cross-posted from: https://lemmy.blahaj.zone/post/46123456
That’s how I felt waking up this morning (as of posting). Who else can relate?
cross-posted from: https://lemmy.blahaj.zone/post/46123456
That’s how I felt waking up this morning (as of posting). Who else can relate?
Obviously it’s for gen ai. Why do ai boosters always conflate this!! You started this thread about “generative ai”, and then when questioned gave a bunch of non-llm use cases, followed by a quick edit and ‘oh yeah it can produce text’.
Every ai pusher does this. You ask why chatbots are useful and they basically accuse you of being anti-science, as if they believe talking to a chatbot should also be included as ‘doing science’. You are a consumer, not a researcher.
I am doing a PhD in computational neuroscience. We use LLMs a lot. Helps write code that I don’t want to bother write and find some bugs due to human error. You sure make a lot of assumptions about strangers online.
Yeah, sure, it can produce text. That’s not my issue with your previous comment. The issue is you bring up “generative ai” and then list a bunch of machine learning applications that have nothing to do with chatbots.
Machine learning is part of AI. I replied with multiple examples because the post and the reply asked about how to use AI non greedily and not generative AI specifically. Generative AI isn’t just chatbots.
So you just completely ignore all context? The fact that you’re talking about generative ai doesn’t give you any fucking context clues? Who is arguing against machine learning?? Nobody! Who is fighting to get rid of disease modeling? Nobody!! These are fake arguments that only exist to distract from the very very limited potential of chatbots, and yes, generative ai is just chatbots.
Generative AI are models that generate data from a learned distribution. Some weather models involve generative AI as in generating weather trajectories, you also use it often in behavioral models where you want to generate animal like behavior from observing how actual animals behave. There are plenty of examples, you just want to be part of the hate bandwagon.
No, absolutely false. Generative ai is literally generating text and images or even videos from a prompt. If you conflate everything it’s very easy to dismiss criticism, but if you actually contend with anti-ai arguments you will find they all use this definition.
Here’s the Wikipedia definition:
What part of it says it is just chatbots? Notice that of says “other forms of data” there? That is because audio, video, and chatbots are the most common cases, not the only cases. Prompts are not even needed. You just have a bunch of data points forming a distribution, and you want a model that samples faithfully from that distribution.
I agree that the word “AI” and even “generative AI” is being used to basically describe big tech that pushes their products on people like OpenAI and Anthropic, but you gotta understand that relying on context is bound to create misunderstanding. Context doesn’t travel along like words do.
Do you not see how incredibly broad your definition is compared to wikis or mine? I agree that wiki definitions often use qualifiers in order to avoid making absolute statements, but wiki and I absolutely do agree about this.
I am struggling to take your claim of simple miscommunication seriously. The context of the gen ai debate that’s been raging for 3+ years is not in flux: we have all been saying the same things about the same chatbots, and the idea that anti-ai people just hate medicine or weather or animals or whatever is just… deeply disrespectful and annoying. You can at least just learn what we’re opposed to, before defending ai from attacks and arguments that were never made and would be silly to try to make.
I thought you were going to say something smart about use of machine learning algorithms, but you went with using LLMs to code for you instead.
That is one of the main applications of LLMs, so what? If you want more specific applications tied to my field you can look at behavioral modeling.