The name Anduril means “Flame of the West”, which sounds cool as hell. But it feels a little icky to me given the ideological proclivities of Palmer Luckey and Thiel.
It takes less than an hour of conversation with either, giving them a few tasks requiring logical reasoning, to arrive at that conclusion. If that is a strong position, it's only because so many people seem to be buying the common scoreboards wholesale.
That’s very subjective and case dependent. I use local models most often myself with great utility and advocate for giving my companies the choice of using either local models or commercial services/APIs (ChatGPT, GPT-4 API, some Llama derivative, etc.) based on preference. I do not personally find there to be a large gap between the capabilities of commercial models and the fine-tuned 70b or Mixtral models. On the whole, individuals in my companies are mixed in their opinions enough for there to not be any clear consensus on which model/API is best objectively — seems highly preference and task based. This is anecdotal (though the population size is not small), but I think qualitative anec-data is the best we have to judge comparatively for now.
I agree scoreboards are not a highly accurate ranking of model capabilities for a variety of reasons.
If you're using them mostly for stuff like data extraction (which seems to be the vast majority of productive use so far), there are many models that are "good enough" and where GPT-4 will not demonstrate meaningful improvements.
It's complicated tasks requiring step by step logical reasoning where GPT-4 is clearly still very much in a league of its own.
I would not say that OpenAI is all that far ahead of the competition. I would not count Google out, and I would be willing to bet that OpenAI/Microsoft + Google will be the duopoly of the AI age.
At this point Gemini is as good as GPT4 (and anecdotally I think it's better at many coding assistance tasks, based on my experiments on the LMSys chatbot arena). Sora is getting a lot of press for text-to-video, but Google's Lumiere has already been out for a few weeks and produces pretty good results.
I have no doubt that OpenAI has been cooking things up. GPT-4 is an older model which others are only just catching up to now. They have an A+ team, a giant war chest, and a lot of momentum. But just because they have momentum does not mean that they have a moat.
I'd be willing to wager that the top-of-the-line foundational models are going to converge and become indistinguishable for almost all tasks. Even the open foundation models (e.g. Mixtral) are getting really good. Foundation models are not moats.
The players that have moats are Microsoft (with deep enterprise software & B2B expertise, allowing them to sell AI-powered software & ward off competition from upstarts) and Google (where their decade of investment into custom silicon allows them to train & run inference for cheaper than anyone else, by far).
I don't have a large amount of time to devote to this.
But no, google is not out.
They just choke slammed ChatGPT with the Gemini announcement.
10m context with that retrieval rate is such a monstrous leap.
And to top it off, we got LargeWorldModel in the same week, capable of 1M token context with insane retrieval rate in the open source space.
So not only is the open source world currently technically ahead of ChatGPT, so is Google.
Which is why they had to announce SORA, because google's model is so far ahead of the competition.
That's also why it will probably be ages before we get access to SORA.
Now don't get me wrong, the average person can't afford 32 TPU's to run LWM, but we already have quants for it, which is a step towards enabling the average person (that somehow has 24-48gb of VRAM to get a taste of that power).
What is also striking is the fact that the new models are all multimodal as a standard. We not only leapfrogged in context size, but also in modalities.
The model seems to only benefit from having more modalities to work with.
I think the statement Bill Gates made claiming that "LLM's have reached a plateau" itself indicates they don't believe they can make more money from training better/larger models. Which indicates that they already did as well as they could with their existing people, and are now "years" behind google.
I never thought google could catch up, especially after their infamous "We have no moat" situation. But it seems they actually doubled down and did something about it.
To a lot of people, last Thursday was a very nihilistic day for Local Models, as the goalposts shifted from 128-200k context to 10M tokens with near perfect retrieval. It's literally insanely scary.
But luckily we got LWM, and that means we have only been 10xed.
Now the local people will work on figuring out how to bridge the gap, before being leapfrogged again. What is really insane is that, we have had LLAMA2 for over a year now, and nobody else figured out how to get this result from it, despite it being around so long.
I still believe there are modifications to the architecture of MoE that will unlock new powers that we haven't even dreamed of yet.
Sorry, this was supposed to be well thought out, but it turned more into stream of consciousness, and I honestly had no intention of disagreeing with you.
> But luckily we got LWM, and that means we have only been 10xed.
If I remember the paper correctly, it was something about a 4M context in there. So not 10x, but 2.5x.
> What is really insane is that, we have had LLAMA2 for over a year now, and nobody else figured out how to get this result from it, despite it being around so long.
This isn't true. For now, the task of extending context to 10M tokens is brute-forced by money (increased HW requirements for training and inference and increased training time are also a financial domain). And for now, there simply is no leapfrogging solution for open source or commercial models, which will decrease the costs by orders of magnitude.
VR is kind of a gimmick, but AR will, 100%, be the next frontier of human-computer interaction. It sparks the imagination in an extremely tangible way.
But IMO, attitudes like "it sparks the imagination" is why VR failed.
FB succeeded because it was in the right place at the right time. Sure, it positioned itself cleverly, but the moment in history was ripe for a social media company.
Things don't just succeed because they spark the imagination or because you work really hard at them: they succeed because all the factors beyond your control are aligned, too.
