have you found a good use case for generative AI?
I know its a bit of a hot topic but I've always seen people (online anyways) are either a hard yes or absolutely no on using AI. There are many types of "AI" that have already been part of technology before this hype, I'm talking about LLMs specifically (ChatGPT, Claude, Gemini, etc...). When this bubble burst its absolutely not going anywhere. I'm wondering if there is case where you've personally used it and found it beneficial (not something you've read or seen somewhere). The ethics of essentially stealing vast amount of data for training without compensation or enshitification of products with "AI" is a whole other topic but there is absolutely no way that the use of the technology itself is not beneficial somehow. Like everything else divisive the truth is definitely somewhere in the middle. I've been using lumo from proton for the last three weeks and its not bad. I've personally found it useful in helping me troubleshoot issues, search or just use it to help with applying for jobs:
- its very good at looking past SEO slop plaguing the internet and it just gets me the information I need. I've tried alternative search engine (mojeek, startpage, searXNG, DDG, Qwant, etc...) Most of them unfortunately aren't very good or are just another way to use google or bing.
- I was having some wifi problem on a pc i was setting up and i couldn't figure out why. i told it exactly what was happening with my computer along with exact specs. It gave gave me some possible reasons and some steps to try and analyze my computer it was very very useful.
- I've been applying for so many jobs and it so exhausting to read hundreds of description see one tiny thing in the middle that disqualifies me so I pass it my resume with links and tell it to compare what i say on my resume and what the job is looking for to see if im a fit. When i find a good job i ask rewriting tips to better focus on what will stand out to a recruiter (or an application filtering system to be real).
I guess what I'm trying to say is it cant all be bad.
It's got lots of uses:
I've used it to help me set up a home server. I can paste text from log files or ask about something not working and it tells me what the problem is. It gets things wrong a lot, but this is the perfect low risk use for AI....for sending me in the right direction when I have no idea why things aren't working. When it's completely wrong, it doesn't really matter.
The real test for AI is: "does it matter when it is completely wrong". If the answer is yes, then that's not a suitable use for AI.
This. I'm a software engineer and I also sometimes use it by providing it a problem and asking it for ideas how to solve them (usually with the addition of "don't write any code", I do that myself, thanks). It gives me a few pointers that I can then follow. Sometimes it's garbage, sometimes it's quite good.
99% garbage.
If you have ever touched C++ you will know that it has godawful errors and not even chatgpt knows what the fuck is happwning
That's why I'm not asking it to give me actual code I should use, but keep it high level. If it then says there are patterns x,y and z that could be usable, I can look it up myself and also write the code itself. Using it to actually write the code is mostly garbage, yes. And in any case you still need to have an idea of what you're doing yourself anyway.
No, I'm not asking it to write code, I'm asking it to interpret the error and point to the actual problem in the code. It just can't...
Solo roleplay. You can make a character and interact. Generate fake conversations etc.
With generative images you can create custom backgrounds, portraits and landscapes instead of having to lookup for them or doing it yourself.
You can also do some interactive story telling that it's kind of fun.
Generating quick test questions over a certain topic. It's another use case I've seen it being quite good at.
Wow, what a cool idea, I never even considered this. Any other suggestions to this idea to add some fun?
If you want it to go unhinged try to get an uncensored llm. Dans PersonalityEngine by bartowski is my current favorite.
Try aidungeon - it does exactly this.
I had no idea this existed, my mind is blown. Looking it up later today.
Yeah I think dialogue for videogame characters so they don't all just repeat the exact same thing again and again would be great.
Works in theory for written dialogue anyway. Spoken would be a bit ropey.
There is a Skyrim mod that does this I believe and it's pretty decent.
Well colour me impressed!
I self host Deepseek R1 and it's been pretty helpful with simple Linux troubleshooting, generating bash commands, and even programming troubleshooting. The thinking feature is pretty cool and I do find myself learning stuff from it.
What took it from gimmick to actual nice to have for me is when my jerry rigged home network broke and wouldn't connect to the internet. Having what is entially an interactive StackOverflow/ServerFault running on a local machine was really helpful.
Running the model locally makes it easier to not overly rely on AI because of the limited token rate.
You self host the full Deepseek R1? What's your hardware?
Also, you might enjoy ![email protected]
No I host the 70b version because I'm limited by my RAM.
Creating low-effort images for ideas that don't warrant effort, like silly jokes.
You know those business books that combine flimsy pop psychology and self help literature with personal development and business goals? Yeah, those books with 300 pages and only one good idea per 100 pages if you’re lucky. Rest of it is just fabricated stories, ideas copied from other books and regurgitation of ideas from the previous chapters to fluff up the page count. Yes, that category!
Well guess what? GPT can generate precisely that level of quality without any effort. In fact, it seems to gravitate towards that style unless you specifically work hard to steer it to aim higher. It has never been easier to become a business book author! Zero editing required. Just prompt and publish.
It feels like this is the one area where GPT truly excels.
I see it as a toy. No different from the Slinky or Silly Putty I had as a kid. Just something to play with.
Inspiration for writing emails, letters, text messages. I always check what the thing wrote though.
Regarding the job application, most companies and sites are using shitty AI to rummage through the piles of resumes they receive.
The whole job application process is frankly one of the worst real world use of most technologies, not only AI
I'd love to have an AI assistant that does shit like call service providers and wait in queue and take care of business for me
There's only a few use cases where I've found I prefer it to doing things the hard way.
However, I try to avoid using it for anything that otherwise requires productive mental effort, because I find that I end up being a lot more informed and capable if I spend 5 minutes going through sites, learning about a topic, identifying wrong answers, and being able to put together better new queries in the first place, than I do if I ask a chatbot, even if it pulls from those same sources.
When you have a chatbot summarize or combine/condense information, you'll always lose nuance and additional context, and very frequently that context will actually be helpful to your overall understanding. There's also many cases where, for example, someone on a forum explains an issue a bit, and their profile has more related information on it that an LLM simply wouldn't go for, only summarizing from their one response on that page. This can lead me down a rabbit hole that then leads me to finding other good sources. Maybe someone mentions that a particular site is helpful for what I'm looking for, and that then becomes something I use more frequently when I do searches for things, whereas an LLM would have just ignored that comment.
You can't steal data, only illegally copy it. The original data holder still has the data, just you do too.
I don't use it for writing code because that's what I love but I use it for documentation and other stuff I hate.....😂
I find it good for music and film suggestions. You feed it a set of ( I want a suggestion like these ) and it provides a good result.
Also good at building mermaid code for diagrams, just tell it write me mermaid code for this, and drop in a descriptive paragraph, then copy paste the code into mermaid.live
That use case became very useful so there is a paid mermaid page to automate that manual process.
I'm very much against AI slop and hate how it's the most prominent use in day to day life.
With that said, I work for a small government contracting company. We are careful about what we bid on, and of course it's not a sure bet that we'll get it. There is a lot of boilerplate stuff in these proposals. When I was on the bench, my boss asked me to help find some AI tools to help with proposal writing.
Honestly? I can see it being used in cases like this. I wish there weren't so much fluff needed in these things, but that's the hand we're dealt. It's not necessarily worth hiring another proposal writer for what we do, and I certainly wouldn't use its output as-is without knowing what you're proposing, but to get some decent starting verbiage, section by section, to be adjusted after? Yeah, I can see that being useful.
Echo this. I work in a similar proposal world which requires too much tailoring and fan fare. Feed in the RFP, load up our USP/methodology, and record a meeting where we talk shit about what the proposal needs to accomplish.
It shortcuts the first 50-60% of the process. But it's helpful to have something to build over, rather than from scratch.
Best (or at least most addictive) use of AI I've seen so far is https://anycrap.shop/
I use it at work for stuff where it would be inefficient for me to pick up entirely new side skills to only be used rarely and sporadically.
For example, I made a spreadsheet tool to compose ordering spreadsheets in Excel for a system at work that needs them. Most of it uses basic macros that you can record with the basic macro recorder in Excel, with no special skill required, but every now and then I need to introduce functionality into it that's far more complex.
Instead of learning obscure VBA coding for something I do once every two months, I can just tell ChatGPT that I have spreadsheet A called this and spreadsheet B called that, assume that they are both open, and write me a macro that does A and then B and then C and then D between them.
It does it in five seconds, I plug the code in, test it, and then go about my day. That's its positive use case for me.
Same here. I also like to use it to save time on things. My work has all my info and their policies anyways so i use it to make meeting minutes, make emails summarizing policies or announcements, finding mutual scheduling times, etc. I can do all these myself but its so much faster.
I only really use AI at work and keep my work off my personal devices.
It's good for rapid output of plausibly human text that can then be sorted or assessed for adequate validity or utility. That's all.
I ask for information sometimes that I cannot find in few minutes of googling (I use a lot of this information in writing fiction). I generate images once in a while for role playing and storywriting. (Not sure if it is AI but) I turn some text over to speech, to listen at my stories.
We have a got at work that’s trained on all our internal docs and handbooks. It’s useful because you can ask it operational stuff like, “how do I request time off?” or “when are performance reviews?”
It’s good for stuff where you might not know who to ask or where to find the answer on internal stuff
I needed the following CSS copied 51 times with a 0.05 s increment, because CSS can't for-loop and didn't want JavaScript:
I know I could've just for-looped it somewhere else and copy-paste the output, but I was curious if DeepSeek could do it.
Giving me a paragraph where Tom Cruise grows increasingly frustrated during his juggalo-themed commercial for shrimp fried rice.
I used to spend days rotoscoping people in videos. Generative infill for background painting and automatic rotoscoping have saved probably a year of my life at this point. Image generation relies on CLIP, which needs a language model for conditioning.
LLMs trained exclusively on documentation and ran locally seem like they'd be nice. Basically the next step in search algorithms. Do note that I am not talking about having an AI summary at the top of every web search page, that's harmful.
That's essentially what I do, I don't have any accounts with ChatGPT or anything, I just run it locally off my laptop. IDK if you get better results with ChatGPT (I'd assume probably) but my local one seems fine for everything I need it for. It's a little slow too, but who cares?
How do you run it locally? I haven't ever tried it. I'm curious to.
Oh sorry, I phrased that kind of badly. I don't run ChatGPT locally, I just meant I run an LLM locally. Usually either Mistral or Qwen. I was using GPT4All until recently, but lately I've been trying out Jan too.
No, you phrased it right, I haven't run any LLMs locally at all.
Ah okay. Well if you grab either of those programs I linked to it's pretty straightforward, you pretty much just choose your model from a list and away you go. You can run them even if you don't have a great GPU, but they'll be slower.
I've tried ChatGPT a few times to see if it's useful for me, and it's worked surprisingly well in most cases.
I made a website that needed two modal images, one on the top and one on the bottom. I wanted them to be enlarged when they were clicked on. I found a load of guides for getting one to work, but I couldn't get both to work. A few minutes with a prompt got it working. It didn't help me to learn JavaScript, but did give me working code that I needed quickly.
I've used it to fluff up some text. I'm not very good at making things sound good in text, so it helped a lot.
The latest one I've tried is getting camera settings for a dark gig setup. I was able to give it an old photo that was under exposed but gave an accurate impression of the room, and ask for recommended settings with the same lens, a new lens, and a flash. It gave me a selection of settings with and without the flash, including settings for rear curtain sync, so when it leaves a ghost trail behind the subject. It's nothing I couldn't figure out, but would have taken a bit more trial and error in the room. I probably wouldn't have thought of the ghost trails.
In short, teaching myself simple stuff.
My take on it is that it's just a tool, and as with most tools you can use it in a sensible way that's positive, although many people choose not to. So as an example, I work in a creative field and I see a lot of people relying on it to do their creative work for them, which I don't really agree with. What I use it for is kind of like an assistant to handle all the admin crap that usually gets in the way of doing creative stuff. So sometimes you have to write form letters, grant things etc. - basically formal stuff that wouldn't require any creative thinking if you did it by yourself anyway, but still eats up time and brain power. I just give that stuff to the AI, make sure it sounds vaguely presentable, and send it off. I could also see a use case for it in areas where I'm weaker like marketing my stuff, maybe for at least coming up with an outline strategy of some sort, although I haven't really tried that out yet.
Essentially, AI will do your creative stuff for you if you let it, or you can just use it to handle the day-to-day piddly crap to free yourself up to do the creative stuff yourself. It's up to you really.
I mean sure a screwdriver is "just a tool", and you can use a screwdriver to pick ice; but it's not the most efficient tool for the job. It was designed for something else entirely, which makes it awkward and finicky to use as an ice pick. This is why so many people use screwdrivers for their intended purpose, rather than manipulating them to fit into a different workflow.
(The intended purpose here is disrupting creative professions, making creative labor cheaper and more efficient in order to maximize capitalists' profit from it. Not to make artists' lives easier or increase their effectiveness as laborers and business owners, although yes it can be used that way and, yes, that is arguably a limited side effect of the original intention.)
I have some Hone Assistant automations, that creates some todos in Habitica for me. These todos are ai generated, so they sound like quests in a rpg 😎 this really motivates me. Also it's funny
Cool, can you please share them?
Sure thing, buddy! However, the actual todos are in german. But I'm sure that doesn't matter, as you want to adapt these to your own cases. The Idea is to gain xp points for daily routines like hanging up the laundry or emptying the dishwasher. Our machines are very old and not-smart. However I noticed, that the Shelly Plus Plug S can also measure the current wattage. You can use this to see if a device is running. I'm just sharing the automations for the washing machine as they're all working the same. I also splitted it up in 2 separate automations to make it easier to debug. This utilizes a boolean helper (can be added in Settings → Devices and integrations → Helpers)
The first one just monitors the usage and sets the bool. Notice the debounce in the triggers section. This is important to avoid false positives:
This one handles the actual notifications and todos. It also adds a second todo 12 hours later to get the clothes off the hanger, fold it and put it in the closet. So you may remove it. It also sends me and my wife a notification on our phones, so the laundry don't get damp in the machine.
Happy gamification, adventurer!
Nice. Ich spreche Deutsch 😅
Was für ein Zufall! 🥳
I took AI courses in college and it was fun to learn about then when it was a bunch of toy examples that showed the potential of these systems, but it was clear enough to anyone in those classes or doing that research how not ready they were for real applications because of all the known flaws with how model training worked. And then some ceos just ignored all that and started blowing up the bubble.
So my answer is the research models that could play video games kinda good. Everything after that was getting ahead of ourselves.
For engineering... Get me a script that calculates the length of a window based on a similar size. Or calculate the tip velocity of a turbine blade given the speed of the gas going into it and the diameter of the turbine. Basically things we would have to take a month to design so as to answer other questions. Cuz nobody pays you to make quick calculation tools.
And then it makes mistakes and you won't be able to tell
No, you don't just get a script and run it blindly! You use your knuckle to figure out if it works first by reading the code and calculating known data as a test.
You can't even rely on AI to have real formulas for the area of a circle. You have to rely on your own knowledge and on books to confirm if the code is doing what you need it to do.
What AI does is it shortens the code creation time to just a few seconds vs days of coding... Because engineers are the best back seat coders I know. Once there's good code they can move mountains. But confronted with a blank page they freeze.
Most if the time is spent fixing AI code, and writing prompts. I don't think you save any time.
Also have fun in integration and deduplication hell.
This doesn't sound like the hardest thing to write a program for especially if you have the gibbity help you write it and quadruple check its output.
Or, you know, you could hire a programmer.
Ha! With no budget?
• I use it for research and then verify its findings.
• It’s excellent at summarizing quickly.
• It’s great at idea creation for specific needs and outlining it.
• While it’s good at writing I enjoy that and do it myself to keep my specific tone.
I used it to extract thousands of rows from tables in PDFs and generate enumerations for them in various programming languages. I had to do some pre-processing with the Python script, and review all the output to make sure it didn't screw up, but it saved me a lot of time.
For image gen I don’t have a good use but it is very complex which means sometimes I can just lose an hour or 2 fumbling around in a complex network of nodes.
What I found fascinating was how strangely good the results were when I created an image, then fed the result back as input, and repeated that process.
The only useful thing I used image gen for was creating references for an artist to create a PFP for me that looks rad as hell.
As for LLMs, also not really. I think about 90% of the time LLMs either give a useless or just wrong answer. I can’t seem to find the thing that LLMs are supposed to be good for. One thing all LLMs I tried have failed consistently at was finding a movie from a vague description I gave.
bash scripting, light python programs, fixing software and hardware issues, learning Japanese, learning other things, writing applications, all the boring stuff and stuff need extreme repetition or data mining. Ideas comes from me, work is done by machine.
It has been great at estimating calories in things. When possible, I compared to the actual food label, and it's usually within a reasonable margin of error. But not everything is labeled, and when you're on a diet it's better to have at least something to go off of.
A couple so far. I have a local copy of Stable Diffusion. It's handy for upscaling some kinds of images. I'll also use it to flesh out my worldbuilding project with landscapes and scenes.
Less often, I'll consult ChatGPT, without logging in. For the times when a search engine doesn't cut it but a forum post would be too much. I'm usually skeptical of AI summaries, but I find it justified for boiling down poorly-written stuff that I have to read, but isn't worth my time in long form.
That stupid song about the Windows registry. Don’t look it up, or at least don’t come crying to me after listening to it 50 times. It has no right being as catchy as it is. I fear for younger people and how God generated music will sound in a year or two.
I used to ask why classical music wasn’t performed by Metal bands. Now I can have AI make it for me and it shouldn’t even be that hard given the songs are already written. Now if I were to say “compose a new symphony in the style of Beethoven and make it sound like Iron Maiden are playing it… that would be different.
I actually just view it as the latest abstraction of search. Yahoo in the 90's did not give you a blurb summary of links or would do math equations for you or give answers to simple questions like what time is it or whats the capital of alaska.
I think there are many thousands of folk in fields beyond IT that use it all the time. It's by no means perfect, but for many of us managing teams or doing boring AF admin, working with procurement, writing user documentation or trying to navigate basic system configs then it's immensely useful.
I have a workflow that translates English srt files to my desired target language. It’s a great use case bc LLMs are proficient at picking up the nuances of language translations, especially related to idioms and the like.
I use notebooklm to act as my research assistant when it comes to labour law and union contract. Claude has helped significantly with compiling the chapters for my upcoming book. Deepseek has created several courses for me to further refine my skills and education.
I've been using it recently for generating alt text for images (my bots on Mastodon and Aunty Madge on ![email protected] specifically). It's pretty good at that, although does sometimes give weirdly wrong details - especially the TED Music Bot, if it gets the usual +4 startup screen it says it's +4 on key F1, instead of 3-plus-1, and tells me the wrong colours for the text and background (I think it may be getting it confused with the C64, bit the colours are right there on the image!). It's infinitely preferable to having no alt text, which would be the alternative.
The other thing it's really good at is summarising articles.
I've also used it when I've had an error in my code I can't track down, or a bracket missing that I can't figure out. It quite often gives nonsense but I've had some success. Usually a normal web search is perfectly adequate though!
I like to use it for stories.
Best use for me has been feeding it logs and pcaps, incorporating it into Wireshark sounds far more useful to me than summarizing a two-line email.
As a teacher, it's good at making assignments for students. Though I don't use it out of principal.
I had ai write a python script for me that checked a directory for zip/rar files that don't also have video files, unzip them and if it fails retry at a doubling interval. It would have taken me an hour or so, took the ai 5min+15min of me fixing it. I also had it add logging which is definitely something I wouldn't have been arsed to add otherwise.
It isnt very helpful at my day job writing professional code though, and I dont think someone who didn't already know how to do it themselves could make use of it.
I find it's good as a "search engine of last resort". I would pay $2 monthly at most for one, and less if it's American.
I only really use LLMs for project ideas, naming things, or recipes, but it’s great for those. For recipes especially, trusting the gaslight machine adds an extra layer of suspense and fun to cooking or baking.
It’s also fun, but not useful, to ask Linux questions - chatgpt specifically told me to restart dbus while in an active GUI session when I asked it about a simple case of unintended behaviour. It’s probably the dumbest Linux advice I’ve heard to date, so I got a lot of enjoyment out of trying it to see how bad it would fuck things up
Despite claims to the contrary ChatGPT is very accurate with its responses when such responses involve web searches. Where it falls apart is complex multi step things like coding questions.
I make heavy use of it to skip past all the clutter of Google search results and end up with clear summaries that answer my questions. That's all I'd really use it for as anything more than that its output is highly variable in quality.