|
webmeister posted:Nah these are PowerPoint slides, mostly with a headline, 1-2 sentences of commentary, then usually 1-3 data points (graphs, charts or small tables). Theyre all in ppt format, no images of charts or anything (which, yeah I wouldnt expect it to understand). I work in marketing, if that helps - these arent detailed technical slides. Ah, Ok, cool. Sorry to be harsh. To be honest I have no idea how good Copilot is for anything really. I also have no idea if it can actually read a powerpoint file in a meaningful way. I know that MS has been shoehorning AI into everything lately with very mixed and disappointing results. For something like that you might want to try and extract them as images and feed them into GPT4 directly or Claude 3 which for a few messages is free.
|
#
?
Apr 8, 2024 03:42
|
|
|
|
| # ? May 17, 2024 16:47 |
|
|
How much can you rely on LLM summarization for any business task of importance given how LLMs work? Seems like a tool that is really only good for idk bullshit could be fine stuff.
|
|
#
?
Apr 8, 2024 03:44
|
|
|
Insanite posted:How much can you rely on LLM summarization for any business task of importance given how LLMs work? Seems like a tool that is really only good for idk bullshit could be fine stuff. Depends on the model, the task and how important it is. You should never completely trust it especially when outside of its context window or trying to extrapolate, it is not a database nor does it think. I always verify the gist at the very least, treat it like a student helper. Always prompt for locations of what it's referencing and check them when working with larger document. If you work within its limitations and understand them it becomes a powerful tool. Broad stroke summarization is pretty good with Opus.
|
|
#
?
Apr 8, 2024 03:51
|
|
|
Tarkus posted:Ah, Ok, cool. Sorry to be harsh. To be honest I have no idea how good Copilot is for anything really. I also have no idea if it can actually read a powerpoint file in a meaningful way. I know that MS has been shoehorning AI into everything lately with very mixed and disappointing results. For something like that you might want to try and extract them as images and feed them into GPT4 directly or Claude 3 which for a few messages is free. All good, it was a semi-vague shitpost anyway  We do use specific AI tools in related circumstances, like here is 10,000 customer feedback comments, divide them into positive/negative/neutral groups and identify the main themes, and itll give fairly decent results.
|
|
#
?
Apr 8, 2024 05:49
|
|
|
Blitter posted:Congratulations on being the kind of credulous moron that assumes that a) you actually understand these topics b) the LLM is reliable and correct. That's why, like some other Goon said many pages ago. You handle the INPUT as well as the output. The more you know about a subject, the better the results turn out because you are able to tailor the specifics to create output that you know to be correct. LLMs are completely useless if you don't already know what you're talking about what you're talking about, but they can save huge amounts of time for people who are already knowledgeable in a field.
|
|
#
?
Apr 8, 2024 07:32
|
|
|
Yesterday, i used chat gpt to make an animated intro for a logo on an oled screen that was running on an Arduino board. Years back, i struggled to do this due to the line by line coding that was needed to create and animate it but chat gpt just banged it out. I'm not sure if this is prophetic of the apocalypse, but it was handy.
|
|
#
?
Apr 8, 2024 14:09
|
|
|
Insanite posted:How much can you rely on LLM summarization for any business task of importance given how LLMs work? Seems like a tool that is really only good for idk bullshit could be fine stuff. Speaking on this I work pretty deep in data science and Tableau recently demoed their upcoming AI tool aimed at non analysts to quickly spin up charts and analysis and its going to gently caress up so many businesses lol
|
|
#
?
Apr 8, 2024 14:13
|
|
|
Like in a way that takes away jobs or just bait for lazy people?
|
|
#
?
Apr 8, 2024 14:20
|
|
|
Rad-daddio posted:Like in a way that takes away jobs or just bait for lazy people? In that it will produce output that seems plausible to the clueless and the c-suites (i repeat myself) but is in fact a meaningless turd.
|
|
#
?
Apr 8, 2024 14:31
|
|
|
Nice
|
|
#
?
Apr 8, 2024 14:32
|
|
|
Rad-daddio posted:Nice 
|
|
#
?
Apr 9, 2024 03:28
|
|
|
It makes higher ups salivate because now they think they could just feed their barely coherent ramblings into something that generates revenue without having to pay for engineers or r&d. Capitalism without the cost (other than human cost, but that's a depreciating asset). When you get shitheads like Besos making small towns into essentially company towns for Amazon warehouses coupled with the unofficial official policy of firing people after a few years because maximum efficiency, it's clear what the end goal is.
|
|
#
?
Apr 9, 2024 14:33
|
|
|
Houle posted:It makes higher ups salivate because now they think they could just feed their barely coherent ramblings into something that generates revenue without having to pay for engineers or r&d. Capitalism without the cost (other than human cost, but that's a depreciating asset). Its amazing how the people that will argue against this assessment are usually people who are quick to tell you its a dog eat dog world and blah blah free market. As if its delusional to point out that the ultimate logical conclusion of unhinged plutocracy is for one dog to finally become the only and biggest and therefore be able to eat the entire world
|
|
#
?
Apr 9, 2024 15:36
|
|
|
|
|
#
?
Apr 9, 2024 16:23
|
|
|
I just want one megacorp to go down in flames in a way that makes it extremely obvious the entire downfall was due to some dipshit saying lets use ai! Just one example. One cautionary tale thatll make CEOs think twice.
|
|
#
?
Apr 9, 2024 17:31
|
|
|
Gutcruncher posted:I just want one megacorp to go down in flames in a way that makes it extremely obvious the entire downfall was due to some dipshit saying lets use ai! Sorry but that's not going to happen because AI is really good and rich people never make mistakes.
|
|
#
?
Apr 9, 2024 18:26
|
|
|
Duck and Cover posted:Sorry but that's not going to happen because AI is really good and rich people never make mistakes. That's not true.
|
|
#
?
Apr 9, 2024 18:34
|
|
|
Gutcruncher posted:I just want one megacorp to go down in flames in a way that makes it extremely obvious the entire downfall was due to some dipshit saying lets use ai! This has already publicly happened multiple times and the result was that nobody in a position to gain from this knowledge cared so it will keep happening with bigger and bigger consequences met with greater and greater apathy Hey on a completely unrelated note is cryptocurrency gone yet? See
|
|
#
?
Apr 9, 2024 18:58
|
|
|
Ignore_Me posted:This has already publicly happened multiple times and the result was that nobody in a position to gain from this knowledge cared so it will keep happening with bigger and bigger consequences met with greater and greater apathy I mean pre-existing megacorp, lets say Amazon. I dont mean corporation that sprung up exclusively for a cryptoscam Like a corporation that makes a bajillion dollars a day just out of nowhere vanishing because some dipshit says lets have chatgpt run customer service. What could go wrong??
|
|
#
?
Apr 9, 2024 18:59
|
|
|
Gutcruncher posted:I mean pre-existing megacorp, lets say Amazon. I dont mean corporation that sprung up exclusively for a cryptoscam You want an existing company that became a crypto scam? oh that's MSTR.
|
|
#
?
Apr 9, 2024 19:13
|
|
|
Gutcruncher posted:I mean pre-existing megacorp, lets say Amazon. I dont mean corporation that sprung up exclusively for a cryptoscam I see. That definitely shrinks the sample size. I only have examples of it happening to preexisting companies that make majillions of dollars not bajillions. For example the Kentucky JCPS school system contracted out their bus route optimization to a company that then decided to plug it into an automation tool and it ground the entire school system to a halt and caused chaos all over the city. Nobody gave a poo poo besides the common person watching in horror. quote:just out of nowhere vanishing because some dipshit says lets have chatgpt run customer service. What could go wrong?? Hasnt happened yet but it will and nobody will care enough to stop doing it until an entire economy gets crumpled by it. I seem to remember speculation that one of the major scheduling database crashes (last year or year before) from one of the major airlines that completely hosed US domestic air travel for several days was due to attempts to automate the maintenance of its decrepit software system by using essentially an AI chatbot
|
|
#
?
Apr 9, 2024 19:17
|
|
|
Gutcruncher posted:Like a corporation that makes a bajillion dollars a day just out of nowhere vanishing because some dipshit says lets have chatgpt run customer service. What could go wrong?? When a corporation is that size it's really hard for even the biggest dipshit to make a big enough mistake to kill it quickly. Like, Boeing recently decided to stop bothering with the whole "fully assembled planes that can fly" thing that is kinda their main business. Somehow this won't kill them and people are still buying their planes.
|
|
#
?
Apr 9, 2024 19:21
|
|
|
Houle posted:It makes higher ups salivate because now they think they could just feed their barely coherent ramblings into something that generates revenue without having to pay for engineers or r&d. Capitalism without the cost (other than human cost, but that's a depreciating asset). There's already a whole lot of "ideas people" churning out AI-generated "ideas boards" and shoving them on production departments and going "Make it like that!" Here's an opera company from LA which was bragging about bringing the director's "vision" and "unique perspective" to life https://twitter.com/secondaryjess/status/1744921666788958568 https://twitter.com/secondaryjess/status/1744923037546270721   https://www.facebook.com/100079892251340/posts/pfbid02WjPNpAgoLaczPAZZAdsA1ijVnLkqB9UP9cCityvK3Eg4aTFsjjmsQYefkuydzSygl/
|
|
#
?
Apr 9, 2024 19:25
|
|
|
Dunno if it's been posted yet, but worth repeating: https://www.wheresyoured.at/bubble-trouble/ LLM-style AI is already at it's peak usefulness. All the information has been consumed, and the output is already getting worse. Everything from here is based on vague promises of 'synthetic data' and computing breakthroughs that will not exist.
|
|
#
?
Apr 9, 2024 19:27
|
|
|
Get his rear end, Jess!
|
|
#
?
Apr 9, 2024 19:53
|
|
|
Failson posted:Dunno if it's been posted yet, but worth repeating: Good article. My thought is what most people are thinking/seeing is 'wow AI has come so far so fast, if progress continues at this rate then maybe we can have real AI soon'. Which ofc is misinformed for the reasons he describes.
|
|
#
?
Apr 9, 2024 20:41
|
|
|
Gutcruncher posted:I just want one megacorp to go down in flames in a way that makes it extremely obvious the entire downfall was due to some dipshit saying lets use ai! Either the megacorp is so large that they get 'too big to fail'd and bailed out by the government, or its small enough that people will say it's poor management, hence why that corporation wasn't even bigger.
|
|
#
?
Apr 9, 2024 22:20
|
|
|
Away all Goats posted:Either the megacorp is so large that they get 'too big to fail'd and bailed out by the government, or its small enough that people will say it's poor management, hence why that corporation wasn't even bigger. This. And even if it does fail it is eaten up by the even larger company. Company man in YouTube covers a lot of these tales.
|
|
#
?
Apr 9, 2024 23:16
|
|
|
Stumbled across a video from a guy I watch sometimes. He tried to replicate that Devin developer AI video that's been going around, to poor result. https://www.youtube.com/watch?v=tNmgmwEtoWE
|
|
#
?
Apr 10, 2024 01:53
|
|
|
Failson posted:Dunno if it's been posted yet, but worth repeating: this article is like a cool drink of water on a hot day for me, thanks.
|
|
#
?
Apr 10, 2024 02:32
|
|
|
It's funny that there's just some underlying foundational and structural issues with how these LLMs work, and if you read the white papers they even acknowledge them as issues that they hope will be solved. Like, later. By someone else. The mad dash to pushing AI to everything is they need to make their money now, because the "next step" in AI is essentially rebuilding from scratch, because these are issues no one can fix, because they struggle to identify why it even happens. It's inherent. But you can't sell what you currently have if you acknowledge that, so they rely on a nebulous "it will definitely be fixed soon" to any issues.
|
|
#
?
Apr 10, 2024 04:38
|
|
|
https://twitter.com/AngstHarrier/status/1777953128484384773
|
|
#
?
Apr 10, 2024 07:56
|
|
|
syntaxfunction posted:It's funny that there's just some underlying foundational and structural issues with how these LLMs work, and if you read the white papers they even acknowledge them as issues that they hope will be solved. Like, later. By someone else. I've seen governments do that for their longterm global warming goals, such as the Australian government's 2021 scheme for achieving zero emissions by 2050 where 70% of the planned emissions cuts will apparently happen via technology breakthroughs which don't currently exist. Job well done, let's all gently caress off down the pub 
|
|
#
?
Apr 10, 2024 08:08
|
|
|
Most AI stuff sucks poo poo but this made me lol heartily
|
|
#
?
Apr 10, 2024 08:32
|
|
|
syntaxfunction posted:It's funny that there's just some underlying foundational and structural issues with how these LLMs work, and if you read the white papers they even acknowledge them as issues that they hope will be solved. Like, later. By someone else. If you want to replace all human artifice, or even just a lot of writing and art, with a robot whirring in the closet you're gonna be waiting for a completely different technological breakthrough. Otherwise there are surmountable engineering and business process problems the people selling the stuff will gloss over but are somewhat surmountable: Public consumption models must be corralled with extra training and just hard exits. The corral will be broken, or the corral will stump the robot at which point a human still needs to step in. Private consumption models must always be reviewed by a human. There's also still very practical business steps to take with LLM development at the existing tech level like making ones that work as generally as GPT or Midjourney that aren't predicated on wide scale IP theft.
|
|
#
?
Apr 10, 2024 12:10
|
|
|
Lmfao Finally a good use of AI
|
|
#
?
Apr 10, 2024 12:21
|
|
|
I find most of these ai pictures have the same sort of old rubberized plastic look to them. Something feels off in that they look sort of too chunky in some areas but not in the way that most of us are chunky.
|
|
#
?
Apr 10, 2024 17:13
|
|
|
Elukka posted:Yeah, that makes it apparent how incredibly unreliable it is. It will regularly fail at Wikipedia-tier knowledge that is right there in the training set and confidently hallucinate absolute nonsense. But it looks believable if you don't actually know the subject. Besides being fed AI nonsense directly, we're getting it second hand by people who take whatever it outputs at face value. It seems like every dance party or conference or basically any event now has some statement that goes like "Immerse yourself in the _______ of ________. Whether you ______, _______, or just want to __________, it's all available to you at ____________" It would bother me less and I probably wouldn't notice it if it wasn't phrased so annoyingly twee
|
|
#
?
Apr 10, 2024 17:37
|
|
|
https://www.youtube.com/watch?v=ys8-HwC77oU
|
|
#
?
Apr 10, 2024 17:44
|
|
|
|
| # ? May 17, 2024 16:47 |
|
|
AEMINAL posted:Lmfao I glued my balls to my butt again is better IMO
|
|
#
?
Apr 10, 2024 19:20
|
|
