Very concerned with this direction of training
“counterfactual events such as whether the Waymo Driver could have safely driven more confidently instead of yielding in a particular situation.” Seems dicey. This could lead in the direction to a less safe Waymo. Since the counterfactual will be generated, I suspect that that the generations will be biased towards survivor situations where most video footage in its training data will be from environments where people reacted well not those that ended in tragedy. Emboldening Waymo on generated best case data. THIS IS DANGEROUS!!!
Not at all. It's not the counter-factual they're generating, it's the "too rare to capture often enough to train a response to" they're generating.
They're implying that without the model having knowledge, even approximate, of a scene to react to, it simply doesn't react at all; it simply "yields" to the situation until it passes. In my experience taking Waymo's almost daily this holds.
I would rather not have the Waymo yield to a tornado, rising flood-waters, or charging elephant...
Driving is always a balance between speed and safety. If you want ultimate safety you just sit in the driveway. But obviously that isn't useful. So functionally one of the most important things a self-driving system will decide is "how fast is it safe to drive right now". Slower is not always better and it has to balance safety with productivity.
Not entering a roundabout when it's clearly safe to do so is a mark against you at a driving exam. So would be always driving at 5mph. It's just not that simple.
now thx to this post (and the infra provider inclination to appeal to hacker news) we will never know if the model actually discovered the 50 spells or memorized it. Since it will be trained on this. :(
But what can you do, this is interesting
I understand this provides a conversation interface for interacting with internet scale data (ChatGPT), but is there any benefit to this over searching in Google then clicking on the top link, (avoiding the ad) clicking accept my cookies, reading the header, scrolling down, Xing out of premium subscription, reading rest of article, repeat for the 4 next links?
> Limitations: ... Finally, while we report strong results on individual datasets, we
have not yet demonstrated generalization to unseen data, an essential step toward general TSLMs.
From the paper itself...
Imagine you asked ChatGPT a question but it could only give you answers from a single blog.
It is designed to be fully modular with the ability to have a human/programmer in the loop at any stage of the process.
Enabling you to:
- effectively execute hierarchical orchestration :chains:
- visualize data in an organized manner
- dynamically classify items
- etc. (I would love to learn about what you may use it for!)
Excited to share the v0 of OpenGPTs-platform https://github.com/OpenGPTs-platform , an open-source project aimed at replicating and surpassing the "GPT store" with a focus on minimal abstraction, high modularity, and extensibility. Our platform is built to seamlessly integrate a wide array of tools, empowering the community to contribute and expand beyond the existing capabilities of conventional platforms. The initial release features an enhanced "Assistants API" with tools like "retrieval", identical to the one from the GPT store, and a custom "web_retrieval" tool designed for efficient retrieval of information from pre-indexed websites, developed in collaboration with https://huggingface.co/spaces/seansullivan/Better_Langchain_... .
Check out our project on GitHub and watch our demo on YouTube https://www.youtube.com/watch?v=yPdIEKb3jWc . Join our Discord community https://discord.gg/23aZEjyjp2 to get involved and help us shape the future of OpenGPTs!
reply