If you meant racism you should have said that. Saying politically and theologically compatible implies a lot more than that. In fact it implies they have to be the same religion, which is being bigoted in itself.

Making sure the parents of your kids friends are compatible politically and theologically sounds incredibly dodgy to me.

I will say this as well: strict parents raise sneaky children

Sugar isn’t a drug lol

Actually we know the system prompt. It doesn’t have “I am a sentient being” anywhere in it. Stop making stuff up.

You are correct. I had to actually look that up.

Reading back over this I think you have me confused with another commentor. I don’t mention anything about IF in the commend you are replying to. Someone else did though.

Yes genetic algorithms are something different. Though they are used sometimes in training or architecting NNs, but not at the scale of modern LLMs.

Fyi you can have all or nothing outputs from a perceptron or other network. It all depends on the activation function. Most LLMs don’t use that kind of activation function, but it is possible. Have you heard of bitnet? They use only one of three states for the neuron output in an LLM. It’s interesting stuff.

Kinda but also no. That’s specifically a dense neural network or MLP. It gets a lot more complicated than that in some cases.

It’s only one type of neural network. A dense MLP. You have sparse neural networks, recurrent neural networks, convolutional neural networks and more!

Not all machine learning is AI. There are plenty of Machine Learning algorithms like Random Forests that are not neural networks. Deep learning would be big neural networks.

To be more specific this is an MLP (Multi-Layer Perceptron). Neural Network is a catch all term that includes other things such as Convolutional Neural Networks (CNNs), Recurrent Neural Networks (RNNs), Diffusion models and of course Transformers.

What you are arguing online is some variant of a Generative Pre-trained Transformer, which do have MLP or MoE layers but that’s only one part of what they are. They also have multi-headed attention mechanisms and embedding + unembedding vectors.

I know all this and wouldn’t call myself a machine learning expert. I just use the things. Though I did once train a simple MLP like the one in the picture. I think it’s quite bad calling yourself a machine learning expert and not knowing all of this stuff and more.

What movie is this?

A techbro? Do you think I work for some big company? I am a PhD student motherfucker.

well that settles it then! you’re apparently such an authority.

I am someone who is paid to research uses and abuses of AI and LLMs in a specific field. So compared to randos on the internet like you, yeah I could be considered an authority. Chances are though you don’t actually care about any of this. You just want an excuse to hate on something you don’t like and don’t understand and blame it for already well established problems. How about instead you actually take some responsibility for the state of your fellow human beings and do something helpful instead of being a Luddite.

I am sure your right. Please tell me where I am wrong. I could always do to learn more about systems engineering.

There is something I never understood about people who talk about scaling. Surely the best way to scale something is simply to have multiple instances with so many users on each one. You can then load balance between them. Why people feel the need to make a single instance scale to the moon I have no idea.

It’s like how you don’t need to worry about MS Word scaling because everyone has a copy on their own machine. You could very much do the same thing for cloud services.

I don’t trust OpenAI and try to avoid using them. That being said they have always been one of the more careful ones regarding safety and alignment.

I also don’t need you or openai to tell me that hallucinations are inevitable. Here have a read of this:

Title: Hallucination is Inevitable: An Innate Limitation of Large Language Models, Author: Xu et al., Date: 2025-02-13, url: http://arxiv.org/abs/2401.11817

Regarding resource usage: this is why open weights models like those made by the Chinese labs or mistral in Europe are better. Much more efficient and frankly more innovative than whatever OpenAI is doing.

Ultimately though you can’t just blame LLMs for people committing suicide. It’s a lazy excuse to avoid addressing real problems like how treats neurodivergent people. The same problems that lead to radicalization including incels and neo nazis. These have all been happening before LLM chatbots took off.

I am sure the terminal IDEs are great. I did used to play around with vim myself, and still use it for editing config files. I have had some success with Jet Brains as well. It’s a solid product.

I don’t really have the energy it takes to configure and learn all the stuff that’s needed for a terminal only setup these days. I guess I am just not as discerning as you are. I might try a ready made solution like LazyVim.

The 50s? Did LLMs exist in the 50s?

This is why safety mechanisms are being put in place, and AIs are being programmed that act less like sycophants.