ROBOT
Lazarou Monkey Terror ๐๐๐lol, "if only someone had warned us about this sort of thing?!"
Jacqueline@Lazarou I drink business owner's tears
Agas Ramirez ๐๏ธ
@agasramirez oh wow, I've heard talk of this attitude but to actually see it in action!
trusty falxter ๐ง 
@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
@agasramirez Sounds like a "decision " made by an LLM. ๐ @Lazarou
fcat@agasramirez @Lazarou To be fair this argument can be valid to some extent.
Though I feel most of "AI" is overhyped and will just end up locking down users/companies into big tech providing inference for them...
Chaotic Unicorn@agasramirez @Lazarou all CEOs went to the same "cool kids" CEO school.
\u2014lfr@alter_unicorn spoiler alert, all they ever learned there was malignant narcissism, effectively @agasramirez @Lazarou
SouprMatt@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 โeverything is AI nowโ 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. ๐
sidereal@agasramirez @Lazarou if these things actually worked as well as advertised, they wouldnโt need people boosting them.
Lightfighter@agasramirez Translation: I've spent so much money on this, if I can't show ROI, I'm toast.
Laurent Bercot@agasramirez @Lazarou He sounds like an AI already. He's probably entirely replaceable by an AI.
Khaleer@agasramirez @Lazarou I am sorry to say that, but your boss is ~4 years late to the party.
Trump & Epstein: Biz partners.@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.
staringatclouds
jbaggs@MugsysRapSheet @Lazarou AI has always been a marketing term. What was "machine learning" back when you were probably studying bares little resemblance to Large Language models, but it all get lumped in the same bucket.
Even when criticizing it, we're encouraged to use terms like "hallucinations" that anthropomorphise the systems, instead of using more correct terms like "statistical error".
Glen T, heated, not stirred@jbaggs @MugsysRapSheet @Lazarou It does not at all meet the definition of statistical error, a jargon term in statistics practice. If so we could ask the AI a hundred times and average the results for the correct answer.
@[email protected] @MugsysRapSheet @Lazarou No. that would be the "normalized answer" and not necessarily correct at all. Just "average".
Nรถmenlลony@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.
์ง์ง แ แฑแแแจ Daniel ้ปๆณๅฎ CyReVolt@nomenloony but it scales well โ๏ธ๐ฅบ ๐
Oblomov@MugsysRapSheet Markov chains, all the way down.
Ben ToddYep, now I see any app that boasts "powered by AI" I have to assume it's using an LLM and stay away from it.
Crystal_Fish_Caves@Lazarou I don't like that google keeps using AI to answer my questions, but I have not bothered to look elsewhere. BUT I run several aquariums since covid and one suddenly turned cloudy and green so I asked and was told, "The Water is too Fat"
oh well that solves everything so I get it. That is frightening.
CathyBikesBook@Lazarou I'm sorry but this is hilarious ๐๐๐๐
Moss (the ฯฮตฯie) <(fsm)>{{@Lazarou
So โฆ whatโs the problem? It functions exactly as designed!
So โฆ whatโs the problem? It functions exactly as designed!
su_liam@Lazarou โWe ran into some trouble setting up an office in West Fobispa. Turned out there was no such place. The whole state of Warmington was a hallucination. That was half of our profits the last fiscal year!โ
@su_liam such incredible faith they put into a machine that they wouldn't put into another human being
Milcom MiasmaI know this probably makes me a terrible person but I honestly can't _wait_ for the lawsuits.
Added bonus laughs if they use AI to write their legal briefs.
Peter Bosch
n1ck@peterbosch66 @Lazarou missed a trick here. In the last panel it should've been "Eleventy Nine" studies
@Lazarou "double checking" and "accident".
EPIC.
seems like today double checking is some kind of failure ๐
EPIC.
seems like today double checking is some kind of failure ๐
Martin Seeger@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 youโre feeding it. Or your questions are heavily biased towards things you want to hear (โShow me how campaign X increased saleโ).
But unless youโre 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? ๐ฌ
Hashbangperl@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.
KathmanduPeople love simple interfaces. People have given up good coffee and good music in favor of lower-quality but single-button-push services.
Asking the LLM is easier than doing the work, even routine work you know how to do.
If the results aren't part of an immediate feedback loop, it could take a while for the results to stop *looking* plausible, or for reality to smack them on the forehead.
Gambero
Andi H@AH_99 I just assume the Regime does, it lies about everything else...
Leonid (fuck USA ๐๐ป)@Lazarou Darwin Business Award ๐ฅ
PaulNickson ๐๏ธ๐๏ธ๐ผ@PaulNickson @User47 lol, let's hope so!
JL Johnson 
ใใใชใ็ใใ@Lazarou IF true lol this is what this company gets. But this is Reddit. This smells like that one post about the food delivery who claimed to have uncovered some massive details of fraud. Turns out that whole post was made by AI.
3 months? Relying on AI for something as critical sounding as this? Hmmm lol. I hope it ainโt true lol ๐
@sayonaraminasan it, um, "sounds plausible" xD
@CyReVolt yeah, itโs kind of sad that this is the world we live in now lol
Dataless
Michal Bryxรญ@Lazarou To be fair: Some people do work in analytics departments, but are forbid to come with โnumber go downโ news, so in many cases stats are โenhancedโ or โinterpretedโ in a way that makes it look nice.
@MichalBryxi Yea but that is usually toward investors, not stakeholders making business decisions - unless your company is really messed up (however many may be ๐ซ ).
@CyReVolt Well I have some bad news for you, I'm afraid...
Afficionado@Lazarou hallucination was not the issue here, it was the user who did not bother to provide context
@fermentedviewer rather whoever didn't verify whatever code came out of that agent
Don Hawkins - W7DAH@Lazarou Thatโs great!