well, I thought we’d get a break with these AI advancements in December with the holidays here and everything but I was wrong because Google just announced Gemini and we have a usable version of Gemini that we can play around with right now.
today in this Blog Post I’m going to dig into what’s new with Gemini what we know so far and How To Use Gemini AI & How It Performs it yourself let’s get into it
So December 6 2023 Google and Deep M introduced Gemini now this is a pretty hefty article with all the details about Gemini.
they’re calling this version Gemini 1.0 and it’s going to start in three different sizes Gemini Ultra which is their largest most capable model, Gemini Pro the best model for scaling across a wide range of tasks this is the one that we have access to now and then, Gemini Nano which which is the most efficient model for on device tasks and this one seems to be mostly designed for Mobile use cases.
when you look at something like GPT3 and GPT 4 these were initially trained from the ground up to be text models and just recently have been upgraded to accept both image’s Audio.
How Gemini AI Performs
now in benchmarking test, Gemini Ultra outperformed GPT 4 in most cases.
now keep in mind Gemini Ultra is the model that we don’t have access to yet this is the model we’ll likely get next year we have access to Gemini Pro right now which is a closer comparison to GPT 3.5 which you get inside of the free version of chat GPT.
here are some examples that they shared in some of their demos today in this example they gave a worksheet with a bunch of math problems on it and the math problems are already done with the handwriting and some of the work being shown and they asked to check the answers essentially
Google also put up this cool video on YouTube that shows all sorts of different ways that you can use the multimodality so check this out
Gemini Ultra is Gemini and Coding
Gemini Ultra is also really good at coding including competitive programming.
Now one thing to note about what we just watched over those last few demos that aren’t explicitly mentioned in any of the blog posts Google put out but were mentioned in a press briefing is that.
Images NOT Supported in Gemini AI
Now I wasn’t at that press briefing but Tech Crunch also confirmed this and another key limitation while the Gemini Ultra architecture supports image generation as does Gemini Pro in theory the capability won’t make its way into the productized version of the model at launch that’s perhaps.
Because the mechanism is slightly more complex than how chat GPT generates images rather than feed prompts to an image generator like Dolly 3 in Chat Egypt’s case.
Gemini outputs images natively without any intermediary step the article goes on to say that they didn’t provide a timeline as to when image generation might arrive only an assurance that the work is ongoing
Safety and Ethics of Gemini AI
The Google article claims that it was built with responsibility and safety at the core they’re adding new protections to account for Gemini’s.
so as we know open AI has kind of gotten into some hot water and a lot of people have tried to sue them recently for training on their data without approval and Google is keeping its mouth shut right now.
On where this training data came from whether it’s due to ethical implications or just because they want to keep it as proprietary as possible I’m not sure now at the Press briefing they did reveal that at least a portion of the data was from public web sources and that Google filtered.
it for quality and inappropriate material but they didn’t address the elephant in the room whether creators who might have unknowingly contributed to Gemini’s training data can opt out or expect SL to request compensation.
Gemini AI Availability
so theoretically you can open up a recording app in your Pixel 8 Pro and ramble for 20 minutes about some thoughts.
Then get a summary of it or set your phone in the middle of a conference table during a meeting record the whole meeting and then get a summary of that meeting directly from your phone and a version of Gemini has already been in use.
Inside of the Google search generative experience on December 13th, they’re going to start making Gemini Pro available via API for developers and again Gemini Ultra is coming soon.
Gemini AI is in Bard NOW
today Google also put out an article on their blog Bard gets its biggest upgrade yet with Gemini.
I do think when Gemini Ultra comes out it’ll probably be even more advanced than what they even announced right now and possibly probably will be even better than what we get with GPT4 that is assuming GPT4 doesn’t roll out some more amazing advancements between now and then we’ll just have to wait and see.
Is ChatGPT Necessary Anymore?
Even Mik here sums it up pretty well what changed today is the simple version there’s no reason to use the free version of chat GPT anymore because that uses GPT 3.5.