3 ms·
https://nautil.us/deep-learning-is-hitting-a-wall-238440/ https://nautil.us/deep-learning-is-hitting-a-wall-238440/ Gary Marcus said that Deep Learning was hit
by lairv 1y ago
https://nautil.us/deep-learning-is-hitting-a-wall-238440/ https://nautil.us/deep-learning-is-hitting-a-wall-238440/
Gary Marcus said that Deep Learning was hitting a wall 1 month before the release of DALLE 2, 6 months before the release of ChatGPT and 1 year before GPT4, arguably 3 of the biggest milestones in Deep Learning
- brazukadev 1y agoSam Altman said GPT-3 was dangerous and openai should be responsible for saving the humanity.
- CamperBob2 1y agoWorth pointing out that no one who doesn't work at a frontier lab has ever seen a completely un-nerfed, un-bowdlerized AI model.
- brazukadev 1y agoBut we know that ChatGPT 5 is better than anything un-nerfed, un-bowdlerized from 2 years ago. And is not impressive.
- ACCount37 1y agoThere are some base models available to the public today. Not on "end of 2025 frontier run" scale, but a few of them are definitely larger and better than GPT-3. There are some uses for things like that. Not that the limits of GPT-3 were well understood at the time. We really had no good grasp of how dangerous or safe something like that would - and whether there are some subtle tipping point that could propel something like GPT-3 all the way to AGI and beyond. Knowing what we know now? Yeah, they could have released GPT-3 base model and nothing bad would have happened. But they didn't know that back then.