and they fine comb every bit that influences their salary relevant metrics.
-
@Lazarou I don't get how LLMs are needed for analytics? This field is basically about counting numbers, statistics magic and visualization. Isn't it?
I mean, asking an LLM what kind of statistical sorcery would be needed to measure this and that and ask it for a small python script will probably work (if you double check everything). But what did they do, to come to a situation where an LLM makes up data for months?
-
@Lazarou Completely deserved
-
-
@Lazarou
I wish they'd stop calling it "#AI".It isn't. Not even close. I went to college to study #AI decades ago.
What they are calling "AI" today is nothing more than "deep database scrubbing".
It *assumes* an answer is correct based simply on the number of results it finds supporting that conclusion.
#GIGO: Garbage In; Garbage Out.
@MugsysRapSheet @Lazarou Garbage in Garbage out is such an old principle in computing and yet it's only now we're really seeing it's effects.
-
@Lazarou As a CEO (which I am not) I would be happy, if there were still persons in my company, who are not totally relying on LLMs.
@agasramirez -
@agasramirez Sounds like a "decision " made by an LLM.
@Lazarou -
@Lazarou I don't get how LLMs are needed for analytics? This field is basically about counting numbers, statistics magic and visualization. Isn't it?
I mean, asking an LLM what kind of statistical sorcery would be needed to measure this and that and ask it for a small python script will probably work (if you double check everything). But what did they do, to come to a situation where an LLM makes up data for months?
@flxtr @Lazarou I've seen a ton of analysis reports done by agents and just trusted with little or no verification, looking for wrong database lock strategy, to reasons for errors in logs to data quality issues. It would not surprise me at all that someone pointed an agent at a database or something like crystal reports or even just a bunch of spreadsheets and asked it to summarise them based on a list of KPI, the results would genuinely be that bad without a bunch of hard work.
-
@Lazarou meanwhile my ceo
@agasramirez @Lazarou If everybody should be using AI, it seems weird that he suggests meeting one-on-one to discuss it.
I recently went to a Microsoft seminar for work. They were full-on “everything is AI now” which I had no interest in. So I found the only breakout sessions WITHOUT AI.
What I found was a much smaller group of people attending and the people leading the sessions actually knew WTF they were doing and I got a lot of one-on-one time with the experts.
-
@Lazarou Darwin Business Award
-
-
@peterbosch66 @Lazarou missed a trick here. In the last panel it should've been "Eleventy Nine" studies
