The "open" in OpenAI stands for "openly purchasable"
Gemini AI
241–250 of 1001 posts
Re: Gemini AI
#242I asked Bard, "Are you running Gemini Pro now?" And it told me, "Unfortunately, your question is ambiguous. "Gemini Pro" could refer to..." and listed a bunch of irrelevant stuff. Is Bard not using Gemini Pro at time of writing? The blog post says, "Starting today, Bard will use a fine-tuned version of Gemini Pro for more advanced reasoning, planning, understanding and more." (EDIT: it is... gave me a correct answer…
Your line of thinking also presupposes that Bard is self aware about that type of thing. You could also ask it what programming language it's written in, but that doesn't mean it knows and/or will answer you.
I had the same issue as OP. Initially Bard seemed clueless about Gemini, then:
Me: I see. Google made an announcment today saying that Bard was now using a fine-tuned version of their "Gemini" model
Bard: That's correct! As of December 6, 2023, I am using a fine-tuned version of Google's Gemini model ...
Re: Gemini AI
#243How do we know the model wans't pretrained on the evaluations to get higher scores? In general but especially for profit seeking corporations, this measure might become a target and become artificial.
Re: Gemini AI
#244Re: Gemini AI
#245So, better than GPT4 according to the benchmarks? Looks very interesting. Technical paper: https://goo.gle/GeminiPaper Some details: - 32k context length - efficient attention mechanisms (for e.g. multi-query attention (Shazeer, 2019)) - audio input via Universal Speech Model (USM) (Zhang et al., 2023) features - no audio output? (Figure 2) - visual encoding of Gemini models is inspired by our own foundational work o…
That's for Ultra right? Which is an amazing accomplishment, but it sounds like I won't be able to access it for months. If I'm lucky.
I hate this "tierification" of products into categories: normal, pro, max, ultra
Apple does this and it's obvious that they do it to use the "decoy effect" when customers want to shop. Why purchase a measly regular iPhone when you can spend a little more and get the Pro version?
But when it comes to AI, this tierification only leads to disappointment—everyone expects the best models from the FAANGO (including OpenAI), no one expects Google or OpenAI to offer shitty models that underperform their flagships when you can literally run Llama 2 and Mistral models that you can actually own.
Re: Gemini AI
#246One observation: Sundar's comments in the main video seem like he's trying to communicate "we've been doing this ai stuff since you (other AI companies) were little babies" - to me this comes off kind of badly, like it's trying too hard to emphasize how long they've been doing AI (which is a weird look when the currently publicly available SOTA model is made by OpenAI, not Google). A better look would simply be to sh…
It's worth remembering that AI is more than LLMs. DeepMind is still doing big stuff: https://deepmind.google/discover/blog/millions-of-new-materi...
Indeed, I would think the core search product as another example of ai/ml...
Re: Gemini AI
#247I've missed this on my initial skim: The one launching next week is Gemini Pro. The one in the benchmarks is Gemini Ultra which is "coming soon". Still, exciting times, can't wait to get my hands on it!
The Pro seem to be available in Bard already.
I've been asking Bard and it's telling me it's latest major update was September and it's backend is LaMDA... not sure if that means anything though
Re: Gemini AI
#248One of the topics I didn't see discussed in this article is how we're expected to validate the results of the output of the AI.
Really liked the announcement and I think this is a great step forward. Looking forward to use it. However I don't really see how we can verify the validity of AI responses with some statistical significance.
For example, one of the video demos shows Gemini updating a graph from some scientific literature. How do we know the data it received for the graph is accurate?
It feels like to me there is a missing prompt step not shown, which is to have a competing advisarial model be prompted to validate the results of the other model with some generated code that a human could audit.
Basically when humans work together to do the work, we review each other's work. I don't see why AIs can't do the same with a human additionally verifying it.
Re: Gemini AI
#249I did some side-by-side comparisons of simple tasks (e.g. "Write a WCAG-compliant alternative text describing this image") with Bard vs GPT-4V. Bard's output was significantly worse. I did my testing with some internal images so I can't share, but will try to compile some side-by-side from public images.
As it should! Hopefully Gemini Ultra will be released in a month or two for comparison to GPT-4V.
Re: Gemini AI
#250Earlier quoted context omitted.
That's for Ultra right? Which is an amazing accomplishment, but it sounds like I won't be able to access it for months. If I'm lucky.
The article says "next year" - so that could be as soon as January, right?
given how google has been functioning, probably as late as December :)