AI can’t be all that bad. The problem I’m always seeing with AI is a double-edged sword. You have corporations shoving AI in just about everything, treating it like its a cure for cancer and that really rubs people the wrong way. Then, on a more of a society level, you’ve got people who use AI for an assortment of things like making art with AI and still accredit themselves as an artist to people who treat AI like a therapist when it is not advised to.
However, I’ve found some benefits with AI. For example, I’m chatting with ChatGPT on credit cards, because it is something I may lean towards getting into. It’s helping me better understand than most people have tried explaining to me. Simply because it is giving me a more stream-lined response than people just beating the bush.
It’s good as an assistant.
And if ChatGPT made a mistake? How would you know before it’s to late ?
Because all other information on credit cards (or anything else) on the internet available to people eager to learn is 100% accurate, all the time?
I’ll take that over an information tool that lie 30% of the time
That is absolutely the worst excuse possible to shill for big tech that comes with no real guarantees about precision or accuracy.
While there are trustworthy human sources in the Internet, there are no trustworthy LLMs.
I would trust a book written by a specialist over any LLM output, though
100%
The technology itself is novel and cool. Its the complete and utter meltdown of all tech companies into brainless hype machines that is harmful, which is course, is a function of capitalist incentive and the need for the tech industry to come out with some new paradigm shifting innovation every decade. A normal, healthy society would have been able to leverage machine learning and LLM technology where its most useful, like parsing large amounts of data, or running a local instance on your computer to ask a few questions, etc. We wouldn’t see LLMs in every text editor, pencilcase and pair on sneakers but these snake oil salesmen who run the US economy are absolutely desperate for a new paradigm shift so they can keep making exponentially more money.
The thing is, we don’t need to build these datacenters siphoning comically evil amounts of energy from the grid and making personal compute a thing of the past. Average everyday person doesn’t need cloud compute, they can run a local 4b parameter (very, very small) model on their laptop or phone if they need to ask chatgpt to make them a workout routine or to ask them who won the 1918 world series. But these fucking cretins don’t care, that’s not the point, they are in this because it’s a golden ticket to growth city and once they cash their check they don’t give one hot fuck about the human-spirit-stealing-machine they built.
TLDR: our society is broken and that’s why we keep getting the shittiest, lowest-common-denominator version of everything. everything has to suck by definition because that’s the only version that the system we built will allow.
Accurate
deleted by creator
It can also identify/predict cancer tissue early.
Do you mean an LLM or a machine learning model specifically trained for this?
deleted by creator
From what I understand here it’s their own ML model then, cool
Different case, obviously, but I remember reading about an AI which can identify pending-heart-attack from x-rays … and nobody could figure-out what the hell it was judging from…
THAT is brilliant.
Specialized to the degree that it is trustworthy.
I’d be surprised if humans could possibly compete against a properly-done set-of-AI’s, which worked through, in the correct order, all possible diagnostic-reasoning.
Democratizing accurate-diagnosis would be THE medical-revolution the world needs, now.
_ /\ _
I have a friend at work that does a lot of video. He films weddings, music videos etc. and is making a pilot for Netflix. He uses AI to go through all his footage and tag it according to content. E.g. if he needs a clip of birds, he can just search ‘birds’ and it will pull up all relevant footage. Incredibly useful.
I feel like this was a thing before it got put under the banner of AI
This could come in pretty handy for me. What’s he edit on that does this?
It’s a custom app he made using ai vision and xml files I guess. He uses DaVinci resolve though
Is the AI vision local?
Multimodal local models like qwen3.5 are pretty good
Basically you grab screenshots from the video at intervals and ask the model to tag / describe them.
I should probably learn how you link all this stuff up, probably not doing myself a lot of favours ignoring it.
A 20€ plan of Claude Code for a month will teach you all of it.
It’s pretty much just scripting ffmpeg and feeding the screenshots to a local model via an API (or use a CLIP model)
Cheers
that seems like a great system, i wonder how it’s structured.
For every small benefit, there are disastrous mistakes. We shouldn’t discuss one without the other:
https://tech.co/news/list-ai-failures-mistakes-errors
March 2026
- Police used AI facial recognition to arrest a Tennessee woman for crimes committed in a state she says she’s never visited
February 2026
- Health advice given by AI chatbots is frequently wrong, says new study
January 2026
-
Study reveals that fixing AI mistakes takes up to 40% of the time that it saves
-
An AI tool used by ICE to identify applicants with previous law enforcement experience falsely flagged applicants with no such experience, leading to the placement of unqualified recruits in field offices.
December 2025
- AI mistakes clarinet for gun at Florida school
November 2025
-
Google Antigravity deletes entire content of user’s computer drive
-
Report finds AI hallucinations in 490 court filings from the past six months
October 2025
-
Teenager handcuffed after AI mistakes Doritos packet for gun
-
Lawyer submits AI-assisted court filing with fake citations
-
Man follows ChatGPT advice over stopping eating salt, develops rare condition. The man was hospitalized, sectioned, and eventually treated for psychosis. He tried to escape the hospital within 24 hours of being admitted.
-
ChatGPT-5 jailbroken with 24 hours of release
July 2025
-
AI Coding app deletes entire company database
-
McDonald’s AI chatbot error exposes data of 64 million job applicants
-
AI program is tasked with running a small shop, goes insane, claims to be human
-
Apple Intelligence falsely presents BBC headline
… and it just keeps going.
So don’t put AI in front of anything mission critical or without going through a review of a human.
So LLMs in agentic mode are a disaster waiting to happen.
God yes.
If we’re strictly talking about LLMs: Certain accessibility services - MAYBE. Writing closed captions / transcription for the most part requires little “human” touch. If we ASSUME that AI will be able to it reliably one day - because it really can’t yet - that’s one thing that would benefit society.
Image descriptions is another thing I might see done by AI one day but that still requires an understanding of what’s actually important about the image.
I built a system that translates subtitles from English to my native language and it beats cheap-ass “official” translations 9/10
It even gets colloquial terms and phrases right, adapting to the correct song for example - something a human translator working for minimum pay usually won’t bother
Please go watch the yt video of Bernie Sanders discussing politics/society/civilization with Claude.ai
That may blow your mind…
It’s … not quite as limited as you, or I, had been believing…
_ /\ _
I’m a therapist. I use HIPAA compliant AI to generate my (editable) case notes for my sessions now. Not only is it a huge time saver to simply edit a generated note as opposed to making one from scratch, but in many cases it takes more detailed notes, including quotes from clients.
I have heard of other therapists and medical doctors also using AI to help with diagnosing.
The danger is when therapistsdon’t review the content to check for accuracy. Because occasionally it will generate something not really reflective of what the therapist might have been doing, or it might lack detail that the therapist might have otherwise inclused. But more often the stuff it comes up with is surprisingly accurate.And editing is even easier when you can just tell the AI something like, “include more details about how the client noticed their pattern of putting their own feelings last,” and it just does what you asked. You don’t necessarily have to edit manually, though you can.
So how does that work? Do you just have an AI listening throughout the session like a note-taker?
Yes basically, but since it is HIPAA compliant the recording is automatically destroyed when the note is saved. Also no protected recordings are used to teach the AI. The therapist can also choose from a number of different case note formats that might focus on different things
no protected recordings are used to teach the AI
How do you know for certain?
People conflate security with risk mitigation. It’s not secure in the way that you can confirm the data has been deleted. The risk however is mitigated due to vendor attestations reinforced by contracts.
Yep, so you can’t actually know if the recording is destroyed, it’s just contractually required to be destroyed. Big difference in my book.
Wished these sensitive audios would be processed locally and never leave the therapist’s network instead.
I can’t know for certain, as I’m not on the product side of things. But I do know that HIPAA standards are very rigorous and if it were discovered that they were intentionally misleading therapists and clients then it would invite a class action lawsuit that would be insanely large.
I do ask for and document my clients’ consent, though, so if anyone is not comfortable with it that’s fine. I just write the note the old fashioned way. Most are fine but a few have said they don’t want to and it’s not a big deal.
A HIPAA violation is a death sentence to a company, along with massive fines.
There’s no incentive for them to fuck around
I dislike this immensely and actively seek health care providers that don’t use these tools.
My core problem is that I want a professional who engages with me as a human and knows me.
I’m a professional (not in health care) but I “know” all of my clients, and I don’t think that’s an unreasonable expectation for a client or patient. When I pay $100 to talk to a GP for 10 minutes, I don’t think it’s too much to ask for them to have a conversation with me, really truly listen to me, and spend a few minutes writing some notes.
In the case of a mental health professional the time spent after an appointment with a patient is much greater. I don’t really want what I’ve said to be automatically converted to notes for a human to review. I want a human to consider the human to human conversation we have had, in the context of other conversations we have had and the relationship I have with them, and use those insights to produce appropriate documentation.
Finally, I have a strongly held belief that relying on the assistance of gen AI reduces one’s skills and abilities. For example, consider two therapists who have just completed their education and accreditation and start seeing patients. One uses gen AI to produce notes for every patient, the other eschews this practice. Ten years later, which therapist would you really trust to listen to patients and be able to distill the key elements of the conversation both spoken and unspoken?
That said, I’m aware that these services are becoming an industry standard. I suppose they may help therapists see more patients, and in the context of public health that might be a good thing. Whether or not I would use a service like this if I were a therapist is a difficult question to answer. If I were just starting out I think I probably would. That is to say my beef isn’t with you personally using a service like this, more that it’s becoming an industry standard.
I understand those concerns and I think there’s validity. But there’s also enormous potential for benefit.
I know of several therapists who are very good at being present with a client but terrible at documentation. And if one of these has a busy day or two it is easy to get behind. By the time they get around to writing the note the details are very fuzzy. Human memory is notoriously unreliable. A therapist I respect has said that if you’re writing a note 24 hours or more after the session, you’re probably writing fiction. A tool like this has the potential to greatly help the documentation process. But I agree that it should never become a replacement. I thoroughly read all my notes and often make edits to make them more relevant to me.
An attorney I know who specializes in representing therapists and regularly conducts legal and ethics trainings has also said that from a legal standpoint, when comparing human to AI generated notes, the AI notes are usually superior. They contain details like quotes and they automatically include all the stuff that matters for legal or insurance requirements. This attorney is VERY risk averse and honestly I thought she would have been against this, expecting horror stories like artifacts. Her opinion was a factor in me trying it out.
Again, I stress that this is a tool and not a replacement. When I read through a note, I am considering the things my clients said and my interventions to see if it matches up. It’s not perfect but it is very good and I’ve regularly been surprised with how helpful it can be.
Thanks for a considered response. As in all things, there’s nuance and I acknowledge there are benefits.
I’m genuinely curious as to whether you think reliance on this service will diminish someone’s opportunity to build the related skills?
I think that given human nature, there will certainly be some providers who overly rely on it. There are already therapists and other professionals who cut corners where they shouldn’t in a variety of ways. Probably the most common example of this is when therapists write bare-bones notes with practically no useful information to bridge one session to the next. That’s been happening since documentation was a legal requirement.
However, as always, any serious professional is going to take the time to do it right. They will understand how to use a tool effectively while keeping their skills sharp. In my field, with this tool, that would mean every note is read and edited so that it is truly useful. For example, editing the content of the note so that it can be interpreted through the therapist’s theoretical orientation.
I would hope that training programs and continuing education providers emphasize that any note they sign, including one generated by AI, is one that they are still legally responsible for. So it behooves them to always read it thoroughly and check it for accuracy.
With any new tool, certain skills will diminish but new skills will be developed. So writing skills may suffer, but good therapists will be good at editing and using effective prompts to get a good note.
Also, for what it’s worth, documentation skills and intervention skills are very different. I have known a few excellent therapists who were absolute shit at documenting. These therapists tend to be so naturally gifted and intuitive that they don’t need to document very well to be effective. And many therapists write very good notes but are mediocre at the actual therapy. So, at least for now, I tend to see the potential pros as outweighing the potential cons. That could change though!
Running automated hacking and blackmail campaigns against AI companies.
translation is pretty good.
they want to make ai npcs on games, which could be awesome if we can ever reduce the system requirements for running it.
I tried out a game/demo thing that was a tester for AI NPC dialogue. I asked an NPC to tell me about himself and he replied that he could not connect to server lol
There’s that one silly vampire game which uses AI NPCs, I think it’s kind of fun looking from people I saw play it
An amazing use for it in audio engineering is for feedback suppression. The old way to give yourself more headroom required you to sit there and turn up the gain until feedback happens and cut that frequency. Now you just turn on the feedback suppression and it does all that for you on the fly. It’s game changing for live sound, every major venue has it now.
Great for film sound too. You’re filming a rainy scene and the rain is way to loud? You had to get the actors into the studio and do voiceover, now you can often just filter it out.
Oh yeah you’re right! It’s the same for all unwanted noise. Rustling, wind, buzz, ac noise. All of it can be filtered out now! You can even take away the reverb from an untreated room and add in your own reverb. Convolution reverb is amazing, you can actually capture the reverb of any space you want and add it into your recording in post. I honestly don’t know how much an expensive treated room matters over some investment in the plugins that let you do those things.
An example for movies: instead of trying to capture the actors talking inside their helmets for Interstellar, they actually made an IR inside of the helmet itself and added that to the overdubs!
The way you create an IR (impulse response) to capture the reverb of a space is you take a speaker and play a sine wave (or a gunshot/balloon pop,) then record it with a good mic. Then just take that WAV file and put it into a convolution reverb plugin. It sounds identical, the technology is amazing! You can use this to capture all kinds of analog circuitry like guitar amps also, that’s how they make those guitar amp plugins.
feedback suppression has been a thing for ages
it does not require AI, all you need to do is identify the consistent tone and subtract it out
I know it doesn’t need AI for older versions of feedback suppression, but there are newer systems using it that are more effective at dynamically subtracting those frequencies
i have never seen a pro live sound engineer use AI to do this
and i dont think a neural net will have significant improvements over standard ways to do it like the neve 5054 at the same latency
Yeah you know what, I definitely am wrong about this, I totally thought they were using AI but that makes way more sense
The best way to learn is to say something confidently wrong on the Internet haha
I agree there is a lot of annoying hype. However i also agree there are some specific use cases where it can be helpful.
I for one find it handy some times when i am writing bash scripts to do things on my system. I obviously check them before running but it does save time.
Although i do recommend running models locally if possible as it is obviously preferable from a privacy and cost standpoint.rubberducking for those with social anxiety. Also small friction to get surface level answers that normally took digging from multiple sources.
it’s a study monster that initially wiped chegg, duolingo, sparknotes etc. The double edge is that people forgot how to take notes, learn fundamentals to handle complex problems.
Accesibility.
Anything that’s fuzzy and impossible to automate with traditional algorithms, but that also has a reasonably high tolerance for error. It just makes up stuff a good portion of the time, you see.
However, I’ve found some benefits with AI. For example, I’m chatting with ChatGPT on credit cards, because it is something I may lean towards getting into. It’s helping me better understand than most people have tried explaining to me. Simply because it is giving me a more stream-lined response than people just beating the bush.
Watch out, personal finance is not one of those things.
