So they hired some guys from other AI companies.
The only universe they're going to end up understanding is the one inside Elon's head.
And Elon's recent polarising nature and the callous nature with which he disbanded the Tesla Supercharger team means that truly talented people aren't going to be as attracted to him as in his early days. They are only going to be there for the money.
Using it for knowledge is bonkers.
Why not buy some educational textbook company and use 99.9% correct data? Oh and use RAG while you are at it so you can point to the origin of the information.
The real evolution still has to come though, we need to build a reasoning engine (Q*?) which will just use RAG for knowledge and language models to convert its thought into human language
In other words - say you have a model that is semi-smart, often makes mistakes in logic, but sometimes gives valid answers. You use it to “brainstorm” physical equations and then use formal provers to weed out the correct answer.
Even if the llm is correct 0.001% of the time, it’s still better than the current algorithms which are essentially brute forcing.
If you need to effectively provide this whole secondary dataset to have better answers, what value do the tweets add to training other than perhaps sentiment analysis or response stylization?
interesting.
It's 6B down the drain. Saying grok 1.5 is competitive is a joke, if it was any good it would be ranked well in chatbot arena (https://chat.lmsys.org/). Elon is a master in hyping underperforming things and this is no exception.
It would be hard to judge rate of improvement at this point, since the company has only been around for 1.25 years, and grok 1.5 is yet to be released for general access.
I wonder why
a16z invested $350 in Adam Neumann's real estate venture - after WeWork. VCs will absolutely knowingly invest on hype if they think it's going to last long enough for them to cash out with great returns.
> Our team is led by Elon Musk, CEO of Tesla and SpaceX. Collectively our team contributed some of the most widely used methods in the field, in particular the Adam optimizer, Batch Normalization, Layer Normalization, and the discovery of adversarial examples. We further introduced innovative techniques and analyses such as Transformer-XL, Autoformalization, the Memorizing Transformer, Batch Size Scaling, μTransfer, and SimCLR. We have worked on and led the development of some of the largest breakthroughs in the field including AlphaStar, AlphaCode, Inception, Minerva, GPT-3.5, and GPT-4.
They are already competitive despite the late start: https://x.ai/blog/grok-1.5v