11–20 of 52 posts
It's funny how generative AI has cast 'numeric' ML into the shadows as traditional ML in such a short space of time.
Just like NNs did to symbolic AI which is sorely missed to build explainable and ethical systems.
I don't think symbolic AI has much to do with ethics? My impression is that ethics in this space are primarily concerned with the application of these technologies, especially when their deficiencies and biases are known.
RLHF is a very important part of the training process. Without it, the LLMs would definitely behave much more similar to a glorified autocomplete than something resembling "intelligence". So the answer is that if you just feed it a bunch of data, it might be good at spitting out some similarly looking data but ultimately has little use. If you manage to get some experts sitting with it and give it instructions and ex…
It's funny how generative AI has cast 'numeric' ML into the shadows as traditional ML in such a short space of time.
Just like NNs did to symbolic AI which is sorely missed to build explainable and ethical systems.
By this I mean there are plenty of scalable, even gigantic, practical uses of NNs popping all around leveraging available data and cheap compute. But not for handdrawn expert systems of old
RLHF is a very important part of the training process. Without it, the LLMs would definitely behave much more similar to a glorified autocomplete than something resembling "intelligence". So the answer is that if you just feed it a bunch of data, it might be good at spitting out some similarly looking data but ultimately has little use. If you manage to get some experts sitting with it and give it instructions and ex…
Isn't RLHF for building _chatbots_ over LLMs (and not part of LLMs themselves)?
Earlier quoted context omitted.
Just like NNs did to symbolic AI which is sorely missed to build explainable and ethical systems.
> ethical systems I don't think symbolic AI has much to do with ethics? My impression is that ethics in this space are primarily concerned with the application of these technologies, especially when their deficiencies and biases are known.
We (I included) have been doing this EXACT thing for a long with with the stock market. We take ALL the data, build a model, then ask for an output buy/sell signal. The idea for using this type of model, from quants, was applied to language, THAT was the new idea. Now it sounds like you are saying "what if" we took a step backwards, but again, we this is exactly how we use these models now, is trained on data.
Deep learning generally brings a lot less to tabular data because the underlying thing being modelled is much simpler compared to language models that effectively are modeling the human mind + culture so there's a lot more to fit the model to.
There are some robotics researchers trying to frame robotics tasks as sequences of action and sensor tokens, giving the model all the data collected by the robot and the actions taken by the robot, so that the model can learn to predict the action tokens based on sensor info. Here's a blog post from a researcher in this field reviewing a relevant paper: https://evjang.com/2023/06/22/robocat.html